UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The aerobiology of the aecial state of the commandra blister rust, Cronartium comandrae Peck, in Alberta. Powell , John Martin 1969-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


831-UBC_1969_A1 P69.pdf [ 25.1MB ]
JSON: 831-1.0103999.json
JSON-LD: 831-1.0103999-ld.json
RDF/XML (Pretty): 831-1.0103999-rdf.xml
RDF/JSON: 831-1.0103999-rdf.json
Turtle: 831-1.0103999-turtle.txt
N-Triples: 831-1.0103999-rdf-ntriples.txt
Original Record: 831-1.0103999-source.json
Full Text

Full Text

THE AEROBIOLOGY OF THE AECIAL STATE OF THE COMANDRA BLISTER RUST, CRONARTIUM COMANDRAE PECK, IN ALBERTA by JOHN MARTIN POWELL B.Sc, University of London, 1956 M.Sc, McGill University, 1959 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of BOTANY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April, I969 0 John Martin Powell 1969 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission Department of Z> 0>"T A NY  The University of British Columbia Vancouver 8, Canada i ' ABSTRACT Cronartium comandrae Peck is a heteroecious native rust which is damaging to several Pinus species in North America. It grows peren nially in the living bark of hard pines producing pycniospores and aeciospores, and develops annually on species of Comandra and Geocaulon producing urediospores, teliospores and basidiospores. Studies were carried out to determine the effect of various meteorological and biological environmental factors on the aeciospore aerobiological phase of the rust. The main aspects considered included the factors affecting aeciospore production and release; and the factors affecting aeciospore transport, dispersion, deposition, germination and viability. Aeciospore production occurred from mid May to late August, with the peak period between late May and mid June. An average aecium produced spores for 35-50 days, and individual cankers produced spores for up to 95 days. Much variation occurred from year to year and between trees. Aecial production was interfered with by the activity of rodents, insects and other fungi on the canker and through resinosis, which pro bably accounted for a 50-55$ reduction of the potential aeciospore pro duction in any year. Fresh rodent damage was recorded on U0-52$ of the cankers at about 20 locations in the years I966 to -I968, insect damage on 39-^6$, Tuberculina and Cladosporium infection on YJ-33?o, and resinosis on 67-71$. Rodent damage was mainly caused by squirrels. Some 6k micro-floral organisms were isolated from the cankers and spores. Tuberculina  maxima was mainly responsible for killing the infected canker bark and ii greatly reduced the production of aeciospores. Penicillium spp. and an undescribed Cladosporium were also prominent on the cankers but played a lesser role in reducing production. One hundred and seventeen species of insects, mites and spiders were associated with the rust canker. Epuraea  obliquus, an unidentified cecidomyiidae and Paracacoxenus guttatus were true mycetobionts. Various other species caused damage to the canker, including Dioryctria, Pissodes schwarzi and Cylindrocopturus deleoni. Daily aeciospore periodicity showed spores to be normally dis persed between 0800 and 1900 hours, with some evidence of a double peak between 1000 and 1600 hours, and little dispersal between 2000 and 0700 hours. Turbulent atmospheric conditions were associated with all peak aeciospore concentrations. Heavy rains initially increased spore concen trations, but no dispersal occurred during long humid cool periods. Spore deposition concentrations were very steep close to source, and showed a typical hollow curve deposition gradient. Spore concentrations from artificial release points were similarly reduced with distance from source, largely by diffusion. Aeciospores had an average velocity of 3.23 cm/sec in calm air. Rapid depletion of spore concentration occurred under the forest canopy, mainly by sedimentation, although aeciospores had good im paction efficiency. Aeciospores germinated on water agar over the temperature range 1-30°C, with optimum for germination and germ tube growth close to 15°C. Most aeciospores germinated within k-5 hours, with a reduction in rate of germ tube elongation after 8 hours and little increase after 2k hours. Free water was necessary for germination; all spores swelled prior to germination. Aeciospores germinated equally well in the dark and light, iii and germinated over the pH range k.5-8. Germination response on sugar media was better than on some other media, but addition of alternate host material to media did not improve germination. Daily aeciospore collections gave high germination percentages for 2-k weeks, but much lower percentages during later sporulation period. Aeciospore germination was lower from exposed than protected aecia, and wet spores germinated poorly. Viability was reduced by contamination from associated fungi. Aeciospores lost viability very rapidly when ex posed to temperatures above 25°C; temperatures close to 0°C were most favourable. High humidity affected viability, but ultra dry conditions were also adverse. Direct sunlight reduced viability rapidly. Generally, daily conditions favouring dispersal were least favourable for germination and viability retention. iv TABLE OF CONTENTS Page ABSTRACT i TABLE OF CONTENTS iv LIST OF TABLES x LIST OF FIGURES xiii ACKNOWLEDGMENTS xxiv INTRODUCTION 1 THE ORGANISM 5 HISTORY OF THE FUNGUS 5 DISTRIBUTION AND HOSTS 10 Distribution of the rust on Pinus 11 Distribution of the rust on Santalaceae Ik LIFE CYCLE 2k SYMPTOMS 37 DAMAGE kl AREA OF STUDY 1+8 LOCATION, PHYSIOGRAPHY AND GEOLOGY 4SOILS 50 CLIMATEVEGETATION 6l STUDY LOCATIONS 2 V AECIOSPORE PRODUCTION 65 PERIODS OF AECIOSPORE PRODUCTION AND ENVIRONMEfflTAL FACTORS AFFECTING SPORULATIONMethods 65 Results 6 Discussion 7BIOLOGICAL FACTORS AFFECTING AECIOSPORE PRODUCTION 78 MethodsResults 82 MicrofloraMicrofauna 91 Rodents 102 Response of the tree 107 Discussion 109 AECIOSPORE DISPERSAL 128 METEOROLOGICAL FACTORS AFFECTING DISPERSAL 12Methods and Materials 128 Experimental sitesSpore collectors 131 Meteorological instruments 136 Results 139 Diurnal periodicity of meteorological factors, and the microclimate of study location 1 139 Diurnal spore periodicity l4l Effect of relative humidity and temperature 146 Effect of rainfall 150 vi Effect of dew 15I4. Effect of wind 155 Availability of spores 15Seasonal spore periodicity 6 AECIOSPORE DISPERSAL FROM A NATURAL POINT SOURCE 159 Methods and Material 15Results 162 OTHER DATA ON DISTANCE OF AECIOSPORE DISPERSAL 171 AECIOSPORE DISPERSAL EXPERIMENTS FROM POINT SOURCES 173 Materials and Methods 17Experimental arrangement 175 Experimental procedure 9 Analysis of data l8l Results 18U Concentration patterns l8k Change of concentration with distance l8h RATE OF FALL OF AECIOSPORES IN CALM AIR 197 Materials and Methods 19Results 200 AECIOSPORE DISPERSAL DISCUSSION 205 Meteorological factors and spore periodicity 205 Spore dispersal 213 AECIOSPORE GERMINATION 228 FACTORS AFFECTING AECIOSPORE GERMINATION 230 Methods and Materials 23vii Spore material and methods of handling 230 Methods for exploratory experiments 231 Methods for subsequent experiments . 23^ Results 2k0 Effect of temperature on germination 2k0 Effect of humidity on germination 2k7 Swelling of spores on ... liquid media 2^9 Effect of hydration of spores on germination 251 Effect of light on germination 252 Effect of hydrogen ion concentration on germination 255 Effect of substrate on germination 257 a. Various agar media 25b. Addition of sucrose to media 259 c. Presence of host leaves on, or leaf extracts in the media 2.6o Discussion 262 DAILY AECIOSPORE GERMINATION 270 Methods and MaterialsResults and Discussion 272 AECIOSPORE VIABILITY 283 Methods and Material 28Results 287 Effect of temperature on spore viability 287 Effect of humidity on spore viability 292 • Effect of light on spore viability 29U viii Discussion 297 SUMMARY AM) CONCLUSION 302 LITERATURE CITED 321 APPENDIX I 353 APPENDIX II 6 ix LIST OF TABLES TABLE Page I Monthly and annual climatic summaries for Kananaskis, 53 lat. 51°02'W, long. 115°03'W; elev. k,560 ft MSL, for the period of record (1939-1968). II Monthly and annual climatic summaries for selected 54 stations in the Marmot Creek Research Basin, lat. 50°57'N, long. 115° 10% elev. 5,300-^8,000 ft MSL (1962-1967). III Monthly temperature and precipitation averages at Kan- 55 anaskis Boundary Ranger Station, lat. 50°55'N, long. 115°08'W, elev. 4,800 ft MSL (1962-1968)., and Pigeon Mountain Lookout, lat. 51°03'N, long. 115°04% elev. 6,000 ft MSL (1960-1968), compared with those at Kananaskis for a similar period (1963J-I968). IV Date of beginning and end of Cronartium comandrae 67 aeciospore production, and main sporulation period at a number of locations in the years 1964 - 1968' inclusive. V Date of beginning and end of spore production, and 68: total number of days of production from individual cankers at location 1 from 1965 to 1968 inclusive. Main spore production period is included for 3 to 6 cankers for the years 1965 to I967. VI Average number of days and range of days of aecio- 74 spore production for individual pustules on several cankers in 1966 and I967 at location 1. VII Summary of sequence of sporulation for 30 and 20 74 pustules on two cankers at location 1 in 1967. VIII The recorded incidence of Tuberculina maxima on 83 Cronartium comandrae cankers on lodgepole pine and the total number of active and inactive cankers ob served at various locations in southwest Alberta during the years 1964 to 1968. IX The incidence of Tuberculina maxima on observed 87 active and inactive Cronartium comandrae cankers at 7 locations during the years 1964 to 1968. X X The number of Cronartium comandrae aecial cankers (l) producing aeciospores, (2) sporulating but infected with Tuberculina, (3) with inactive aecial zones in fected with Tuberculina, (k) with inactive or dead cankers, at two locations during the years 1966 to 1968. XI The recorded incidence of Cladosporium tax. sp. 1 on Cronartium comandrae cankers observed at various locations in southwest Alberta during the years I965 to 1968. XII The incidence of insect damage on Cronartium comandrae cankers on lodgepole pine and the num ber of cankers observed at various locations in southwest Alberta during the years 1964 to 1968. 92 XIII The incidence of new rodent damage on Cronartium 10k comandrae cankers, and the number of cankers observed at various locations in southwest Alberta during the years 1966 to I968. XIV The percentage incidence of fresh rodent chewing 106 of Cronartium comandrae cankers at selected locations during the years 1966 to 1968. XV XVI The incidence of fresh resinosis on Cronartium  comandrae cankers on lodgepole pine and the number of cankers observed at a number of loca tions in southwest Alberta during the years 1966 to 1968. The hourly mean temperature, relative humidity and wind speed at two canker stations, compared with records from the Bay station for temperature and humidity, and from Kananaskis for wind speed, for the period May 17 to July 11, 1966. 109 iko XVII Number of aeciospores deposited on natural leaf surfaces collected along the four cardinal radii at various distances from a sporulating Cronartium  comandrae canker on July 7, 1968, at location 3-170 XVIII Average number of aeciospores deposited per square centimeter on spore collector coated slides at dis tances of 5, 10 and 15 feet along eight radii a-round a sporulating Cronartium comandrae canker at location 33 on 9 days between June 2k and July 15, 1968. 170 xi XIX Particulars of experiments on dispersion of Cron- l80 artium comandrae aeciospores from a point source. XX Total number of spores trapped at one foot above 185 ground at each distance on all radii. XXI Total number of spores trapped at 5 and 10 feet a- 185 bove ground at 20, 50, 100 and 150 feet on five radii. XXII Results of Experiment X, showing number of spores 190 trapped on an area of 13-5 sq cm at the one foot le vel at each sampling point. XXIII Comparison of the observed and expected concentra- 195 tion of spores at four distances from the 5 foot release point during eight tests. XXIV Percentages of individual tests of aeciospores 201 of Cronartium comandrae deposited 3 meters from point of liberation on glass slides exposed suc cessively for 15 second periods in a closed cylin der. XXV Average rate of fall in still air of ten aecio- 202 spore tests. XXVI Number of aeciospore clumps deposited 3 meters from 203 point of liberation on glass slides exposed successively for 15 second periods in a closed cylinder. XXVII Average rate of fall in still air of dry and wet 20k aeciospores during two tests. XXYIII Average length and width (|a) of aeciospores 205 deposited on slides during the dry and wet-spore tests of rate of fall. XXIX Average number of germ tubes per spore, and the 2^3 range of the average for four test series at various temperatures. XXX Average length and width (in microns) of 25 aecio- 250 spores from 5 fresh or stored samples. Measure ments were made on dry spores and when spores were placed on a Czapek-Dox agar medium. XXXI Average germination percentages for spores 252 stored for various periods of time at -k°C, and then germinated, dry or after 2k hours hy dration in a saturated humidity, on Czapek-Dox agar at 15°'C. Xll Effect of light and dark conditions on percent germination and germ tube growth of three series of aeciospore samples, germinated on Czapek-Dox agar at 15-C or under fluctuating outside tem peratures for 2k hours. Effect of dark and three colored light wave bands on percent germination and germ tube growth of three series of aeciospore samples, germinated at 15°C on Czapek-Dox agar for 2k hours. Average percentage and range of germination of two series of aeciospores after 2k hours on different media at 15°C. Average number of germ tubes and lengths of germ tubes per spore for three samples on seven different media after 2k hours at 15°C. Average germination percentages of five aecio spore samples stored at six temperatures and germinated on water agar at 15°C after various intervals of storage. Average germination percentages of aeciospore samples stored for four to six days at various temperatures and germinated on water agar at 15°C Average percentage germination of a number of spore samples given two methods of rapid cooling and stored at -20°C for periods between 7 and 18 days, and then germinated on water agar at 15°C for 2k hours. A portion of the liquid nitrogen cooled spores was rapidly thawed for one minute at 38°C prior to seeding on the agar. Average number of germ tubes per spore and average length of longest germ tube in three dry-humid treatments after various intervals of storage. xiii LIST OF FIGURES Figure Page 1 Distribution of Cronartium comandrae on 15 Pinus species in western Canada. 2 Distribution of Cronartium comandrae on 16 Pinus species in eastern Canada. 3 Aerial stems of Geocaulon lividum, with 17 axillary greenish flowers. h Aerial stems of Comandra umbellata ssp. 17 pallida, with terminal white flowers about to break open. 5. Series of aerial stems of Comandra umbellata 17 ssp. pallida, which branch just below the soil, and are attached to an underground rhizome. Some aerial stems dead from pre vious year. 6 Large haustorium of Comandra umbellata ssp. 17 pallida attached to Populus tremuloides Michx. root, with smaller haustorium on smaller root. 7 Distribution of Cronartium comandrae on the 22 Santalaceae, Comandra umbellata and Geocaulon ' lividum, in western Canada. 8 Distribution of Cronartium comandrae on the 23 Santalaceae, Comandra umbellata and Geocaulon lividum, in eastern Canada. 9 Transverse sections taken from the central 26 portion of four Cronartium comandrae cankers on lodgepole pine showing eccentric growth caused by the rust. Note the resin impreg nation of the outer sapwood. 10 Transverse sections taken from the central 26 portion of five Cronartium comandrae cankers on lodgepole pine showing eccentric growth caused by the rust. xiv Transverse section taken through the aecial zone of a Cronartium comandrae canker showing the aecia situated in the bark tissues. Note the resin impregnation of the bark between aecia, and the swelling of the outer ring-of sapwood caused by the fungus. Pycnial drops on the pycnial zone of a Cronar tium comandrae canker on the stem and branches of a lodgepole pine. Note the swelling of the stem associated with the canker. Group of pyriform pycniospores of Cronartium  comandrae. X 2500. Pyriform aeciospores of Cronartium comandrae showing size and shape variation of the spore tail. X 700. Pyriform aeciospores of Cronartium comandrae showing the spore wall ornamentation. X 700. Canker.of Cronartium comandrae on a young branch of Pinus contorta, with the peridia of the aecia still unruptured and forming typical blisters pushing through the bark. Canker of Cronartium comandrae with abundant ruptured aecia covering the aecial zone. Bright orange-colored aeciospores of Cronartium  comandrae covering the exposed aecia of the canker. Young canker of Cronartium comandrae on Pinus  contorta with rough bark covering the old aecial zone around the branch stub that acted as entry point to the stem. Abundant aecia were produced around the old aecial zone. Slight hypertrophy of the young Pinus contorta stem caused by Cronartium comandrae, with characteristic rough, cracked bark, in the older canker area caused by aecia rupturing. Typical basal stem canker of Cronartium coman drae showing older rough bark zone and current year aecial zone with many dispersed aeciospores caught in the crevices of the bark. XV 22 Aeciospores of Cronartium comandrae, stained' 33 with HC1 - Giemsa, each with two nuclei. X 700. 23 Aeciospore of Cronartium comandrae, stained with 33 HC1 - Giemsa, with two nuclei. X 1200, 2k Initiation of aeciospore germination, with four 33 germ tubes developing through the spore wall, after 1 hour. X 1200. 25 Cronartium comandrae aeciospore with several 33 germ tubes, but only one germ tube well developed after 6 hours. Note the short-multiple branches on the non-septate germ tube. X 380. 26 Group of germinated aeciospores with typical 33 branching of germ tubes after 2k hours. X 125. 27 Aeciospore of Cronartium comandrae with one 3^ developed germ tube, after 3 hours. Note the non-septate condition of the germ tube and that the two nuclei have migrated towards the tip of the germ tube. X 830. 28 Germ tubes of two Cronartium comandrae aecio- - 3^+ spores after 3 hours, showing development of a globose and club-shaped appressorium on the developed germ tube, into which the two nuclei have migrated. X 700. 29 Germ tube development of a Cronartium comandrae 3^+ aeciospore after 3 hours, showing a globose ap pressorium with two nuclei and an infection peg. x 1080. 30 Germ tube development of two Cronartium coman- 3^ drae aeciospores after 3 hours, showing two types of appressoria each with two nuclei and an infection peg. X 6l0. 31 Uredia of Cronartium comandrae on central portion 35 of a Comandra umbellata ssp. pallida leaf. Dark structures on bottom are young telia. 32 Telia of Cronartium comandrae on both surfaces 35 of a portion-of a Comandra umbellata ssp. pallida leaf. xvi Branch infection of Cronartium comandrae on Pinus contorta, with mycelium causing swelling in the main stem around the branch stub. Tree with dead spike-top above a canker of Cronartium comandrae girdling the tree; lower branches are progressively killed by downward growth of rust. Heavy resinosis forming a dried crust over a portion of a Cronartium comandrae stem canker. Note rough bark of old branch canker (upper left) which served as entry for the rust to the stem. Annual rodent damage on a large Cronartium  comandrae stem canker on Pinus contorta. Note the strip of dried dead bark not removed on each annual visit, and the abundant exudation of resin. Rodentcdamage on a Cronartium comandrae stem canker around a branch which acted as an entry point for the rust into the stem. Sporu lation of the aecial zone can be seen outside the chewed area. Map of the lower Kananaskis River Valley, Alberta, showing the locations of the main study areas. Study location 1, on the Kananaskis Forest Experiment Station, near the northeast shore of Barrier Lake, showing the location of infected lodgepole pine, Comandra plant plots, and in struments used during the study. Summary of the daily maximum and minimum temperatures and relative humidity taken from hygrothermograph records, the daily rainfall, and the phenology of the aecial, pycnial, uredial and telial states of Cronartium  comandrae, May to August in I965 and 1966. Summary of the daily maximum and minimum tem peratures and relative humidity taken from hy grothermograph records, the daily rainfall, and the phenology of the aecial, pycnial, uredial and telial states of Cronartium comandrae, May to August in I967 and 1968. XV11 A plastic screening cylindrical-sleeve cage used for collecting insects from Cronartium  comandrae cankers on stems of Pinus contorta. Rough, cracked bark of the aecial zone of a Cronartium comandrae canker infected with the purple mold, Tuberculina maxima, which is conspicuous as a darker area where the surface bark has been removed, or cracked. Basal canker of Cronartium comandrae with typical rough bark in aecial zone and showing evidence of insect damage. Note exit holes and Lepidoptera frass at top of canker, and further frass in lower rough zone. Pupal chambers of Pissodes schwarzi scored into the sapwood throughout the Cronartium  comandrae canker area on small Pinus contorta stem. View of site no. 2 at study location 1, showing a 24-hour impaction spore collector by Cronartium comandrae canker no. 2721 on a small Pinus contorta, and instruments for recording the weather. Instruments are, from left to right, black porous disc atmometer, instrument shelter containing hygrothermograph, Wallin-Polhemus dew duration recorder on ground, mast with anemometer cups for wind speed recorder and spore collector. (Rain gauge is out of the picture) Standard Stevenson screen in an opening, con taining hygrothermograph and thermometers used as reference weather station at study area 1, with a bi-metal actinograph for recording in coming radiation seen at the back. A record ing rain gauge and a wind direction recorder (out of the picture) were maintained at this open site. A Hirst spore trap with sampling orifice one foot above ground, close to a Comandra um bellata plot. xviii 49 Seven-day pollen sampler of the Sarvas type, 130 used to collect aeciospores at set distances from sporulating Cronartium comandrae cankers. Spores pass through the sampler orifice and are deposited on a vaseline-coated'sampling band placed around a clock-driven drum housed within the intake cylinder. 50 A 24-hour impaction spore collector, with the 133 sealed lid removed to show the acrylic plastic disc holding 24 microscope slides upon which spores are deposited, and the fan for drawing air through the box at a controlled rate. 51 A 24-hour impaction spore collector in opera- 133 tion, with its sampling orifice placed close to the sporulating surface of a Cronartium  comandrae canker. 52 Central instrument power and recording box, 133 with Thorntbmte four-unit wind speed register recorder (left), wind direction recorder, and digital printout recorder with polaroid camera for wind speed recorder system (right). Behind are the transformers and rectifiers for reducing the 110 volt power supply and controlling an output of 12 volts for operation of wind in struments, dew instruments and 24-hour impaction spore collectors. A standby 12 volt battery is also present. 53 Set of anemometer cups of the Thornthwaite 133 wind speed register recorder system operated at canker height near a sporulating Cronartium  comandrae canker. 54 Mean diurnal aeciospore periodicity curves 143 for three cankers over varying periods of trapping in the years 1964 to I967, expressed as a percentage of the peak geometric mean hourly concentration. 55 Frequency of hourly maximum aeciospore concen- l45 trations from three Cronartium comandrae can kers on 91 dry and 40 rainy days in 1966 and 95 dry and 29 rainy days in 1967, during the main spore production periods. 56 Average hourly aeciospore concentrations 147 collected from three Cronartium comandrae cankers in June 1966. xix Average hourly aeciospore concentrations collected from three Cronartium comandrae cankers from June 6 to 25, I967. Average hourly aeciospore concentration col lected from Cronartium comandrae canker no. 2713 during periods without rain (average 19 to 2k days) in June 1966, compared with the aver age hourly air temperature, relative humidity and wind speed. Hourly number of aeciospores collected from Cronartium comandrae canker no. 25l6 during the period June 7 to July 3, I965, related to hourly air temperature, relative humidity, rainfall, incoming radiation and wind speed. Hourly number of aeciospores collected from Cronartium comandrae canker no. 2721 during the period May 29 to June 95 1966, related to air temperature, relative humidity, rain fall, incoming radiation and wind speed. Daily number of aeciospores collected from two Cronartium comandrae cankers during the 1966 spore production period, plotted on semi-log scale. Daily number of aeciospores collected from Cronartium comandrae canker no. 2713 during I965, I966 and 1967, plotted on semi-log scale. Pattern of aeciospore deposition around Cronartium comandrae canker no. 2713 during a ICH5- hour period on June 23, 1967-The average percentage change of aeciospore deposition concentration with distance on eight radii around Cronartium comandrae canker no. 2713 on June 23, 1967, and the percentage change on the southeast radii, plotted on semi-log scale. Pattern of aeciospore deposition around Cronartium comandrae canker no. 2689, on three days (0800 to IcOO'J-hours) in June 1968. XX The average percentage change of aecio spore deposition with distance around Cronartium comandrae canker no. 2689, on three days (0800 to l600 hours) in June 1968, plotted on semi-log scale. Spore-ejector used for liberating aecio spores in dispersal experiments from point source. Compressor unitvwith connecting rubber tubing, for regulating air flow for re lease of aeciospores in dispersal experiments. Wind-directional plastic slide holder spore collectors (without microscope slides) used for collecting aeciospores in dispersal experiments and from natural sources. Portion of spore collector network, with collectors on different radii at various distances from release point. Wind direction vane and wind speed anemometer can be seen to the left of the stand for the spore-ejector. Note spore collectors at 5 and 10 feet on the masts. Portion of spore collector network with collectors on different radii at various distances from spore-ejector at release point. Wind speed anemometer at left of spore-ejector. Diagram of point source aeciospore dispersal experiment sampling grid, with 113 sampling collectors laid out for southwest winds. Spore collectors were also located at the 300 and kOO foot distances along the mid-line. Aeciospore concentration patterns at three sampling heights for a spore release one foot above ground (Experiment VII). Aeciospore concentration patterns at three sampling heights for a spore release five • feet above ground (Experiment X). xx i Composite aeciospore concentration patterns at three sampling heights for all eight spore releases five feet above ground (Experiments II - VI, VIII - X). Vertical profiles of aeciospore concen tration patterns along the mid-line of the sampling network for two spore re leases at one foot (Experiments I and VII), and two spore releases five feet above ground (Experiments V and X). Vertical profiles of aeciospore concen tration patterns along the mid-line and for k5° and 90° sectors of the sampling network, for one spore release at five feet above ground (Experiment VIII). Composite vertical profiles of aeciospore concentration patterns along the mid-line and for ^5° and 90° sectors of the sam pling network, for all eight spore releases five feet above ground (Experiments II -VI, VIII - X). Effect of temperature on per cent germina tion of three Cronartium comandrae aecio spore samples after 2k hours on water agar (pH 6.8). Effect of temperature on Cronartium com andrae aeciospore germination. Average percent germination of five series and various numbers of replicas with three to five samples each, after 2k hours on water agar. Average length of. germ tubes of three Cronartium comandrae aeciospore samples germinated oh water agar (pH 6.8). aft er 2k hours at various temperatures. (a) Length of longest germ tube per spore, (b) Length of all germ tubes per spore. Influence of various temperatures on rate of Cronartium comandrae aeciospore germina tion after various time intervals on water agar. xxii 83 Average rate of germination of three Conartium 246 comandrae aeciospore samples at different tem peratures and after various time intervals on water agar. 84 Average length of the longest germ tube from 248 three to five Cronartium comandrae aeciospore samples germinated on..water agar at different temperatures and after various time intervals. 85 Effect of hydrogen ion concentration on germ- 256 ination of Cronartium comandrae aeciospores after 24 hours at 15 and 20°C. 86 Effect of hydrogen ion concentration on 256 germination of Cronartium comandrae aecio spores after 24 hours at 5, 15 and 25°C. Spore samples were stored for 4 and 6 weeks at 0°C prior to use. 87 Average daily per cent germination of five 273 to seven Cronartium comandrae aeciospore samples after 24 hours on water agar at 15°C, for the years 1965, 1966 and 1967. 88 Daily per cent germination of aeciospores 275 from two Cronartium comandrae cankers for the years 1965, 1966 and 1967. 89 Variation in the daily per cent germination 276 of aeciospores from seven Cronartium com-drae cankers in I965. 90 Daily per cent germination of aeciospores 277 from five individual aecia of a Cronartium comandrae canker during the 1966 sporulation period. 91 Daily per cent germination of aeciospores from 278 five individual aecia of a Cronartium comandrae canker during the I967 sporulation period. 92 Average per cent germination of three Cronartium 288 comandrae aeciospore samples stored at three temperatures, and germinated on water agar at 15°C after various hours of storage. XX 111 Average per cent germination of five Cronartium comandrae aeciospore samples stored in a wet, dry or ultra dry atmos phere, and germinated on water agar at 15°C after various intervals of storage. Effect of exposure of Cronartium comandrae aeciospores to direct light for varying lengths of time on clear sunny days (June 16, and July 5, 1966; June 23, 1967), and on a cloudy day (June 16, 1967). xxiv ACKNOWLEDGMENTS I wish to express my sincere gratitude to the late Dr. J. E. Bier for his enlightened guidance and communicative enthusiasm throughout-much of this investigation. I wish also to thank Dr. R. J. Bandoni, Dr. A. L. Farley, Dr. P. G. Haddock, Dr. T. M. C. Taylor, Dr. G. H. N. Towers, and Dr. B. J. van der Kamp, for their counseling and assistance in the preparation of the thesis. Thanks are extended to the Canada Department of Fisheries and Forestry for permission to use, for this thesis, data gathered in the course of work on a Department project. Appreciation is expressed to Mr. W. Morf and Mr. L. S. Skaley, technicians of the Forest Research Laboratory, Calgary, Alberta, for assistance in the collection of data. Thanks are due to Dr. B. C. Sutton, Forest Research Laboratory, Canada Department of Fisheries and Forestry, Winnipeg, Manitoba, for the identification of the majority of the fungi. Other specialists who con tributed to the identification of various floral groups are: Dr. A. M. Adams, Horticultural Experiment Station, Ontario Department of Agriculture, Vineland Station, Ontario; Dr. F. D. Cook, Department of Soil Science, University of Alberta, Edmonton, Alberta; Dr. Y. Hiratsuka, Forest Research Laboratory, Canada Department of Fisheries & Forestry, Calgary, Alberta; and Dr. D. B. Prest, Department of Bacteriology, University of Wisconsin, Madison, Wisconsin. Many institutions provided Cronartium comandrae ma terial for the distributional aspects of this study, and I would like to especially thank the curators of the following herbaria for their assis tance: Arthur Herbarium, Purdue University, Lafayette, Indiana; The xxv National Fungus Collections, Beltsville, Maryland; The New York Botanical Garden, New York; Mac Donald College, Ste. Anne-de-Bellevue, Quebec; The W. P. Fraser Memorial Herbarium, University of Saskatchewan, Saskatoon, Saskatchewan; University of Alberta, Edmonton, Alberta; Mycological Herbarium, Canada Department of Agriculture, Ottawa; Ontario; and the Mycological Herbaria, Canada Department of Fisheries and Forestry, at Calgary, Quebec, Victoria and Winnipeg. I am indebted to the following taxonomic specialists of the Entomology Research Institute, Canada Department of Agriculture, Ottawa, Ontario, who made authoritative identifications of faunal material: Dr. E. C. Becker, Dr. D. Brown, Mr. W. J. Brown, Dr. J. M. Campbell, Dr. T. N. Freeman, Dr. M. Ivanochko, Mr. G. Lewis, Dr. E. E. Lindquist, Miss M. R. MacKay, Mr. J. E. H. Martin, Dr. W. R. M. Mason, Dr. J. F. McAlpine, Mr. C. D. F. Miller, Dr. A. Mutuura, Dr. 0. Peck, Dr. B. V. Peterson, Dr. W. R. Richards, Mr. R. de Ruette, Dr. L. K..Smith, Dr. J. R. Vockeroth and Mr. G. S. Walley. Other specialists who contributed to the identification of various faunal groups are: Dr. G. A. Bradley, Forest Research Laboratory, Canada Department of Fisheries & Forestry, Winnipeg, Manitoba; Dr. C. W. O'Brien, Department of Entomology, Univer sity of California, Berkeley, California; Dr. C. V. G. Morgan, Entomology Laboratory, Canada Department of Agriculture, Summerland, British Colum bia; Dr. C. T. Parsons, Manchester Depot, Vermont; Mr. B. A. Sugden, Forest Entomology Laboratory, Canada Department of Fisheries & Forestry, Vernon, British Columbia; Dr. A. L. Turnbull, Department of Biological Science, Simon Fraser University, Burnaby, British Columbia; and Dr. T. A. Woolley, Department of Zoology, Colorado State University, Fort Collins, xxv i Colorado. Thanks are extended to my wife, Margaret, who proof-read the manuscript during its preparation, and to Miss Barbara Brown who typed the final manuscript. - 1 -INTRODUCTION A knowledge of the effect of environment on a pathogen during that part of its life cycle when it is outside the host is of prime im portance to an understanding of the factors influencing the spread and intensification of an airborne disease. For a pathogen to be able to spread and intensify effective dispersal must be achieved. This implies the production of spores in positions from which they can be transported and deposited alive on host plants under conditions and in quantities that will lead to infection. The environmental conditions under which a pathogen can intensify have been fairly accurately defined for a few diseases, but estimates of the distance of spread or loss through non-viability under different environmental conditions have been largely guesses. Besides affecting dispersal, environment also affects the pro duction of inoculum on the host, spore germination and subsequent infec tion of the host, all of which should therefore be considered in this phase of disease epidemiology. Factors of the environment are both li miting and optimal for spore survival and germination and must be known before infection can be predicted. The phase in the life cycle of a pathogen, when a spore state proceeds independently of its hosts, con cerns the field of aerobiology. Gregory (1961) considered aerobiology as "the study of passively airborne-.macro-organisms their identity, behaviour, movements and survival". This phase has been studied in a few diseases affecting agricultural crops (Carter 1963; Hirst and Stedman 1961, 1962; Jarvis 1962; Meredith 1961a, 1961b; Shanmuganathan and - 2 -Arulpragasam 1966; TSireeramulu 1962; Wilson and Baker 1946), but few-studies have involved aspects of the aerobiology of forest pathogens (Bega i960; Risbeth 1959; i^reeramulu 1963; Van Arsdel et al. 1961). The comandra blister rust, Cronartium comandrae Peck, an im portant native disease damaging several species of hard pine in North America, was selected for this study, because of its distinctive aecio spore, a unique pyriform shape, which is easily identified against the background air spora. Except for an outline of the C_. comandrae rust — its history, distribution, life cycle, symptoms and damage, this study is limited to the aerobiological aspects of the aecial state of the rust. The rust has been known on pine for nearly 90 years and was first recorded in western Canada from Alberta in 1907 on lodgepole pine, Pinus contorta Dougl., and from British Columbia in 1913 on ponderosa pine P. ponderosa Laws. (Arthur and Kern 1914). Widespread damage by C_. comandrae in western North America was not recorded until the mid-1950's (Mielke 1957), but during the last decade there have been several re ports (Andrews and Harrison 1959; Krebill 1965; Peterson 1962a), and it-has recently become a plantation problem in other parts of North America (Cordell et al. 1967; Dance and Lynn 1965; Wolfe et al. 1968). The genus Cronartium includes the world's most damaging tree rust fungi (Peterson 1962b); C. comandrae has been reported the most destructive stem rust on hard pines of western North Americai•(Peterson 1962a; Hawksworth 1964). Hawksworth (1964) suggested that C. comandrae was one of the three major diseases of P. contorta. The life-cycle of C. comandrae has been summarized by Hedgcock - 3 -and Long (1915a), Mielke (1961), and Mielke et al. (1968), but the envir onmental conditions under which each spore state develops or survives are not well known. C. comandrae is a heteroecious rust which grows perennially in the living bark of hard pines, and develops annually on the alternate hosts of the genera Comandra and Geocaulon. The pine is infected during the late summer and fall; the pycnia appear in the summer 2 or 3 years after initial infection, and aecia are produced the following year. The aecio spores are dispersed and infect the alternate hosts, where several genera tions of uredia may be produced. Telia follow the uredial state, and these germinate to produce basidiospores which are able to infect pine, com pleting the life cycle of the rust. Heavy pine infection occurs at infre quent intervals (Krebill 1965; Mielke 1957; Peterson 1962a; Wagener i960) giving rise to the use of the term "wave" years of infection. This sug gests that satisfactory environmental conditions for infection do not occur every year, or even every decade. Characteristically a native pathogen is in stable balance with its hosts and natural environment, but if this balance is upset then local or widespread epidemics will result. For an outbreak — an intensification or spread of the comandra blister rust to occur, environmental factors would have to be non-limiting for three se parate aerobiological phases during the life cycle of the rust in any one year. Although the present study concentrates on the first phase of aecio spore sporulation on pine to dispersal and germination on the alternate host, similar factors would also affect the repetitive uredial phase on the alternate host, and the phase between basidiospore production on the alternate host and infection on susceptible pine. The basidiospore phase, - k -is probably the most critical for determining outbreaks or wave years of pine infection. The basidiospore phase was not selected for study here, as its spore state was not very suitable for distance dispersal studies. The information gained in this study can be related to the probable aero biology of the other rust spore states, and to that of other important Cronartium rusts, and forest pathogens in general. The aerobiological aspects which play a role in the sequence of development of each spore state of the rust and which must be considered in any aerobiology phase, include both meteorological and biological en vironmental factors. The main aspects which are considered in the present study include, a) the character of the aeciospores, their production and release, b) the boundary layer and turbulent factors of the atmosphere as they, affect escape, transport, dispersion and deposition of aeciospores, c) the environmental factors affecting aeciospore germination, d) the environmental effects on aeciospore viability during these events, and e) biological factors affecting aeciospore production, dispersal, germina tion and viability. - 5 -THE ORGANISM HISTORY OF THE FUNGUS In the 1870*s C. H. Peck described two new rust fungi. Thirty-five years later they were connected as the aecial and telial states of the same fungus. In l875.Peck described as a new species, under the name Periderm!um pyriforme, a caulicolous or stem-inhabiting Peridermium with aecia having obovate, pyriform, or oblong-pyriform spores, from a speci-. men collected by J. B. Ellis on pine branches at Newfield, New Jersey. Peck was informed by Ellis that the collection may have been made in Georgia, and placed by accident among his New Jersey specimens. Arthur and Kern (l9lk), after examining the type specimen, believed the in scription on the original specimen voucher was correct. There are still no records of any collection of this rust in Georgia. Hedgcock and Long (1915b) suggested Pinus rigida Mill, was the host of the type speci men, as this was the only native species of pine in the type locality which was found infected elsewhere (Vermont). In 1879 Peck described the second fungus, the telial state on the stems of Comandra pallida A.DC. from Colorado, as a new species, Cronartium comandrae. He stated that this species appeared to be a paler form of the same species which occurred on leaves of Comandra umbellata (L.) Nutt. A specimen of C. comandrae was collected by Ellis on _C. umbellata (Ellis and Everhart, North American Fungi, No. 1082) near New-field, New Jersey, in 1879, which was used by Hedgcock and Long (1915a) as more evidence that the type location for P. pyriforme was correct. An - 6 -early collection by T. G-. Lea from Ohio On Thesium umbellatum (C. umbel lata) was given the name Cronartium asclepiadeum Kze. var. Thesii by Berkeley (l845), to distinguish the rust from the European C. asclepia deum. Berkeley stated then that this variety might be a distinct species. In 1895 Lagerheim, elevated it to species status and termed it Cronar tium thesii. In the meantime Peck (1879) had correctly given the telial state of the rust full species status as C. comandrae, its present valid name. Following Peck's description of Peridermium pyriforme, con siderable confusion developed in the use of this binomial, which produced many mis-identifications of specimens and erroneous reports of the occur rence of the comandra rust. The:specimens which probably initiated this confusion were collections by Ellis of Peridermium rusts from near the type location of P. pyriforme at Newfield. One was collected on Pinus inops Ait (P. virginiana Mills) in May 1882 (issued as P. pyriforme, Wo. 1021, Ellis, North American Fungi), and the others were collected on P. virginiana in April 1890, and on P. rigida in May 1890 (both in the Herbarium, New York Botanical Garden). The latter two were originally labelled Peridermium pini Wallr., a European Peridermium, but were later referred to P. pyriforme. However, these have now been referred to P. comptoniae (Arth.) Orton and Adams. These later collections of Ellis lead Arthur and Kern (1906) and Arthur (1907a) to describe as P. pyri forme what is now known as P. comptoniae, completely ignoring Peck's description stating that the aeciospores were pyriform. Clinton (1908) later connected this so called P. pyriforme (P. comptoniae) with - 7 -Cronartium comptoniae Arth., a telial state occurring on Comptonia  peregrina (L.) Coult. and Myrica spp. Prior to this, Underwood and Earle (1896) had cast doubt on the status of P. pyriforme and had sug gested that it was possibly a synonym of Peridermium cerebrum Peck, a species described in 1873. They suggested this, because they found the striking pyriform spores of P. pyriforme occurred "in both forms of P. cerebrum as they occur in the South". This indicates further con fusion of the identity of Peridermium forms on pine, as only the true P. pyriforme has pyriform spores. Hedgcock and Long (1913) described a caulicolous species of Peridermium with ovoid to pyriform aeciospores on Pinus contorta col lected by E. Bethel from Colorado, as a new species, Peridermium betheli. Unable to compare the specimen of Bethel with type material of P. pyri forme , they relied (Hedgcock and Long 1915a) on the published statement of Arthur and Kern (1906) which concluded that Peck's description was inaccurate concerning the pyriform shape of the aeciospores. Arthur and Kern (19135 1914), however, discarded their original interpretation of P. pyriforme after seeing a specimen submitted in 1913 by W. P. Fraser, collected on Pinus ponderosa at Vernon, British Columbia. This specimen had the typical pyriform spores described originally by Peck. Arthur and Kern (1914) speculated that the alternate states of this Peridermium would be found on species of Comandra, as this genus had the only unattached species of Cronartium then known. A I907 col lection of P. pyriforme on Pinus contorta from Devil's Lake (now Lake Minnewanka), Banff, Alberta, by E. W. D. Holway, gave support to this - 8 -speculation, as Holway wrote on the packet that the Peridermium on pine was undoubtedly associated with a Cronartium on Comandra. Orton and Adams (igik) found true P. pyriforme on Pinus pungens Lamb., and Cronartium comandrae on Comandra umbellata within ho feet of the in fected pine. They concluded that this Cronartium was the alternate state of P. pyriforme. They also suggested that P. betheli was a synonym of P. pyriforme. Hedgcock and Long (lyik) successfully infected Comandra  umbellata with aeciospores of P. pyriforme, confirming the life cycle of the rust and gave it the new combination Cronartium pyriforme. In the following year (1915a) they gave the morphology, life history, and distri bution of this rust as then known under the new combination C. pyriforme. Kirkwood (1915) inoculated Pinus ponderosa with spores from Comandra (pallida?) through incisions in the bark, which resulted in the develop ment of rust hyphae within the bark and wood of-.--the pine. This was the first record of the successful transfer of C. comandrae from Comandra to pine. He also successfully inoculated Comandra plants with aeciospores from pine. More recently, Andrews et al. (1963) reported successful in fection of Pinus contorta with basidiospores of C. comandrae. Telial horns were placed among and above the needles, with pycnia being observed a year later. Boyce (1916) first described the pycnia and pycniospores of C. comandrae from a collection on Pinus ponderosa made in July 1916 in California. In the same year, Weir and Hubert (1917) observed pycnia and pycniospores occurring on P. ponderosa and P. contorta, confirming the findings of Boyce. Meinecke (1929) tried unsuccessfully for 6 years to inoculate directly from pine to pine with aeciospores. This indicates _ 9 -that C. comandrae was only heteroecious and can not be autoecious as is Peridermium harknessii J.P. Moore (Meinecke 19l6, 1920, I929; Hedgcock and Hunt 1920). Many early herbarium specimens of C. comandrae are still la belled with synonyms or combinations of them. Some early collections were placed under the form-genus Aecidium (Farlow 1905) instead of Peri dermium, some of which were even non-aecial collections (eg. Aecidium  asclepiadeum var. thesii on Comandra pallida from Kamloops, British Columbia). Hedgcock and Long (1915a) gave Caeoma comandrae Peck as a synonym of Cronartium pyriforme. Caeoma comandrae was described by Peck (1884) from a specimen collected by M. E. Jones on Comandra pallida in Utah., However, a recent check of the New York Botanical Garden Herbarium, where Hedgcock and Long reported the specimen deposited, re vealed no specimen held as Caeoma comandrae, but one held as Uredo  comandrae by the same collector. This specimen proved to be Puccinia  comandrae Peck, a species also described by Peck in 1884 but from a col lection by T. S. Brandeges in Washington Territory. This leaves the question of the validity of Caeoma comandrae as a synonym in doubt, as it has not been listed by others. - 10 -DISTRIBUTION AND HOSTS Comandra blister rust has been reported on one introduced and ten native hard pine species in North America, ranging from New Brunswick to the Yukon and southward to Tennessee, northern Alabama, Mississippi, New Mexico and California. It has not been reported from Alaska (Cash 1953; Laurent, personal communication 1966) or Mexico although suscept ible pines do occur. It has not been introduced outside North America but is a potentially dangerous rust to native hard pines of Europe and Asia. The uredial and telial states of the rust have been recorded over a similar range on the alternate hosts, Comandra and Geocaulon of the family Santalaceae. The earliest lists of distribution and hosts were given by Hedgcock and Long (1915a) and by Arthur (1907b, 1925, 1927, 1934). More recent collections have extended the known distri bution of the rust and re-identification has eliminated some of the earlier reports. All the Cronartium comandrae material has been seen in the following herbaria (Herbarium codes are those used in the "index Herb-ariorum", Lanjouw and Stafleu (196U): CFB, DAOM, DAVFP, PUR, UBC, WIN, and critical..'material has been seen from the following herbaria: BPI, MISSA, NY. Lists or information have been obtained from the following herbaria not included above: ALTA, FFB, MFB, MONT, QFB, QMP, SASK, WINF, WSP, MacDonald College, Ste. Anne-de-Bellevue, and many of the regional laboratories of the United States Forest Service. - 11 -Distribution of the Rust on Pinus Pinus attenuata Lemm. This host was not recorded by Arthur (1934) or more recently, by Peterson (1962b), but there are specimens from Grant Pass, Oregon, collected by J. R. Weir, Sept. 6, 1916, in the PUR and BPI herbaria. A recent publication by Peterson (1967) does include this pine as a host. Pinus banksiana Lamb. Arthur (1934) reported C. comandrae occurred on this host in Michigan, Wisconsin, Alberta and Saskatchewan. Hedgcock and Long (1915b) reported several collections from Michigan and. Minnesota on P. divaricata (=_P. banksiana). It was also reported from New Brunswick (Davidson and Newall 1957), Quebec (Pomerleau 1942), Ontario (Conners 1934) and Mani toba (Riley et al. 1952). Pinus contorta Dougl. Recorded by Arthur (1934) from Colorado, Idaho, Wyoming and Alberta. Hedgcock and Long (1915b) reported the occurrence in Michigan, and Peterson (1962b) in California, Montana, Oregon, Utah and Washington. Also recorded in British Columbia (Ziller 1953) and the Yukon (Molnar i960). In I968 two specimens were collected from the Cypress Hills, Saskatchewan (CFB 8406, 8407). Pinus echinata Mill. Four recent reports of the rust on this host are from Missouri (Berry et al. 1961), Arkansas (U.S. Forest Service I962; Pooling et al. 1964), and Tennessee (Powers et al. I967). - 12 -Pinus jeffreyi Grev. & Balf. Peterson (1962b) reported this species as a host in California and Oregon, and I have seen specimens collected by Lewis, from near the Siskiyou National Forest, Oregon (BPI, PUR). Pinus ponderosa Laws. Arthur (193^) reported it from California, Colorado, Montana, Oregon, South Dakota, Utah, Washington and British Columbia. Peterson (1962b) added the states of Arizona, Idaho, Nevada and Wyoming. Collec tions on P. ponderosa var. arizonica (Engelm.) Shaw are included in this distribution. Plantations of ponderosa pine have recently been reported infected in Tennessee (Powers et. al. 1967). Pinus pungens Lamb. Orton and Adams (l^lh) reported the rust from Pennsylvania. Pinus rigida Mill. There is a non-verifiable specimen from this host, collected at Mt. Calvary, New Jersey in 1917 (PUR), but Hedgcock and Long (1915b) reported a collection from Vermont. Pinus sylvestris L. The rust has been reported on this introduced pine, from Wash ington (Shaw 1958) and Connecticut (Spaulding 1958). In Canada it has been reported from Manitoba (Thomas 1953), Saskatchewan (Ives et al. I968), and Alberta (Powell and Morf 1965). Specimens from Roscommon, Michigan are deposited in the BPI and PUR herbaria, neither is verifiable, - 13 -and is probably C. comptoniae. A specimen collected in 1887 in the Mis souri Botanical Garden, St. Louis, Missouri (Arthur and Kern I906) and now in the BPI herbarium, was labelled Peridermium pyriforme, but is ' probably Cronartium comptoniae as it does not have the pyriform spores. Pinus taeda L. Miles (1934) first reported the rust on this host from Missis sippi, and more recently it has been, reported from plantations in Tennessee and the northeastern corner of Alabama (Powers et al. 1967). Pinus virginiana Mill. There is a lQlk collection by Hedgcock from Peterburg, Penn sylvania, in the PUR herbarium attributed to this host, but the collection is not verifiable, and was not listed by Hedgcock and Long (1915a, 1915b) or Arthur (193*+). Peterson (1967) reported a Pinus sp. specimen as .likely from this pine. Cordell et al. (1967) noted that planted and natural P. virginiana were not infected in Tennessee, although at several locations they were intermixed with infected P. taeda. Other reports The introduced species P. nigra Arnold and P. pinaster Ait. have also been reported as hosts of C. comandrae (Boyce 19^3; Spaulding 1958; Mielke 196I; Verrall 1964), but Peterson (1962b) points out that these were based on a report by Clinton (1912) at the time when Peri dermium pyriforme was often linked with C. comptoniae (Arthur and Kern I906; Clinton 1908). There is also a I918 non-verifiable specimen la belled as C. comandrae from Roscommon, Michigan on P. nigra in the BPI - Ik -herbarium. Peterson. (1966b) indicated the erroneous report (Gremmen 1964). of P. resinosa as a host for C. comandrae in Vermont. The collect ion upon which this report was based was again C. comptoniae. The distribution of C. comandrae on Pinus spp. in western and eastern Canada is shown in Figs. 1 and 2. Distribution of the rust on Santalaceae Two genera of the Santalaceae serve as alternate hosts for C. comandrae. The distribution for the genus Comandra has recently been modified as Piehl (1965) rearranged the genus, and placed what were five species (C. umbellata (L.) Nutt., C. californica Eastw., C. richardsiana Fernald, C_. pallida A. DC., C. elegans (Rochel) Reichenb.) as four sub species of one species (C. umbellata). This is not too satisfactory, especially in Canada, as his distribution map indicates that there are large zones which have intermediates between two subspecies. However, his arrangement has been followed, as it does eliminate the previous problem of distinguishing between C_. umbellata and C. richardsiana, "since they are now one subspecies (ssp. umbellata). Earlier, Fernald (1928) had separated the species C. livida Richardson from the genus Comandra, and formed a new genus, Geocaulon with G. lividum (Richardson) Fernald as the type. Piehl (1965) agreed with this separation, as Geo caulon has several features that separate it from Comandra. Both Comandra and Geocaulon are hemiparasites (Figs. 3, k, 5 and 6). Several papers have listed the plant species upon which Comandra is parasitic (Brooks '.v 1937; Fisher 1922; Harrington 19^5; Hedgcock 1915; Moss 1926; Piehl. Fig. 1. Distribution of Cronartium comandrae on Pinus species in western Canada. Fig. 2. Distribution of Cronartium comandrae on Pinus species in eastern Canada. Fig. 3- Aerial stems of Geocaulon lividum, with axillary greenish flowers. Fig. k. Aerial stems of Comandra umbellata ssp. pallida, with ter minal white flowers about to break open. Fig. 5. Series of aerial stems of Comandra umbellata ssp. pallida, which branch just below the soil and are attached to an . underground-rhizome. Some aerial stems dead from previous year. Fig. 6. Large haustorium of Comandra umbellata ssp. pallida atta ched to Populus tremuloides Michx. root, with smaller haustorium on smaller root. 17 -- 18 -1965)5 but I am aware of only one published report for Geocaulon (Moss 1926, as C. livida). Geocaulon is a genus with a distinctly northern distribution in North America, extending from Newfoundland to Alaska, and only as far south as northern New England, Wisconsin, Michigan, Minnesota, Idaho and Washington (Fernald 1928; Harris 1965; Hitchcock et al. 1964; Raup 1947 pi. XXII). Comandra umbellata (L.) Nutt. ssp. umbellata Piehl (=C. umbellata; CJ. richardsiana). Arthur (1934) reported C. comandrae occurred on this host in California, Delaware, Illinois, Indiana, Iowa, Massachusetts, Michigan, Mississippi, New Hampshire, New Jersey, New York, Ohio, Oregon, Pennsyl vania, Washington, Wisconsin, Ontario and Quebec. Piehl's (1965') dis tribution map suggests that the records for California, Oregon, and Washington were on the ssp. californica and not umbellata. Hedgcock and Long (1915a) reported a. collection from Vermont, and Powers et al. (1967) record it from Alabama and Tennessee. Specimens from the Dis trict of Columbia and Minnesota are deposited in the BPI herbarium. Col lections listed as C. umbellata or C. richardsiana from Manitoba and Saskatchewan are in the UBC and WINF herbaria, and are probably true C. umbellata spp. umbellata, or an intermediate between this and ssp. pal lida. Comandra umbellata (L.) Nutt. ssp. pallida (A.DC.) Piehl ( =C. pallida) Arthur (1934) reported it from Arizona, Colorado, Idaho, Iowa, Minnesota, Missouri, Montana, New Mexico, North Dakota, Oregon, South - 19 -Dakota, Utah, Washington, Wyoming, Alberta, British Columbia, Northwest Territories and Saskatchewan. Hedgcock and Long (1915a) reported col lections from Nebraska, and a collection from Michigan is deposited in the PUR herbarium. There is a specimen of this host from Texas in the BPI herbarium, and one reported as Comandra sp. (NY herbarium) which should probably be referred to ssp. pallida. Bisby (1938) reported its occurrence in Manitoba, and Conners (1934) in Ontario. There is some doubt about the specimen upon which Arthur (1934) based his Northwest Territories record. J. A. Parmelee (personal communication 1966) be lieved this record was based on the DAOM specimen, No. l86l, from near Martin Cabin, Slave Lake Alberta, collected in 1929, and that Arthur interpreted 'Slave Lake' as Great Slave Lake, Northwest Territories, instead of Lesser Slave Lake, Alberta. However, Arthur (1927) included the Northwest Territories in his distribution list prior to the collection of the above DAOM specimen. There is no record of Comandra from the Northwest Territories or the Yukon in the DAO or CAN herbaria, or more regional herbaria .(ALTA, CAFB, UAC, UBC), and it has not been collected north of about 59°N. latitude in Alberta. Porsild and Cody (1968), in their checklist of vascular plants in the continental Northwest Terri tories, indicated that C_. pallida was expected to occur, but at present there is no record from the. area. Comandra umbellata (L.) Nutt, ssp. californica (Eastw.) Piehl. I am aware of no herbarium specimens under this subspecies name, but according to Piehl's (1965) distribution map, the specimens recorded by Hedgcock and Long (1915a) and Arthur (1934) under C. - 20 -umbellata from California, Oregon, and Washington, should be ssp. califor-nica. Geocaulon lividum (Richardson) Fernald Arthur (1934) gave the following distribution of the rust on this host (as Comandra livida), Wisconsin, Manitoba and Quebec. Ziller and Molnar (1953) first reported the occurrence in British Columbia, and Molnar (i960) in the Yukon. Baranyay et al. (1962) first reported the infected host in Alberta, and a year later Baranyay and Bourchier (1963) reported it in the Northwest Territories, though inaccurately reported to be on Comandra. A collection made in 1937 by D. V. Baxter (DAOM 5558 and PUR 48514) from Great Slave Lake, Northwest Territories, and labelled Comandra sp. proved to be G. lividum. Other unreported herbarium specimens, some labelled as C. livida, extend the known dis tribution to Idaho (NY), Washington (DAOM) and Ontario (DAOM). Other reports Arthur (1934) referred to Buckleya distiochophylla (Nutt.) Torr. as a host of C_. comandrae in Tennessee. This species is no longer regarded as a host of C. comandrae, as Hepting (1957) made the connection of the Cronartium occurring on Buckleya with Peridermium  appalachianum Hepting and Cummins, giving the combination Cronartium  appalachianum Hepting, a rust which occurs in North Carolina, Tennessee and Virginia. The distribution of C_. comandrae on Comandra and Geocaulon in western and eastern Canada is shown in Figs. 7 and 8. Very recently a - 21 -distribution map for C. comandrae on Pinus spp. and Comandra spp.has been published by Krebill (1968b). Generally his northern distribution line for Canada does not extend far enough north, especially in the Yukon and Northwest Territories, and in Manitoba and Quebec. The col lection from Great Whale River, has been misplaced in Ontario instead of Quebec. Fig. 7. Distribution of Cronartium comandrae on the Santalaceae, Comandra umbellata and Geocaulon lividum, in western Canada. Fig. 8. Distribution of Cronartium comandrae on the Santalaceae, Comandra umbellata and Geocaulon lividum, in eastern Canada. - 2k -LIFE CYCLE Pine infection by Cronartium comandrae usually takes place during summer and autumn through the needles on branches or main stems (Andrews et al. I963). It is possible that the fungus infects young trees directly through the bark or through wounds (Kirkwood 1915)5 but this mode of entry is of minor importance. Penetration of the needle by the germ tube of the basidiospore or sporidium may be through the stomatal pore or directly through the epidermal cells. Following;pene tration a mass of hyphae develops within the needle tissue near the point of infection, then hyphal growth penetrates downwards through the vascular bundle of the needle and into the bark which becomes discolored The length of the period between needle infection and bark discoloration the incubation period, varies according to the length of growing season, the seasonal weather, and the distance the hyphae must grow between point of infection and the bark. The mycelium of the fungus, which branches irregularly, ramifies intercellularly through all live pine bark tissues, and extends into the outer sapwood tissues, mainly in the medullary rays (Adams 1919; Krebill 1968a). Krebill found the advancing hyphae in the phloem, never more than 2.cm.beyond the visible swelling of the infected stem. The infected bark increased in thickness due to the intercellular hyphae and expanding parenchyma cells. With the in crease in thickness the conducting sieve cells in the phloem collapse. One or more unicellular haustoria may penetrate into a single parenchyma cell from the intercellular hyphae. Krebill also reported haustoria sometimes present in sieve elements and xylem ray tracheids, and noted - 25 -that the haustoria vary in size and morphology depending on their posi tion. Peterson (1966a) reported the interesting occurrence of C. comandrae hyphae in tissues of the dwarf mistletoe, Arceuthobium ameri-canum Nutt. ex Engelm., on Pinus contorta, with haustoria penetrating into Arceuthobium parenchyma cells. The reaction of the sapwood to in fection is a decrease or cessation of growth in the canker area (Figs. 9 and 10). Resinosus also occurs and the resin may thoroughly impregnate the bark and outer sapwood (Figs. 9? 10 an(i ll) • The pycnial or spermogonium state appears as a distinct crust like layer of parallel pycniophores or spermatiophores pushing up between the outer layer of cortical parenchyma cells and the overlying thin walled cells of the phellogen (Adams 1919). The pycniophores arise from a zone of intertwined uninucleate hyphae which make up the base of each pycnium. The pycniophores contain a single nucleus which divides, one nucleus migrating into the immature pycniospore or spermatium which are abstracted from the free ends of the pycniophores at maturity. The pycn iospores appear to escape through the irregular cracking of the over lying pheHum tissue, and are exuded with a thin gelatinous fluid to form orange-colored droplets (Fig. 12), which eventually dry on the sur face of the bark to form dark-colored spots called "pycnial scars". Insects are attracted to the droplets and help distribute the pycniospores. The pycniospores are 3 to k by 3 to 7M-, pyriform in shape (Fig. 13), and probably have a sexual function similar to that shown by Pierson (1933) for pycniospores of Cronartium ribicola Fischer. The pycnia appear one to three years after the initial infection and precede the aecial state Fig. 9- Transverse sections taken from the central portion of ' four Cronartium comandrae cankers on lodgepole pine showing eccentric growth caused by the rust. Note the resin impregnation of the outer sapwood. Fig. 10. Transverse sections taken from the central portion of five Cronartium comandrae cankers on lodgepole pine • showing eccentric growth caused by the rust. Fig. 11. Transverse section taken through the aecial zone of a Cronartium comandrae canker showing the aecia situated in the bark tissues. Note the resin impregnation of the bark between aecia, and the swelling of the outer ring of sapwood caused by the fungus. - 26 -Fig. 12. Pycnial drops on the pycnial zone of a Cronartium coman drae canker on the stem and branches of a lodgepole pine. Note the swelling of the stem associated with the canker. Fig. 13. Group of pyriform pycniospores of Cronartium comandrae. x 2500. Fig. Ik. Pyriform aeciospores of Cronartium comandrae showing size and shape variation of the-spore tail. X 700. Fig. 15. Pyriform aeciospores of Cronartium comandrae showing the spore wall ornamentation. X 700. - 28 -on the identical area by at least one season. The pycnia occur in tis sues just beneath the periderm (Adams 19195 Krebill 1968a) and overlie the aecia which develop several cells deeper in the cortex (Fig. ll). Hiratsuka and Cummins (1963) found that the Cronartium species were the only rusts to develop an intracortical pycnia, and one of the few rust genera in which pycnial growth was indeterminate. Pycniospore product ion usually begins in June, or shortly after the main aecial production period, and in some seasons they may continue to be produced into October. The primordia of the aecia appear as a mass of uninucleate hyphal cells, originating in the cortical parenchyma of the host some 3 to 8 cells below the phellogen (Adams 1919)- The cells of the central portion of the matured primorddfa become differentiated first to form bi-nucleate basal cells. Each of the first basal cells formed cuts off an intercalary cell and a cell which becomes metamorphosed to form a peri-dial cell of the central arch. The peridium so formed consists of a continuous membrane one to three cells thick, and the aecium is delimited by the peripheral peridial chains. The binucleate basal cells cut off chains of cells which further divide into an intercalary cell and an aeciospore. The cell walls of the aeciospores begin to thicken when they are about the third or fourth spore from the basal cell. The aecio spores then become acuminate above (Adams 1919? Arthur 193^) or below (Arthur and Kern l^lk; Hedgcock and Long 1915a) to give a typical pyri form shape, are quite variable in size, 17 to 37 by 38 to 77u, with colorless 2 to 3u thick walls, which are finely verrucose with small ir regular tubercles (Figs. Ik and 15). As the aecia continue to mature the Fig, 16. Canker of Cronartium comandrae on a young "branch of Pinus  contorta, with the peridia of the aecia still unruptured and forming typical blisters pushing through the bark. Fig. 17. Canker of Cronartium comandrae with abundant ruptured aecia covering the aecial zone. Fig. 18. Bright orange-colored aeciospores of Cronartium comandrae covering the exposed aecia of the canker. - 29 -- 30 -overlying host tissue becomes raised and gradually ruptures exposing the white blisters consisting of the peridium (Fig. 16) enclosing a powdery mass of pyriform orange-yellow aeciospores (Fig. 18). The aecia (Figs. 17 and 18) are produced over the area which previously produced pycnia, and occur over or between pycnial scars, usually in the year following pycnial production, although aecia may not occur until the second year or even later. Aeciospore production varies widely with seasonal con ditions, locality, and bark thickness. In the study area the main period of production began in late May and continued into July and up to the end of August at a reduced rate. In some years the cankers may produce a second crop of aeciospores late in the season, but in the current year pycnial zone. There may be a great variation between individual aecia from one canker which can prolong the production period, but usually production from an individual aecium only lasts for two to five weeks. Frequently the individual aecia are so close together that several ap pear to coalesce to form one large irregular aecium. After the aecio spores are released, the peridium disintegrates, and bark necrosis oc curs around the aecial ruptures. Adams (1919) reported that a new cork layer is formed below the aecia, and the dead tissue is sloughed off or remains as a rough, deeply-cracked bark (Figs. 19, 20 and 2l). Krebill (1968a) found no evidence of the development of a secondary cork layer beneath aecia, but a layer occurred at times beneath pycnia. The mature aeciospores, which are generally dispersed by the wind to susceptible Comandra or Geocaulon plants, are binucleate (Figs. 22 and 23). Upon germinating one to six germ tubes may emerge through Fig. 19. Young canker of Cronartium comandrae on Pinus contorta with rough bark covering the old aecial zone around the branch stub that acted as entry point to the stem. A-bundant aecia were produced around the old aecial zone. Fig. 20. Slight hypertrophy of the young Pinus contorta stem caused by Cronartium comandrae, with characteristic rough, cracked bark, in the older canker area caused by aecia rupturing. Fig. 21. Typical basal stem canker of Cronartium comandrae showing older rough bark zone and current year aecial zone with many dispersed aeciospores caught in the crevices of the bark. - 32 -the spore wall (Fig. 2h), but usually the two nuclei migrate into one of the developing germ tubes (Fig. 25) and only this germ tube continues to elongate to any extent. The germ tubes have multiple short branches which may occur irregularly throughout the length of the germ tube (Figs. 26 and 27). The germ tubes are non-septate, with the two nuclei continu ing to migrate toward the tip of the tube during elongation. If condi tions are suitable an appressorium forms (Fig. 28) from which an in fection peg may develop (Figs. 29 and 30) and penetrate the host through the stomata to infect and develop a binucleate mycelium. This type of aeciospore germination is typical of the heteroecious rusts, and is in marked contrast to that of the autoecious rusts (Hiratsuka 1968; Hirat-suka et al. 1966; Powell and Morf 1966). The stem, leaf petioles, leaves, floral bracts and fruits of the alternate host may also be infected. One to three weeks after aeciospore infection, the uredia appear on the sur face of the leaves as small yellow pustules (Fig. 31)• The uredia de velop from a hyphal plexus often beneath a stoma of the host leaf (Moss I928). The vertical hyphae arise in a palisade fashion and become di vided into three cells; the peridial, intercalary and sporogenous cells. At a later stage the intercalary cells disorganize, the basal sporogenous cells give rise to spore-initials, and the peridial cells give rise to the peridium of the uredial sorus. The peridium of the uredial sorus quickly ruptures exposing a, mass of elliptical to globoid, orange-yellow urediospores, 16 to 23 by 19~28u, with nearly colorless 1.5 to 2\x thick walls, which are sparsely echinulate. The urediospores are easily disseminated by wind or insects and are able to germinate and re-infect Fig. 22. Aeciospores of Cronartium comandrae, stained with HC1-Giemsa, each with two nuclei. X 700. Fig. 23. Aeciospore of Cronartium comandrae, stained with HC1-C-iemsa, with two nuclei. X 1200. Fig. 2k. Initiation of aeciospore germination, with four germ tubes developing through the spore wall, after 1 hour. X 1200. Fig. 25. Cronartium comandrae aeciospore with several germ tubes, but only one germ tube well developed after 6 hours. Note the short multiple branches on the non-septate germ tube. X 380. Fig. 26. Group of germinated aeciospores with typical branching of germ tubes after 2k hours. X 125. - 33 -Fig. 27. Aeciospore of Cronartium comandrae with one developed germ tube, after 3 hours. Note the non-septate condition of the germ tube and that the two nuclei have migrated towards the tip of the germ tube. X 830. Fig. 28. Germ tubes of two Cronartium comandrae aeciospores after 3 hours, showing development of a globose and club-shaped appressorium on the developed germ tube, into which the two nuclei have migrated. X 700. Fig. 29. Germ tube development of a Cronartium comandrae aecio spore after 3 hours, showing a globose appressorium with two nuclei and an infection peg. X 1080. Fig. 30. Germ tube development of two Cronartium comandrae aecio spores after 3 hours, showing two types of appressoria each with two nuclei and an infection peg. X 6l0. Fig. 31. Uredia of Cronartium comandrae on central portion of a Comandra umbellata ssp. pallida, leaf. Dark structures on bottom are young telia. Fig. 32. Telia of Cronartium comandrae on both surfaces of a por tion of Comandra umbellata ssp. pallida leaf. - 35 -- 36 -the same alternate host, thus spreading and intensifying the infection during the summer, often over a distance of a hundred miles or more. In the present study uredial infections were found hundreds of miles out on the Prairies away from any pine source. All parts of Comandra plants were highly susceptible to infection at any stage in their develop ment. Four to seven generations of the uredial state may be produced in a single season, if weather conditions permit, which greatly increases the possibilities of pine infection. Telia may form about two to three weeks after the uredia ap pear on the plants and may occur on lesions initiated by uredia or on new lesions late in the season. The telia are reddish-brown, slender, often curved, hair-like structures, 80 to 120u thick and about 1 mm in length, which stand up from the host tissue (Fig. 32). They are com posed of columnar rows of teliospores, 12 to 16 by 28 to kk\i, with nearly colorless, 2 to 3M- thick, smooth walls, held together in a gela tinous matrix. In time, usually under moist conditions, each teliospore germinates in situ by means of a five-celled promycelium, each of the four upper cells bearing one small thin-walled globose basidiospore, 2 to 6 by 2 to 6\i, on the tip of a stout sterigma. The basidiospores are disseminated by the wind, and if deposited on susceptible pine needles while still viable, can germinate under favourable conditions and in fect the pine, thus completing the life cycle of the fungus. - 37 -SYMPTOMS The first visible sign of comandra blister rust infection on a pine is a yellow-brown spot on a needle. This is at the point of in fection, but it is difficult to distinguish macroscopically from similar spots caused by other agencies. In the first, second or third season following infection an area of the bark at the base of the needle fasc icle becomes discolored, and a small spindle-shaped swelling may develop. This is the beginning of a canker, which mostly begins on needle-bearing branches and stems. Infections on main stems usually enter via small branches (Fig. 33). Mycelial growth in the host tissue is more rapid longitudinally than laterally, and more rapid towards the base than the tip of the branch or stem. Reddish-orange pycnial drops appear on the bark two or more years after infection . In subsequent years the pycnia form in a zone behind the periphery of the infected tissue and outside the aecial zone. Pycnial drops are not visible long; they may be eaten by larval and adult insects, dry up, or be washed by rain. "Pycnial scars" may still be seen if the bark is relatively smooth. Aecia appear as white blisters pushing through the bark over pycnial zones of the previous years. The aecial peridia rupture releas ing masses of orange aeciospores. Aecia are only present in one area for one season as the pycnial and aecial zones continually form around the periphery of the canker. Not all aecia appear as blisters through the bark. In thick-barked trees the aecia are often buried and the aeciospores Fig. 33- Branch infection of Cronartium comandrae on Pinus contorta, with mycelium causing swelling in the main stem around the branch stub. Fig,. 35-. Heavy resinosis forming a dried crust over a portion of a Cronartium comandrae stem canker. Note rough bark of old branch canker (upper left) which served as entry for the rust to the stem. Fig. 3^. Tree with dead spike-top 'above a canker of Cronartium comandrae girdling the tree; lower branches are progres sively killed by downward growth of rust. - 39 -appear through cracks in the bark. On removing the bark large areas of aecia with aeciospores are exposed. Under natural conditions only a small proportion of these aeciospores would be dispersed. Some trees, especially those with marked resinosis, seldom produce aeciospores, and it is difficult to establish the limits of the canker microscopically. Where aecia have been produced, the bark cracks and the cambium and underlying wood die. As the fungus grows the branch or stem even tually is girdled and parts beyond the canker die. Spike-topped trees (Fig. 3*0 and flagged branches are conspicuous symptoms of the disease. When cankers develop on the main stem, the canker zone is occasionally constricted instead of swollen, as increased diameter growth of the healthy stem continues above and below the canker during canker gird ling. Resin exudation is a conspicuous symptom of disease, especially on Pinus contorta (Fig. 35). Much of the resinosis is caused by the fungus but some is caused by rodents chewing the succulent infected bark of the canker. Rodent damage is usually a good indication of the pre sence of the disease. (Figs. 36 and 37). The first symptom of the rust on the alternate host is the yel low swellings of the developing uredial sori, which rupture to release the urediospores. A few weeks later the telial "horns" are easily visi ble developing on the same area as the uredial sori or from new lesions. Uredia and telia may be present over the whole plant. There is some chlorosis of the leaves in heavily infected plants, slight evidence of reduced leaf growth and defoliation is usually premature. Fig. 36. Annual rodent damage on a large Cronartium comandrae stem canker on Pinus contorta. Note the strip of dried dead bark not removed on each annual visit, and the abundant exudation of resin. Fig. 37. Rodent damage on a Cronartium comandrae stem canker a-round a branch which acted as an entry point for the rust into the stem. Sporulation of the aecial zone can be seen outside the chewed area. - ko -- 1+1 -DAMAGE Comandra blister rust damage in natural pine stands is not usually spectacular, infection being limited to some extent by the dis tribution of its alternate hosts, which, although ranging over much of North America, are restricted locally to small areas. Comandra umbel lata is rarely found within pine stands as it favours dry open areas, but Geocaulon lividum occurs on wetter areas within spruce-pine stands. Heavy pine infection is usually found adjacent to areas of infected al ternate hosts. The rust attacks trees of all ages and sizes and causes mortality by basal or stem girdling. The period between initial infec tion and death of the tree may be as much.as thirty years, or more, and then only a few trees die each year. Seedlings may be killed within a few years following infection, because the rust can enter through needles on the main stem and the fungus soon girdles the small stem. The fungus characteristically enters the main stem of the older tree via branch in fections. The fungus only grows one to. three inches down the branch each year so that the rust.may be present in the branch many years be fore it reaches the main stem. If the branch is killed before the rust reaches the main stem, the main stem will not usually become infected. Dead tops or spike-tops were common damage features in mature stands where the fungus had girdled the stem thereby killing the upper stem and branches. Childs (1968) showed that downward growth on the stem by the rust is less than 6 inches per year, thus the lower portion of the tree may live for many years if vigorous lower branches remain. Infection within a stand may have occurred in only a few years, or even in one year, when conditions - h2 -for infection were favourable and a supply of basidiospores was avail able. Several reports show that pine infection by:the rust is in frequent. Observations were made for over 50 years in an infected Pinus  ponderosa stand in Siskiyou County, California, and it appeared that current damage was from infection kO years ago with little or no recent infection (California Forest Pest Control Action Council 1958; U. S. Forest Service 1955; Wagener i960). Infection of P. ponderosa in south ern Nevada was reported as sporadic by Wagener (i960) with a wave of infection by the rust in .1932 or 1933" and almost no new in fection on pine since Wagener (1950) reported the rust on P. ponderosa in a southeastern Idaho plantation " . . . . chiefly on the 19^0 and 19^1 growth of branches and no evidence was found in I962 to indicate any more recent infection (Krebill, personal communication 1962). Little or no recent infection was found on P. ponderosa in the Black Hills, South Dakota, and areas of Colorado in stands where damage was reported 25-35 years previously (Peterson 1962a). Peterson (1962a) pre sented evidence that infection occurred two to several decades ago on P. contorta in the Bighorn and Shoshone National Forests, Wyoming, with almost no recent infection. One of the first reports of damage in the United States was made by Boyce (Hedgcock and Long 1915a) in Klamath National Forest, Cali fornia, where he found 34% of the young P. ponderosa dead and a further 17% infected in a stand representing average rather than heavy infection. Meinecke (1928) reported 35% of P. ponderosa dead or doomed in an area - k3 -near Mt. Lassen, California, and calculated that this resulted in a 21$ reduction of forest cover. Other reports of damage to P. ponderosa have been noted from other parts of California (U.S. Forest Service 195'5, 1956), a 30-year old plantation in Caribou National Forest, Idaho (Wegener 1950), Charleston Mountains, southern Nevada (Wegener i960), the Black Hills of South Dakota (Hedgcock and Long 1915a; Luckinbill 1935)> and the east side of the Cascade Range in Oregon and Washington (Childs 1968). Damage to young P. pungens in Pennsylvania was noted by Adams and Orton (191^) and Hedgcock and Long (1915a); the latter reported that 58$ of the pine which were producing aecia at the end of June were dead by autumn from the girdling effect of the fungus. Until recently, P. contorta was considered only an occasional host for this pathogen (Mielke 1957) although Peterson (1962a) quotes an unpublished I925 report by Bethel of damage to this host in Colorado. Mielke (1957) reported mature stands covering several square miles where 50 - 98$ of the trees were infected in areas of Idaho, Utah and Wyoming. Andrews and Harrison (1959) also reported severely infected P. contorta stands in Wyoming adjacent to sagebrush areas supporting the alternate host. Peterson (1962a) studied damage on 2k plots in infected stands in northern Wyoming and found 21$ of the trees infected. The stands were largely 50 - 100 years old with infection centered in tissue usually over 20 years old, with the cankers at an average height of 12 feet. Krebill (1965) studied infected stands in 12 National Forests located from central Montana to northern Utah and found that some cankers were 100 years old, and that there had been a build up between 1910 to 19^5, - 44 -with only sparse recent infection. In most cases over 90% of- the sampled cankers were in trunks and from one-fourth to one-half of these had caused spiketops. The centers of cankers ranged from 1.5 to 69 feet above ground with the mean height around 26 feet. About 90% of the cankers had been scarred by rodent chewing. In one sample plot 26% of all live lodgepole pines were infected, and there was considerable evi dence of recent mortality by rust girdling. Recently C. comandrae has become a problem in young plantations of P. ponderosa and P. taeda in the Cumberland Plateau area of eastern Tennessee and northeastern Alabama (Powers et al. 1967). Powers et al. report plantations with over 9°% of the less than 10 year-old trees in fected, and one case, with 57% of the 2 year-old P. taeda trees infected. More recently, Cordell et al. (1967) reported that 40% of the P. taeda plantations less than 10 years old were infected on the Cumberland Plateau, and on these plantations 6% of the trees were infected. They also detected severe infection and mortality in a P. echinata plantation in the same area. C. comandrae is also proving a plantation problem on P. echinata in the Clark National Forest, Missouri (Berry et al. 1961), and Ozark National Forest, Arkansas (Dooling et al. 1964). In the latter area 44% of the trees on three plots planted in 1958 in the Buffalo Dis trict were dead by 1967 (Dooling 1967). Wolfe et al. (1968) recently found that the rust was now present on all 8 districts of the Ozark National Forest, being observed in 19% of the P. echinata stands examined. In Canada there have been few recorded reports of extensive damage, although the rust has been reported as fairly common throughout - 45 -the Prairie Provinces. In 1959 two area reports of mortality were re corded in P. contorta stands: one in young trees near Robb, Alberta (Thomas et al. i960), and the other in the Teslin-Whitehorse area of the Yukon with 5 - 10$ of the sapling size trees infected (Molnar i960). Later, Baranyay and Stevenson (1964) reported 3$ of the living trees on seven 0.05-acre plots in 20 year-old regeneration near Robb were infected by C. comandrae, the range between plots varying from 1.4 to 10.8$. In eastern Canada, 22$ of the trees in a P. banksiana plantation at Thunder Bay, Ontario, were infected (Dance and Lynn 1965), and more than 10$ of young P. banksiana in a small area at Saint Urbain, Quebec, were infected (Martineau and Ouellette 1966). Meinecke (1928), Horton (1955), and Mielke et al. (1968) men tion the value of the rust as a biotic thinning agent causing mortality in overstocked young pine stands. Peterson (1962a), however, states that mortality caused by the rust "could seldom or never be written off as harmless or beneficial natural thinning". He found that the infected trees were larger, and that suppressed trees did not respond to release by the rust. He also found that over half the infected live lodgepole pine trees had a lower vigor rating than they would have if rust-free. Childs (1968) found that top-killing of P. ponderosa by the rust was not indicative of high-risk trees, unless the remaining crown was small and of poor vigor. Nordin (195*0 reported the rust was more prevalent in pure than in mixed stands in Alberta. Damage appraisal surveys carried out in association with the present studies, indicated that 44$ oftthe 725 trees on a 0.1 acre plot - k6 -in a 20 year-old P. banksiana stand south of Rae, N.W.T., were infected in I965 with an average of 1.6 infections per tree;.. By 1966, 7.9% of the infected trees had died and many others had dead tops (J. A. Baranyay, unpublished data I965, I966). Nearly a third of the cankers noted in I965 were inactive in 1966, but many new active cankers were recorded. A 1966 survey on a 0.1 acre plot in a 25 year-old P. contorta stand near Saskatchewan River Crossing, Banff National Park, Alberta, indicated 32.7% of the trees infected by the rust, Peridermium stalactiforme Arth.. & Kern, and 10% by C. comandrae (J. A. Baranyay, unpublished data 1966). Over half the C_. comandrae infected trees were dead by 1968. A I92U plantation of Pinus sylvestris near Beaver Mines, Alberta, was reported to have extensive rodent damage, some with fresh chewing which lead me to suspect a rust infection. After considerable search, one strip of active canker with spores was found at the edge of a chewed area (Powell and Morf 1965). Of 50 trees tallied all had rodent damage, suggesting rust infection was once widespread throughout the plantation. Heavily in fected C. umbellata ssp. pallida grew within a few feet in an open meadow. About 500 cankers were measured and observed on trees varying in age from 3 to 125 years from 23 areas in the Rocky Mountains and Foot hills regions of Alberta. The average height of the center of the can kers was l.k feet, with the highest centered at 15 feet. The average length of the cankers was 1.2 feet, with the longest 15 feet. In 1966 and 1967, kkO live cankers were tagged (103 in 1967), but by 1968 101 of of these were dead, of which only 13 were branch cankers. Canker mortality at one location amounted to 82%, and at another 55%5 both occurred in 10 - kl -to 20 year-old stands. In a regeneration area where few trees were over 5 years old, kk% of the trees with cankers died in one year. Total tree infection in this area amounted to 23% on a 0.22 acre plot in 19°75 but it was unlikely that any of the infected trees would survive for more than a few years at the current mortality rate. The regeneration oc curred in an area of Comandra and chances for further infection in these young age trees were high. - hd -AREA OF STUDY LOCATION, PHYSIOGRAPHY AND GEOLOGY The main area of study was restricted to a portion of the Kananaskis River valley centered on the Kananaskis Forest Experiment Station some 5 miles south of Seebe, Alberta (Fig. 38), where the foot hills merge with the front range of the Rocky Mountains. The Kananaskis Forest Experiment Station, an area of about 2k square miles, is bounded to the north by the Stoney Indian Reservation, and on all other sides borders on the Bow River Forest, of the Alberta Rocky Mountains Forest Reserve. The Kananaskis River, which flows north to join the Bow River at Seebe, lies in a north-south valley situated between two of the east ern series of ranges of the Rocky Mountains. In the northern part of the valley the Kananaskis Range lies to the west and the Fisher Range to the east. The altitude of the area varies between U,200 and 95 500 feet, with timber line between 6,500 and 7,000 feet. Barrier Lake is a man-made water body over four miles long, filling a portion of the flat-bottomed U-shaped valley between the northern end of the Kananaskis Forest Experiment Station and a point about a mile south of the Evans -Thomas Creek. The valley bottom, which at some points is !§• miles wide, is flanked at many places by terraces of glacial till and morainic ma terial, and lacustrine and alluvial deposits; evidence of previous glaciations and retreats to which the area was subjected. From approxi mately the 5,000 foot level the slopes rise gradually or precipitously to masses of upthrust and folded sedimentary rocks deposited from Fig. 38- Map of the lower Kananaskis River Valley, Alberta, show ing the locations of the main study areas. - k9 -- 50 -Cambrian to Cretaceous times, which form the mountains and ridge tops of the ranges. The geological formations, of the area were studied by Bawling (1905) and Beach (19U3). SOILS A preliminary survey of the soils of the area was made by Crossley (1951), who found well drained alluvial soils and eroded soil and rock areas dominant in the flood plains of the Kananaskis River and its major east bank tributaries. Shallow stomy/ lithosol soils are pro minent on moderate to steep slopes where erosion is quite rapid. The dominant soils on the lower slopes of the valley are grey podzols of calcareous or non-calcareous origin, with brown podzols and podzols gen erally at higher elevations. Localized areas support hangmoor peat, half-bog and sod soils, which result from excessive moisture, poor drainage, or soils at high altitudes under grass or open forest. Brown forest soils, and localized chernozem and rendzina soils are found around the northern end of Barrier Lake and to the north, where the Foothills abut against the Rocky Mountains. A forest land classification map, based on surficial material, was recently prepared for some of the Kananaskis Research Forest (Duffy and England I967), and a generalized soils map was prepared for the Marmot Creek basin (Jeffrey 1965). CLIMATE The main climatic characteristic of the Kananaskis Valley is - 51 -its variability, typical of most mountain regions in continental locations. This variability is only now becoming more accurately determined through the increase in the number of climatological stations in the valley, and especially through the intensive instrumentation associated with the Alberta Watershed Research Program in the basin of Marmot Creek (Munn and Storr 1967), and other specialized projects (MacHattie 1966, 1967). Kananaskis (elevation k,560 feet MSL, latitude 51°02'N, longitude 115°03' W) is the only station with a relatively long climatological record upon which acceptable average conditions can be based, but even some of the parameters recorded at this station are based on short term, and some times irregular periods. This station is located 300 feet to the east of study location 1 (Figs. 38 and 39)j on a knoll on the edge of a large grassed clearing, open to the south and east, with k-0 foot trees im mediately to the north, thus the site is not particularly typical of this forested region. All other climatological stations in the lower Kananaskis valley have been established during the last decade, and again records for some parameters are•intermittent, and in the case of Pigeon Lookout (elev. 6,000 ft MSL, lat. 51°03'N, long. 115°0Vw) depend on the duration of the summer forest fire season. Tables I, II and III, give the climatic summaries for the meteorological parameters recorded at Kananaskis, Kananaskis Boundary Ranger Station (elev. k,800 ft MSL, lat. 50°55'W, long. 115°08'W), Pigeon Lookout, and for a few stations in the Marmot Creek Basin (elev. 5,200 to 9,200 ft MSL, lat. 50°57% long. 115°10'W). Data for Kananaskis, Kananaskis Boundary and Pigeon Lookout stations were extracted from the "Monthly Record", Canada Department of Fig. 39- Study location 1, on the Kananaskis Forest Experiment Station, near the northeast shore of Barrier Lake, showing the location of infected lodgepole pine, Comandra plant plots, and instruments used during the study. Table I. Monthly and annual climatic summaries for Kananaskis, lat. 51°02'N, long. 115 03'W, elev. 4,560 ft MSL, for the period of record (1939 - 1968). TEMPERATURE "F1 JAN FEB MAR. APR MAY JUN JUL AUG SEP OCT NOV DEC YEAI 1 Daily mean 15.2 20.4 24.8 34.3 45.0 51.4 57-6 55.7 48.7 40.9 27.2 21.8 36. c Extreme maximum 59 61 64 75 82 88 93 92 86 80 66 64 93 Extreme minimum -50 -42 -41 -24 - 7 23 23 29 15 - 8 -32 -44 -50 1 PRECIPITATION inches Total 1.05 1.41 i.4o 2.49 3.06 4.08 2.49 2.78 2.28 1.44 1.15 1.17 24.7S Snowfall 10.24 14.07 12.89 20.18 7.48 1.53 0.00 0.16 0.34 8.88 10.52 11.24 97.5: SUNSHINE DURATION hours 2 Average 69 138 150 246 214 236 308 254 163 121 71 61 2031 Per cent of possible 26 41 59 45 48 62 57 43 37 27 25 43 Years of data 1 2 2 l 2 4 5 5 5 2 1 1 WIND SPEED mph3 Mean 6.8 6.8 5.9 5-9 5-3 5-4 5.0 4.8 5-1 7.2 7-2 7.7 6.1 WIND DIRECTION FREQUENCY <j? North 8 9 10 .'8 8 8 7 8 ' 5 . 4 6 6 7 Northeast 8 9 9 5 9 10 7 9 9 6 7 8 8 East 12 14 ll 10 14 13 14 14 13 9 9 9 12 Southeast 9 6 6 3 8 10 .10 10 10 9 8 10 8 South 4 5 6 4 6 5 8 10 7 5 4 5 • 6 Southwest 29 "27 • 28 39 23 25 26 25 29 40 38 37 30 West 17 20 21 18 26 20 17 15 16 16 18 17 18 Northwest 5 5 10 9 8 7 9 7 7 6 5 5 7 Calm 9 9 6 4 4 3 2 3 5 5 6 6 5 Years of wind data 12 13 12 10 17 20 21 22 14 15 11 12 -Data for period August 1939 to December 1968 Data for 1939-19^1, 1946-1947 and 1968, largely summers on 3Data for 1939-1941, 1946-1968 (summers only 1946-1954). Table II. Monthly and annual climatic summaries for selected stations in the Marmot Creek Research Basin, lat. 50°57'N, long. 115°10'W, elev. 5,300 - 8,000 ft MSL (1962.- 1967). Station Elevation Jan Feb Mar Apr 'May Jun Jul Aug Sep Oct Nov (feet) MEAN DAILY TEMPERATURE °F Confluence 1 5300 14.9 24.1 23.0 30.6 ko.2 49.2 56.9 56.2 48.4 39-6 23.4 Confluence 5 5800 12.8 22.6 21.2 31.0 38.3 47-3 54.3 53-5 46.2 36.9 22.1 Cabin 5 6800 12.6 17.4 15.9 21.9 34.2 43.2 53.0 51.4 42.3 32.7 19.4 MEAN DAILY MAXIMUM TEMPERATURE °F Confluence 1 5300 22.4 32.5 31-2 38.9 50.0 59-9 68.9 68.0 59-6 48.1 31.0 Confluence 5 5800 21.4 33-0 31-7 4l.O 48.7 58.8 67.8 66.8 57-7 46.9 30-3 Cabin 5 6800 19.5 26.5 25.6 30.8 43.9 52.8 66.0 62.8 51-9 4o.6 26.6 MEAN DAILY MINIMUM TEMPERATURE °F Confluence 1 5300 7.4 15.7 l4.7 21.5 30.4 38.5 kk.$ 44.3 137-2 ; .30.3..15.8 Confluence 5 5800 4.2 12.1 10.6 20.9 28.0 35-8 40.9 40.1 34.6 27.1 14.0 Cabin 5 6800 5-6 8.3 6.2 12.9 24.5 33-6 39-9 39-9 32.6 24.8 12.2 TOTAL PRECIPITATION inches Confluence 1 5300 -- -- -- -- -- 4.26 2.17 2.29 1-84-Confluence 5 5800 1.26 1.26 1.34 2.54 3.06 4.55 2.50 2.53 2.04 2.17 2.25 Cabin 5 6800 -- -- -- -- -- 3-47 2.26 3.09 2.33 MEAN WIND SPEED mph and DIRECTION Confluence 4 5600 4.3 5-2 5-3 4.7 5-2 4.8 5-2 5-1 k.8 5.9 4.2 W NWS , W, SE / NW /NW, , W „ w 'w N ms , Ws (W) (SE) (W/SE) (SE) (NW/SE) (NW/SE) (SW) (NW) Upper Ridge 8000 -- -- . -- -- -- -- 10.7 8.1 9.7 12.4 11.2 -- NW/SW SW/W SW SW sw/w MEAN EVAPORATION inches Confluence 5 5750 -- -- — — — 3-29 4.86 3-88 2.70 lAnemometer moved to 150 ft tower at 5,900 ft in July 1967, previously on 33 ft tower. Dec 17.2 15.3 14.7 24.2 23.4 21.3 10.2 7-3 8.0 k.9 13.5 w Years Yr T,of , x Record 35-3 5 33.4 5 29.8 2-4 44.5 44.0 39.0 1.55 27.1 4.9 NW/W 5 5 2-4 25.9 5 23.0 5 20.7 2-4 6 4-6 4 2-3 2-3 2-4 1-4 4-6 Table III. Monthly temperature and precipitation averages at Kananaskis Boundary Ranger Station, lat. 50°55'N, long. 115°08'W, elev. 4,800 ft MSL (1962-1968)1, and Pigeon Mountain Lookout, lat. 51°03'N, long. 115°0VW, elev. 6,000 ft MSL (1960-1968)% compared with those at Kananaskis for a similar period (1963-I968). Station JAW FEB MAR APR MAY JUW JUL AUG SEP OCT WOV DEC YR MEAN TEMPERATURE °F Kananaskis Boundary 12.9 19.2 23-5 30.8 1+2.5 49.5 56.3 54.8 48.8 42.2 24.6 12.3 34.8 Pigeon Lookout2 -- -- -- -- -- 47*3 56.1 55-1 ^5-3 Kananaskis 15.I 26.7 25.6 34.5 43.6 50.6 57-4 55-8 49.3 40.9 26.3 18.1 36.2 TOTAL PRECIPITATION inches Kananaskis Boundary 1.45 I.87 1.6l 1.72 3-02 '4.74 1.68 2.05 2.03 1.62 1.71 2.59 26.09 Pigeon Lookout2 -- -- -- -- 5.98 2.71 2.29 3.09 I.99 Kananaskis l.l4 0.82 1.57 3-39 3-22 5-12 I.83 2.47 2.78 1.43 1-33 1-88 26,98 •^-Temperature data began 1963. 2Temperature data 1963, I965-I968; Precipitation data i960, I962-I968. - 56 -Transport (very recent months from the Climatological Station Report, Forms 230k and 2306), and those from Marmot Creek stations. from the "Compilation of Hydrometeorological Record, Marmot Creek Basin", Vol ume I - III, Water Survey of Canada, Department of Energy Mines and Resources, Calgary, Alberta. The winter climate is characterized by an alternation of cold, dry, rather still periods, with periods of comparatively warm, dry, windy, chinook air, which gives to the general area of southwestern Alberta one of the greatest winter temperature ranges. In December 1968, Kananaskis experienced a 108°F temperature range. The variation in winter weather is determined largely by type and circulation of air masses. The main circulation is from the north and west which results in the predominance of maritime Polar (mP) and continental Arctic (cA) air masses, with the occasional occurrence of unstable and very cold maritime Arctic (mA) air masses (Penner 1955). Maritime Tropical (mT) air masses may occasion ally enter the region aloft from the south in winter ascending the Polar airmasses but never reaching the surface (Anonymous 1956). Week-long periods of thawing may occur in all winter months when mP air enters the region, with temperatures in the 50's not being uncommon. LongLey (1967b) found that Kananaskis, on average, had 29 Chinook days (above !+0°F) during the winter months December to February, which was two more than Calgary, and 19 more than Banff. In contrast, periods of sub-zero temperatures of a duration of a week or more are comparatively rare. Ex treme low temperatures occur when stable cA air stagnates over the eastern slopes of the Rockies and western Prairies. Often twjr. c 1. _ :'• '..*;: - 57 -temperatures in the valley are lower than on the higher slopes as cool air collects in the valley under inversion conditions, and Chinook air, or warm air masses under subsidence are only experienced at the higher levels. The Chinook is characterized by a strong westerly flow of mP air with lee waves forming troughs and crests roughly parallel to the mountain ranges, which result, when warm dry air is drawn into a mid-latitude low pressure center in the lee of the Rocky Mountains. The air descends the leeward side of the mountains at the dry adiabatic lapse rate which brings high temperatures and low humidities to the areas where the Chinook reaches the ground. Condensation and clouds form near the crest of each standing wave with the ascending of air, giving the well known Chinook Arch. Much of the red belt conifer foliage injury observed on valley slopes, which is very prominent in some seasons, has been attributed to the abrupt alternations of cool arctic air and warm chinook air (Henson 1952; MacHattie 1963). The winters are relatively dry, only about 30% of the annual precipitation at Kananaskis (Table I) occurs during the six winter months, October to March. Snow accounts for nearly 40% of the annual precipitation of about 25 inches at Kanan askis, but much of this falls in April, which is the highest snowfall month. The heaviest single snowfall on record at Kananaskis occurred in June 1951 when 33 inches fell in two days. At the higher levels of the valley, Storr (1967) found that 70 to 75% of the annual precipitation occurred as snow or a mixture of rain and snow. June is the wettest month of the year, and August the month of greatest precipitation variability. The variation of summer precipitation and temperature depends upon the - 58 -duration and frequency of subsiding dry mP air from the west, mT air from the Gulf of Mexico which may bring considerable moisture to the Front Ranges, and in late summer and autumn the continental Tropical (cT) air which may penetrate from the south and give hot, dry spells. In spring and autumn, mA air may enter the region for a few days but it is absent in summer. Valley bottom stations receive less precipitation than valley slope stations. Storr (1967) found that on the average sum mer rainfall increased about 1.5 inches per thousand feet in the east-facing Marmot Creek basin with a levelling off at about 7,600 feet MSL. Although the years of record of sunshine are short, the low total hour values for May and June, when less than 50 per cent of the possible duration was recorded (Table I),'reflect the passage of lows just to the south of the area which bring cloudy, moist air to the region. There is generally less variation in the summer month temperatures than in winter temperatures. Temperatures above 8o°F are experienced in several months but the high elevation of the area, all above 4,200 feet, is responsible for cool summer nights and lower daytime tempera tures than occur on the hot, dry Prairies to the east. Maximum tempera tures usually occur near the end of July. Differences in the average daily maximum and minimum temperatures during the summer months are approximately 10 to 15°F between the upper and lower areas of the valley (Munn and Storr 1967). Frosts can occur in any month, and the average frost-free period for the years 1951 to 1964 at Kananaskis was 62 days, with the average date of the last spring frost, June 21, and the first autumn frost, August 22 (Longley 1967a). During the summer - 59 -months there were many occurrences of low night relative humidities in the Kananaskis valley (MacHattie 1966) when chinook-type winds occurred. Webb (1965) showed that these low humidity nights were associated with air subsidence under conditions favouring lee wave formation. MacHattie (.1966) found that daily minimum humidities were' remarkably independent of vegetation cover, site and topography, and that the increase with elevation was very slight from valley bottom to 1,000 feet elevation up the valley sides. Nightly maximum humidities were more variable and frequently decreased abruptly with elevation just above valley bottom (11% in 300 feet) with a more gradual decrease above this level (less than 1% per 100 feet). The mean monthly wind speeds at Kananaskis were higher in the winter months, when chinooks are frequent, than in the summer months (Table I). December had the highest mean wind speed and July, closely followed by August, the lowest. The strongest winds come from the south west or west at all times of the year. The dominant wind direction in practically all months of the year was from the southwest, although in certain years winds from the southeast or east were dominant in some of the summer months. MacHattie (1967) indicated that this southeast wind component was dominant at night, although of only low speed, and was typical of a downvalley wind coming from the Lusk Creek sub-valley. He showed that the wind components across the main valley had a more pro nounced day-night cycle than the wind components along the valley, both at Kananaskis and in Marmot Creek at 5680 feet. No appreciable diurnal oscillation of winds occurred up and down the valley at Kananaskis in - 6o -summer, but he found that the southwest component could be dominant for most of the day, or under certain conditions, during only the daylight hours. In the summer months there was usually a marked maximum wind speed in mid- and late-afternoon, with a minimum occurring around sun rise. At a valley bottom station near the confluence of Marmot Creek and the Kananaskis River there was a morning-evening slope wind cycle, with the wind blowing toward the more intensely insolated slope. Munn and Storr (1967) also showed the prominence of the northwest-southeast sub valley winds in August at the 5,600 foot level on Marmot Creek. They found a wind speed maximum just before sunrise with a downvalley wind, and another maximum in the early afternoon with an upvalley wind. The minima, about 0800 and 1800 hours, were associated with wind shifts from downvalley to upvalley and vice versa. At the Confluence h station at 5,600 feet (Table II), very close to the location used by Munn and Storr (1967), there was also a high occurrence of downvalley northwest winds in many months, and of upvalley southeast winds in the months March to July, but at the ridge station at 8000 feet, the winds were predominantly from the southwest, and at a monthly mean velocity two or more times those of the lower station. A comparison of monthly mean temperatures and monthly preci pitation for the period 1963 to 1968 (Table III), showed that Kananaskis had warmer summer months than Kananaskis Boundary by about 1°, and warmer winter months by as much as 7°. During the winter months Kananaskis re ceived less precipitation, but more during the summer months than Kanan askis Boundary. Temperatures at Pigeon Lookout were 1° cooler in July - 61 -and August, and 3 to 5° cooler in June and September than at Kananaskis, and Pigeon Lookout generally received more precipitation. Monthly temp eratures were also generally lower in the Marmot Creek Basin (Table II) than at Kananaskis. Precipitation from April to June was higher at Kananaskis than in Marmot Creek. VEGETATION The area falls largely within the Srub-alpine Forest Region (Rowe 1959); which is found between the approximate limits of 5,000 and 6,500 feet, where the climax species are Engelmann spruce, Picea  engelmanni Parry, and western white spruce, Picea glauca (Moench) Voss var. albertiana (S. Brown) Sarg., although large areas are covered by the sub-climax species, lodgepole pine. At higher elevations, between 6, 500 feet and the timberline, the Alpine Forests are characterized by the dominance of alpine fir, Abies lasiocarpa (Hook.) Nutt., and alpine larch, Larix lyallii Pari. At the northern end of the Kananaskis valley, Douglas fir, Pseudotsuga menziesii (Mirb.) Franco, a climax species of the Montane Forest Region is present on warm, dry slopes at lower ele vations (up to 4,750 feet). Balsam poplar, Populus balsamifera L. occurs on the alluvial soils of the valley, and trembling aspen, P. tremuloides Michx., competes with lodgepole pine, as a pioneer species following fire, on the lower slopes of the valley where brown forest soils predomin ate. Fire has been the most important factor in forest stand development in the area, and throughout the Sub-alpine and Foothills Forest Regions of Alberta (Horton I956; Smithers 1962). Smithers (1956) records a - 62 -major fire in the area around 1865, and other important fires occurred in 1910, 1920 and. 1936. This lead to lodgepole pine stands of even age, sometimes overstocked with very high densities per acre. A number of workers have studied the silvical and ecological characteristics of lodgepole pine in the area and much of this work has been summarized by Smithers (1962). Forest cover maps have recently been published for the Kanan askis Forest Experiment Station (Canada Department of Forestry and Rural Development, I967), and Marmot Creek Watershed Research Basin (Depart ment of Forestry of Canada, I965), which indicate the major species pre sent, their average height class, and crown closure density, as well as showing poorly drained and rock outcrop areas. The forest and alpine cover types and habitat types of the Marmot Creek Watershed Research Basin, will be published by Kirby and Ogilvie in 1969. This will re present the first detailed ecological vegetation report for any.area in the Kananaskis valley. The report describes eleven forest habitat types and ten alpine habitat types, and indicates that kQfo of the Marmot Creek Basin consists of non-productive forest area — alpine forest, meadow and rock. The rest of the area is largely of the spruce-fir-pine cover type, with areas of pine and pine-aspen cover types at lower elevations, where they became established after the 1936 fire. STUDY LOCATIONS The study locations (Fig. 38) occurred in both mixed and even aged stands of lodgepole pine, except location 8 which was in a grass - 63 -clearing. The even aged stands were approximately 10, 2k, 28, 32 and kO years-old, and became established after fire. There was also one regeneration area with few trees over 5 years old (location 9). The mixed stand (location l) had trees ranging in age from 7 to 60 years, with most trees grouped in the 21 or 52 year age class. Location 3 was situated in the lower levels of the Marmot Creek Watershed Research Basin near the climatological station, Confluence 1 (Table II). Laboratory experiments with aeciospores, daily counts and ger minations tests, were carried;.'out in the laboratory at the Kananaskis Forest Experiment Station, or at the Forest Research Laboratory, Cal gary. The spore dispersal and microclimatic studies were carried out at locations 1, 2, 3 and 8 (Fig. 38). The aeciospore collections for daily germination tests came from location 1. Spore collections for ex perimental purposes came from most of the locations. Collections of the associated microflora and microfauna came from locations 1, 2, 3, 5 and 7. Measurements and observations of canker growth and activity, and the incidence of associated microflora, microfauna and rodent damage, were made at all.locations except location 8. Besides the main area of study, additional information was obtained from many points within the Sub-alpine, Montane and Foothills Forest Regions between Robb (53°13'N, ll6°58'W) and Waterton Lakes Na tional Park (49°03'N, 113°55'W"). The information mainly consisted of measurements and observations of canker growth and activity, and inci dence of associated microflora and microfauna. Collections of the associated microflora were made from a number of places. Material for - 6k -the rearing of associated microfauna from Pinus contorta was obtained from Baril Creek, Saskatchewan River Crossing, Cline River, and near Robb, and from P. banksiana, kO miles south of Rae, N.W.T. Extensive survey for distributional data were carried- out over the southern portion of the Province south of 56°N. Additional distri bution data and material for rearing associated microfauna was gathered by personnel of the Forest Insect and Disease Survey in the course of their regular duties. - 65 -AECIOSPORE PRODUCTION PERIODS OF AECIOSPORE PRODUCTION AND ENVIRONMENTAL  FACTORS AFFECTING SPORULATION ": • ' Methods Aecial cankers were observed at regular intervals from early May to October at locations 1 and 2 (Fig. 38) in the years 196k to .1968, at location 3 from 1966 to 1968 and at location 9 in 1968. At location 1 observations were made 2 or .3 times a week in the years 196k to I967 and weekly in 1968. At location 2 observations were made several times a week in 196k and once or twice a week from 1965 onwards. Locations 3 and 9 were visited at least once a week. In some years observations were made on a daily basis on selected trees at location 1 to establish ini tiation of sporulation and duration of 'production from individual cankers and individual aecial pustules. Sporulation was considered to have com menced with the rupture of the peridium of the aecium and the release of the first aeciospores, and to have ceased when no further aeciospores were present in the ruptured aecium. Three to 30 individual pustules were marked on individual aecial cankers with lettered or numbered pins prior to rupturing of the peridia. Efforts were made to select isolated pustules which, would not coalesce with nearby pustules. Observations of aeciospore sporulation were also made at other locations at irregular intervals in the various years and provide general information on the sporulation period. Daily temperatures, relative humidity, rainfall, wind and radiation data were recorded on hygrothermographs, rain gauges and rainfall recorder, wind speed and direction recorder and actinograph - 66 -placed at location 1 from mid May to late August. Location of instru ments and infected trees is shown in Fig. 39 j and- additional details of the instruments are given in the aeciospore dispersal section. For April and the first half of May in each year weather data was obtained from the Kananaskis'climatological station, 330 feet from the nearest infected tree of the study area. Weather data was also recorded at locations 2 and 3 for the period of study. In addition to observing the phenology of aeciospore produc tion in the different years at the various locations, the phenology of the other spore states of the fungus was observed. At location 1, Comandra plants were observed two or three times a week for the presence of the uredial and telial states on four plots, each one square meter in area (Fig. 39)- Some observations were also made on the phenology of lodgepole pine and Comandra plants at this location. r'''y.v'.Vi;. Results Table IV summarizes the beginning and end of aeciospore sporu-lation at locations 1, 2,t:3 and 9 for 1964 to 1968, and indicates the main spore production period. Aeciospore production varied widely with seasonal conditions and locality. Aeciospore sporulation began from mid-May to early June. At higher elevations the period was one to two weeks later, and trees on sites with a north or west aspect were later in initiating sporulation than those with a south or east aspect. The main period of aeciospore sporulation lasted1 three to five weeks, us ually commencing at the beginning of June, although in I965 and 1968 spore production was concentrated in a two week period. For three years, sporulation ended in the second1 -half of July, but in the other two years - 67 -Table IV. Date of beginning and end of Cronartium comandrae aeciospore production, and main sporulation period at a number of loca tions in the years 1964 - 1968 inclusive. Year Location no. Began Ended Main Period 1964 1 June 2 July 19 June 3 - June 30 2 June 1 July 18 June 8 - July 7 I965 1 May 24 July 30 May 30 - June 13 2 May 18 July 24 # 1966 1 May 13 Aug. 18 May 26 - July 3 2 May 26 Aug. 10+ June 1 - July 4 3 Aug. l6 1967 1 May 31 Sept. 1 June 5 - July 8 2 May 30 Aug. 3+ June 10 - July 10 3 May 31 Aug. 17 June 12 - July 10 1968 1 June 6 July 18 June 18 - June 28 2 May 31 July 22+ June 13 - June 23 3 JJune 2 July 31 June 18 - June 28 9 June 2 July 30 June 13 - June 28 * Information not complete spore production continued to the middle or end of August. In I966 a sporadic and very light resumption of aeciospore production, induced possibly by warm weather, occurred at location 2 in September and October. The scattered aecial pustules, rarely more than 2 mm in diameter, occurred in the pycnial zone of the same year. Table V gives the date of beginning and end of spore production for a number of cankers at location 1 for the years 1965 to I968. Figs. 40 and 4l summarize the weather and phenology data for the various rust states at location 1 in the same years.' Most aecial cankers in any one year began sporulation at approximately the same time although some can kers were consistently early, e.g. #2721, but others were late, e.g. #2712. From observations there was a tendency for sporulation to occur Tree No. Began 2516 June 1*+ Table V'., 1965 Date of beginning and end of spore production, and total number of days of production from individual cankers at location 1 from 1965 to 1968 inclusive. Main spore production period is included for 3 to 6 cankers for the years 1965 to 1967. Ended July 30 2710 May 10 July 2k 2711 2712 2713 2715 2716 2717 2719 2720 2721 May 29 June 1+ May 27 June 8+ June h No data May 2k May 30+ May 22 2722 May 2k 2723 May 2k 2728 June 2+ June 26 July 26 July 26 July 17 July 20 July 16 June 20 July 12 July 5 July k July 22 Main period June 5-11 July 2-10 May 30-June 11 June 24-27 June 5-13 June 5-17 May 26-June 11 Total no. of days 57+ 60 29 56+ 61 40+ kl 5k 22+ 52 1+3 k2 51+ 1966 Began Ended No production June 1 June 15 (1 pustule only) May 2k May 30 May 20 June 19 Aug. 18 Aug. 15 10 May 26 Aug Tree dead June 2+ July 30 May 26 Aug. 15 Tree dead May 13 July 27 May 22 July 8 Tree dead May 25 July 8 Main period June 2-, July k May 25-July 3 May 22-July 3 Total no. of days 1U 27 81 77 59+ 82 76 kd k5 1967 Began Ended Tree dead May 28 Aug. 25 May 31 June 5 June 2 June 23 Aug. k Sept. 1 June 6 Aug. 7 June 2 May 30 Aug. 15 Aug. 15 Main period Total no. of days June 2-June 25 June 11--July 18 June 5-Aug. 7 May 30 June 27 (tree dead July 18) May 31 July 7 May 31 Aug. 10 90 2k 60 92 63 75 78 29 38 72 1968 Began Ended June 6 July 6 No production June 15 July 8 June' 8 June 30 (tree dead July 22) Total no. of days 31 June 10 July 16 June 10 July 18 June 10 July 3 June 6 July 18 0 2k 23 Canker area dead — 37 39 21+ >*3 CA co Fig. hO. Summary of the daily maximum and minimum temperatures : and relative humidity taken from hygrothermograph records, the daily rainfall, and the phenology, of the aecial, py cnial, uredial and telial states of Cronartium comandrae, " May to August in 1965 and 1966. - 69 -10 15 20 15 30 10 15 20 25 30 10 15 20 15 30 10 15 10 MAXIMUM 20 AND MINIMUM TEMPERATURE °C -10 RELATIVE 24 HUMIDITY 16 NO.OF HOURS 80 AND 100% * 0 2.0 1.5 RAINFALL INCHES 1.0 0.5 0 AECIOSPORES PYCNIOSPORES UREDIOSPORES TELIOSPORES K> 15 20 25 30 MAY 10 15 20 25 30 JUNE 10 15 20 25 30 JULY 5 10 15 20 AUGUST 10 15 10 25 30 10 15 20 25 30 10 15 20 25 30 MAXIMUM jo AND MINIMUM TEMPERATURE "> °C RELATIVE HUMIDITY 16 NO. OF HOURS 80 AND 100% RAINFALL 1.0 INCHES AECIOSPORES PYCNIOSPORES UREDIOSPORES TELIOSPORES 10 13 20 25 30 MAY 10 15 20 25 30 JUNE 15 20 25 30 JULY 5 10 15 20 AUGUST 80n HUMIDITY 100| % Light Moderate Heavy Fig. 1+1. Summary of the daily maximum and minimum temperatures and relative humidity taken from hygrothermograph records, the daily rainfall, and the phenology of the aecial, py cnial, uredial and telial states of Cronartium comandrae, May to August in 1967 and 1968. - 70 -10 15 20 25 30 10 IS 20 25 30 10 15 20 25 30 10 15 20 MAXIMUM 20 AND MINIMUM TEMPERATURE 10 °C -10 24 r RELATIVE HUMIDITY 16 NO. OF HOURS 8 80 AND 100% o 2.0 1.5 1.0 0.5 RAINFALL INCHES AECIOSPORES PYCNIOSPORES UREDIOSPORES TELIOSPORES l_ MAXIMUM AND MINIMUM TEMPERATURE °C RELATIVE HUMIDITY NO. OF HOURS 8 80 AND 100% o RAINFALL INCHES AECIOSPORES PYCNIOSPORES UREDIOSPORES TELIOSPORES L_ 5 10 15 20 25 30 MAY 10 15 20 25 30 10 15 20 25 30 5 JUNE 10 15 20 25 30 10 15 20 25 30 JULY 10 15 20 25 30 1 1 5 10 15 20 AUGUST 10 15 20 25 30 MAY 80n HUMIDITY , 100i % Ugh, Moderate Heavy 10 15 20 25 30 JUNE 10 15 20 25 30 JULY NR = No Record 5 10 15 20 AUGUST - 71 -first from cankers on the smaller branches, seedlings and young trees, all having thin smooth bark, somewhat later on large branches with thicker bark, and later still on stems of older trees with a thicker rough bark. This was probably because the developing aecia were deeper seated in the thicker roughened bark than in thin smoother bark, and because of the mechanical problems of rupturing thick bark. Usually the aecial blisters of thick barked cankers erupted in the bark fissures, but removal of bark often exposed a large aecial area underneath where spores remain unexposed for much of the season. Cankers on trees in the open or on a south or east aspect of a tree also began sporulation slightly earlier than those in a dense stand or located with a dominantly west or north aspect. Even on trees which were nearly girdled by the fungus, sporulation began from pustules on the southeast or south aspect and was last from those on the north. Cankers a few feet up the stem also sporulated before basal cankers situated in the thicker bark zone of a tree. Initiation of sporulation was about two to three weeks after evidence of shoot bud break on the lodgepole pine which usually commenced in late April or early May. Similarly, growth of Comandra plants began before aeciospore sporulation. Growth amounted .to an inch or more of aerial plant development by May 12 in 1965 and May 10 in I966, and the Comandra plants were flowering by the last week in May in favourable sites. Development was later in 1967 and 1968. In 1963 and I96U Comandra plants were first noted as producing urediospores at location 1 on June 18 and 28 respectively. In later years the incidence of infection and uredial and telial development was recorded on Comandra plants on the four plots at .-.location 1 (Fig. 39). Each plot having an average of 53 - 72 -aerial shoots each year. These plots were laid out within 100 feet of infected trees, and in most cases were within 1+0 feet. In 1966 no in fections were.recorded on the plants until 39 days (on June 28) after aeciospore release from the adjacent trees (Fig. 1+0). A similar long interval occurred in 1965 (Fig. 1+0), but in I968 (Fig. l+l) uredia were present 12 days (on June 17) after the first aeciospores were recorded. Early springs may advance and late springs retard somewhat the advent of aecial production. Sporulation was three weeks earlier in 1966 than in 1968. Early May (May 1-15) was unusually warm in 1966 there being twice the number of degree days above 0°C than in the same period in 1967 and 1968, and one and a half times the number in 1965. During May 1966 the region was influenced by a prevalence of mP air, but in 1967 and 1968, May was dominated by mA and cA air masses. Wo cA air was experienced in May 1966", but this air mass was present for a few days in the first half of May 1967 and 1968, bringing well below normal tem peratures to the region. May 1965 did not experience the cold cA air, as occurred in 1967 and 1968, but was dominated by mA air during the month. The presence of warmer air masses, with their corresponding warmer surface temperatures, could well account for sporulation being earlier in 1966 than the other years, and~.'why sporulation was delayed to a lesser extent in 1965 compared with 1967 and 1968. April had also been warmer in I965 and 1966 than in the latter two years. Initiation of sporulation in most years seemed to start on a day of, or a day following, rain, suggesting that moisture was required to rupture the peridia. The end of the main spore production period often coincided with a heavy rainfall which washed the spores from the aecia. This was most noticeable in 1965 and - 73 -I966. The warm dry weather of July 1966 and 1967 tended to extend the spore production period. There was a great range in the duration of spore production from individual cankers. (Table V) and pustules (Table Vi). Period of spore production on some trees was consistently short, e.g. #2711, and on others long, e.g. #2713 and 2719. Generally production tended to be longest on the largest trees, shortest on the smallest, with the inter mediates in between. As well as variation in the spore duration from in dividual cankers there was considerable range between individual pustules on the same canker (Table Vi). In both years individual pustules pro duced spores for a shorter period on #2715, & younger, thin suppressed tree, than on the older dominant tree, #2719- Tree #2717 was similar to #2719, but #2710 was again a smaller diameter tree, and on average, sporulated for a shorter period although the pustule sporulation range was similar. Pustules tended to sporulate first towards the center of a canker and last towards the periphery. Generally pustules on one canker sporulated at about the same time, although on occasion there could be two waves of pustule sporulation, as occurred on tree #25l6 in 1965 (Table V) which accounted for the two main periods of spore production. Table VII summarizes the sequence of development of individual pustules on two cankers in I967, and shows how a large percentage of the pustules rupture within a few days of each other and produce spores for very simi lar periods of time. - 7h -Table VI. Average number of days and range of days of aeciospore production for individual pustules on several cankers in 1966 and 1967 at location 1. Year Canker No. No. of pustules Spore Production followed Aver. no Range of days 1966 2715 8 21 17-38 2719 6 k6 21-64 1967 2715 20 3k 25-58 2719 9 60 43-75 2710 30 52 37-87 2717 3 60 53-71 Table VII. Summary of sequence of sporulation for 30 and 20 pu; tules on . two cankers at location 1 in 1967. Canker Date State of Aeciospore pustules No. Closed Ruptured Sporulation Finii Main Diminished 2710 May 31 13 2 15 0 0 June 1 6 4 20 0 0 June 3 4 1 25 -0 0; June 6 ,'0 5 25 0 0 June 8 0 1 29 0 0 June 9 0 0 30 0 0 June 23 0 0 Ik 16: 0 June 28 0 0 :o 30 0 July 25 0 0 0 . 11 19 Aug. 9 0 0 0 6 24 2715 June 7 19 1 "0 0 0 June 8 12 7 1 0 0 June 9 9 10 1 . 0 0 June 12 4 14 2 0 0 June 16 2 6 12 0 0 June 17 0 3 17 0 0 June 22 0 0 20 0 0 July 4 0 0 0 20 0 July 25 0 0 0 k 16 - 75 -"V ' ' Discussion Annual production of the aecia may be short lived on small seedlings, young trees and small diameter branches. Production may only last for two or at the most, three years on this small diameter stem or branch material before the fungus effectively girdles and kills the stem or branch. If branch infections are sufficiently close to the main stem the fungus will grow into the main stem and continue to pro duce spores annually, although there is often a year without aecial production at time of transfer from a killed branch to the stem. On large branches and stems with thick bark, aecia may be produced annually until that portion beyond the canker is killed.as a result of the gird ling action of the fungus. This may require many years and depends on the size and state of the infected tree. Krebill (1965) found that the majority of cankers of his study originated from 20 to 50 years ago, and that a few were more than 100 years old. In the present study some trees under observation succumbed every year, but the period of study was not long enough to follow fresh stem cankers from initial stem in fection to stem girdling, although one branch to stem infection had al most girdled a 1.4 inch d.b.h. stem in three years. Evidence of annual rodent chewing on one 10.0 inch d.b.h. tree indicated that the infection had been present for at least 13 years and probably for 20 years. Dur ing the period of infection there is an annual increase in canker size, thus theoretically, a greater volume of aeciospores is produced each succeeding season until death of the affected part occurs. In most cases, however, aecial. production is interfered with by the activity of other fungi, insects or rodents, at least in some portion of the canker, - 76 -and there is often a reduction in annual aecial production despite the increased canker size. The effect of these biological agents on produc tion is discussed later. Mielke (19^3) noted the correlation between size of canker, age of host tissues and bark thickness, and aeciospore production for C. ribicola. Death of stem or branch cankers in any one year in one locality will greatly reduce the volume of aeciospores pro duced the following year in the same area. In 1968, a dying tree pro duced only two aecial pustules whereas in the three previous years"it had produced abundant spores (#2713, Table V). In 1968 infection on the nearby Comandra plants was low and late occurring, whereas in other years it appeared about 10 days after the first aeciospores were released and was abundant. Widespread canker death may therefore be a factor con tributing to the wavelike 'character of spread and intensification of the rust. Krebill (1968c) has recently published information on the phenology of C!. . comandrae in the Rocky Mountain States. Data from a number of plots showed great variation in duration of aeciospore pro duction, varying, from 3 to 17 weeks, but aecia were only abundant through June and July. Development of aecia was earliest at the low est elevations, and the duration of aecia was shortest at the higher elevations. Boyce (1916) reported that CJ. comandrae aecia usually commence sporulation in late April or early May, and finish by late June or early July. Spaulding (1922) gives the duration of C_. ribicola aeciospore production in different years for a number of areas in east ern North America. Duration in most years laster for k8 to 90 days, although he gives extremes of 12 and lk-0 days. The 12 day period appears - 77 -to be very short especially as he records closed blisters appearing kl days earlier I The long aeciospore period resulted from late aecia being noted in mid-September at Amery, Wisconsin, possibly a second period of aecial production as occurred at location 2 of the present study in I966. Pennington (cited in Spaulding 1922) noted that more aecia were produced per canker in 1919 than in 1918 in the Adirondacks, also that there were more new sporulating cankers. However, he believed that fewer aeciospores were set free in 1919 than in 1918, from results of spore-trapping and observations on first generation uredia. York (cited in Spaulding 1922), working in the White Mountains, found that aecia matured for about 30 days on single cankers, but may take longer on larger branch or stem cankers. Rhoads (1920) followed production of 300 aecia on various trees in Maine from the time they first broke on May 3- He found that by May 20, only 63 aecia were still sporulating, by May 29 only 19 and none by June k. Others have reported periods of 20 to 30 days, and more than Ik days (cited in Spaulding 1922) for aeciospores to be produced by an individual aecium. In western North America, Lachmund (1933) reported that C_. ribicola aeciospore production at low elevations may begin as early as late February although more usual dates occur in late March or early April. At the higher elevations sporulation usually begins by June. The main aeciospore production period occurs between mid May and mid July, and heavy spore production may extend into August. Lachmund also noted that, rather frequently at lower elevations, a sporadic and light aeciospore production may occur in October or November. The aecia were characteristically small, and this second wave of aeciospore production - 78 -occurred under the influence of warm fall weather in years having an early spring. He also noted its occurrence on C. comptoniae and C_. filamentosum Pk. cankers (Lachmund 1929). Dates of aeciospore produc tion given by Mielke and Kimmey (1935) indicated that the main aeciospore period occurs from mid-April to mid-June, but may be a month later at higher elevations. Mielke (19*1-3) stated that during warm and dry springs the main aecial sporulation may only last two or three weeks, but during cool, wet springs may last two or more months. Lachmund (1933) noted that thick bark tended to retard aecial production, and Mielke (19*1-3) further discusses this factor in relation to the different soft pine species that C. ribicola infects. Lachmund (1933) found that for cankers that produce pycnia for the first time in a given year only 57$ produce aecia the following spring, the rest produce aecia in the second year, although a few do not produce aecia until subsequent years. . Cankers affected by secondary organisms may not produce aecia at all. BIOLOGICAL FACTORS AFFECTING AECIOSPORE PRODUCTION ;[••.:) Methods Observations were made on the frequency and type of associated fungi seen on the aecial zone of cankers at different locations. These fungi were isolated from the canker, cultured and identified. On one occasion pieces of aecial canker were shaken in distilled water for sev eral hours, and the resulting solution streaked on agar plates. Micro organisms which developed on the streaks were isolated and cultured. A large number of fungi, yeasts and bacteria were frequently found as sociated with fresh and stored spore collections. These microorganisms - 79 -were allowed to establish colonies on test agar media for a few days before being isolated and grown in pure cultures for identification. Records were kept of the frequency of these associated microorganisms during the daily germination tests at location 1 from 1965 to I967 in an effort to establish the importance of the various microorganisms. Tentative identifications were made ,;6f many of the isolates of organisms frequently encountered. Many!isolates, including the less frequently occurring, were identified or verified by taxonomic specialists. Insects were observed to be infesting the rust cankers in most areas visited, and in many of these areas the frequency and amount of insect damage was noted. On a few occasions insects were collected directly from the aecial canker surface and placed in preservation vials. • To gain a better idea of the number of insects which used the rust can ker as a habitat niche for all or part of their life cycle a cylindrical-sleeve insect cage was developed which could be placed around the rust canker on the infected tree stem or branch (Elliott and Powell 1966). The cage was constructed of plastic screening and cotton fabric closed by an open-ended zipper, the cage being attached to the tree by nylon draw strings at each end (Fig. k-2). The drawstrings held the cage in position when the zipper was opened for removal of insects, but to ensure that all insects were collected, the cage was removed and all folds of the cage material examined. Ten to 12 cages were operated at location 2 from 1965 to 1968, 8 to 10 cages at location 3 from 1966 to 1968, and 1 or 2 cages near location 1 for the same period. The cages were visited for insect collection at weekly intervals from mid May to October. Many of the cages were left on the trees overwinter and were protected by l/2-inch Fig. k2. A plastic screening cylindrical-sleeve cage used for col lecting insects from Cronartium comandrae cankers on stems of Pinus contorta. - 8o -- 81 -wire mesh to prevent damage from bears or other mammals. In many cases a cage was left around a canker for two or more seasons, thus any in sects with a long life cycle would have been obtained. In each year some new cankers were caged in each area. In addition to the above collection methods, insects were reared from infected cankers which were placed in cardboard or metal rearing containers in the insectary. These containers were usually placed in a cool area (about 2-5°C) for a period of 3 to k months in an effort to satisfy any insect requirements of a diapause period. Each collection for rearing came from a single canker, or as many as 20 can kers. The collections came from many locations in Alberta, a few from the Northwest Territories, and included some C_. comandrae cankers on Pinus banksiana. Generally the larger collections for rearing were made in September or October, thus any insects which overwintered in the soil had usually departed the environs of the canker by this time. It was often noted late in the season, that a number of late instar larvae or pupae had collected at the bottom of the cage, thwarted in their efforts to crawl or drop to the soil surface. Collected or reared insects were pinned, mounted or placed in an appropriate amount of alcohol for pre servation. Aeciospores that had passed through adult or immature insects were checked for viability by attempting to germinate spores obtained from faecal pellets. Estimates were also made of the number of spores in the faecal pellets from some species, and whether the faecal pellets were composed entirely of spores or partly of bark and other material from the canker area. - 82 -Observations were made on the frequency of old and fresh rodent chewing associated with the rust canker at about 20 locations in the years 1966 to 1968 and at a few locations in 196k and 1965. At most locations in I967 and I968 the amount of chewing and its effect on future aecial production was noted. The effect of resin flow on aecio spore production was noted. Results Microflora All the identified microfloral organisms that were isolated from aeciospore collections or directly from the aecial zone of the canker are given in Appendix I. The list gives a total of 6k organisms, although the number might be reduced if organisms only identified to genus proved to be identical with others identified to species. The list includes 8 bacteria and 56 fungi, including 5 yeasts or yeast-like fungi. The fungi and yeasts largely belong to the class Deuteromycetes (43) with the largest number belonging to the family Moniliaceae (26). Within the Moniliaceae, 18 belong to Penicillium and its closely re lated genera Paecilomyces and Spicaria. Many of these organisms are probably only air spora contaminants although, as such, they may well affect the viability of the aeciospores with which they come into con tact. Several of these organisms are common soil inhabiting fungi, e.g. Mucor, Botrytis, Aspergillus, Penicillium and Epicoccum. Polyporus  adustus Willd. ex Fr. is known to be a prominent pathogen on species of Populus, but was isolated from a very large number of spore collections at location 1 in 1966. An unidentified sterile mycelium with clamp con nections was also quite common in the same year. - 83 -Two of the organisms were commonly found sporulating on the aecial zone of the canker and played a large role in reducing aeciospore production and aeciospore viability. They were the purple mold, Tuberculina maxima Rost. (Fig. 43), and an undescribed dark green species, Cladosporium tax. sp. 1. The latter fungus does not fit any of the species described from rusts and has been given a taxonomic niche, number XLII, in the Commonwealth Mycological Institute, Kew, England, by Dr. M. B. Ellis (B. C. Sutton,personal communication 1966). Table VIII summarizes the incidence of T. maxima and the num ber of cankers observed during the years 1964 to 1968 at 3 to 23 locations. Table VIII. The recorded incidence of Tuberculina maxima on Cronartium  comandrae cankers on lodgepole pine and the total number of active and inactive cankers observed at various locations in southwest Alberta during the years 1964 to 1968. Year No. of No. of Total no. of Tuberculina Per cent locations locations active and infected Tuberculina observed with no inactive can cankers infected Tuberculina kers observed 1964 5 2 38 7 18.5 1965 3 0 26 5 19.2 I966 20 6 31+1 82 24.0 1967 23 10 424 55 13.2 1968 22 9 313 42 13-4 The number of cankers for which observations were recorded varied each year depending on the number of locations visited and on canker mortality. The number of cankers observed at any one location varied from 1 to 51 with most locations having 15 to 20 active cankers in the initial year Fig. U-3. Rough, cracked bark of the aecial zone of a Cronartium  comandrae canker infected with the purple mold, Tuber culina maxima, which is conspicuous as a darker area where the surface bark has been removed, or cracked. - 8k -- 85 -observations were recorded. On subsequent visits the same cankers were observed, but some had become inactive through the action of Tuberculina or other agents.. T^ maxima was not observed at six of the locations in any year, and the numbers of cankers infected at other locations varied from year to year. Examples of the variation.in different years is given for 7 locations in Table IX. At Altrude, Vicary and Marmot Creeks, and the area near Saskatchewan River Crossing, there was less tree mortality resulting from canker infections, but considerable mortality was ob served at the other locations, especially at Cline River, where the stands were generally slightly younger. At Cline River the site was particularly dry and 25 trees succumbed during a two year period. The number of Tuberculina infected trees at Altrude Creek and at the Wedge has remained relatively constant, but the number has fluctuated at the other locations, with a peak year in 1966 at most locations. On some cankers the Tuberculina completely covered the aecial zone and prevented any aeciospore production. On other cankers only small areas were covered and aeciospore production continued from the uninfected areas. Tuberculina did not appear to sporulate to any degree on the pycnial zone, but was most abundant on the aecial zone. Tuberculina was pre sent from the beginning of aeciospore production, although there was a marked increase in the production of its purple spores as the season advanced. Initially it appeared as a purple, velvety layer, but soon a mass of spores were produced which continued well past the normal aecio spore sporualtion period. No evidence could be found that Tuberculina was a hyperparasite of the rust, but it was certainly a parasite on the rust canker, not being found outside the rust zone on the tree. It was - 86 -usually restricted to the current year's aecial zone, or what would normally have been the aecial zone if aecial production had not been affected, although on occasions it spread into old aecial zones or into the current pycnial zone. Only current season aecial or pycnial cankers appeared to be susceptible to Tuberculina as no cankers-which had been inactive the previous season were recorded with Tuberculina. In 1966 and 1967 a general view was obtained that the amount of aecial zone covered by Tuberculina increased as the season progressed. In I968 eight cankers with Tuberculina, initially covering between 5 and 40% of the aecial zone, were checked at weekly intervals and there appeared to be no appreciable increase in the amount of the aecial zone covered by Tuberculina. Although there may be only a slight increase in the Tuber- . culina covered area during the aeciospore production period, there was generally an increase in the following season, and in many cases no aeciospores were produced. When the pycnial zone became infected there was no aeciospore production the following year, thus Tuberculina was a favourable biological control of C. comandrae. Tuberculina appeared to hasten the death of the rust infected bark. A tree practically girdled solely by C. comandrae tended to live and continue producing aeciospores for at least a year, but one girdled by C. comandrae and infected by Tuberculina was soon killed. At the two locations with Tuberculina ob servations for five years no tree was infected by Tuberculina for more than three years, and at the locations with three years of data no tree was recorded with Tuberculina for more than two years. Generally, Tu berculina infected the whole of the canker and the canker became inactive, but occasionally one area of the canker escaped infection and aeciospore Table IX. The incidence of Tuberculina maxima on observed active and inactive Cronartium  comandrae cankers at 7 locations during the years 196k to 1968. Year Altrude Creek Vicary Creek Mist Creek No. of cankers (A) No. infected (B) (A) (B) (A) (B 1964 5* 3 10 1 - -1965 2* 2 - - - -1966 7 5 23 9 20 k 1967 8 3 22 3 15 6 1968 5 0 21 0 11 l * Incomplete tally Marmot Creek The Wedge Cline River Saskatchewan RiVer Crossing (A) (B) (A) (B) (A) (B) (A) (B -• - 12* 3 - - - -- - 10* 2 - - - -25 13 16 3 33 9 17 k 25 6 20 2 18 2 16 6 2k 5 15 ' 5 7 0 16 2 - 88 -production continued from this area for a year or two. Table X shows some of the effect of Tuberculina and other factors on continued aecio spore production at two locations. At both locations there was a trend for the number of cankers producing aeciospores to decrease and the num ber of cankers becoming inactive to increase. Most of these cankers were inactive as a result of Tuberculina, although in both areas a few cankers were inactivated through rodent chewing. When the cankers with aeciospore producing zones or zones infected with Tuberculina were tagged at Marmot Creek in 1966,'.a number of other inactive cankers were observed in the stand but not tagged. None of these cankers became active after I966 and evidence indicated that many were probably in activated by Tuberculina. From the evidence of the present study the Table X. The number of Cronartium comandrae aecial cankers (l) producing aeciospores, (2) sporulating but infected with Tuberculina, (3) with inactive aecial zones infected with Tuberculina, (4) with inactive or dead cankers, at two locations during the years I966 to 1968. Year Cankers producing aeciospores.; 1966 19 1967 14 10 1966 14 1967 12 1968 4 Aecial zone Aecial zone active with inactive with Tuberculina Tuberculina Marmot Creek 7 6 3 3 l 4 Saskatchewan River Crossing 2 1 3 4 2 2 0 Other inactive cankers 0 Dead cankers 10 0 3 12 0 0 1 0 0 1 - 89'-build up and spread of Tuberculina in an infected stand is slow, but once established it is effective in inactivating the canker and markedly reducing the aeciospore production. Germination tests of Tuberculina spores gave average per cent germination of about 80$. Viability was maintained for a period of at least a year. The role of Cladosporium tax sp. 1. in reducing aeciospore production is not as clear. Generally it became obvious only on the aecial zone towards the end of the aeciospore production period, although evidence of Cladosporium infection could be observed at the beginning of the aeciospore sporulation period on the old aecial zones of the canker. The fungus was typically shades of dark green although at times, it appeared black-green. It sporulated abundantly in the latter half of the summer and covered both the aecial and pycnial zones of the canker. Some evidence was obtained that it parasitized the rust aecio spores, and it was an extremely common isolate from spore collections where its: mycelium was found throughout many collections. When present in stored spore collections, the viability of the aeciospores was usually reduced to zero. The incidence of Cladosporium is shown in Table XI for the years 1965 to 1968, and was generally lower than for Tuberculina at most locations. The percentage for 1965 was obtained from a very small location sample, and'if percentages from the same three locations were only considered in the other years percentages of 19.0, 23.4 and 15.8 were obtained, values much higher than those for the larger location sample. Nearly half (46$) of the Cladosporium infections occurred on the same cankers as Tuberculina was observed, but due to the very low incidence in 1968, Cladosporium was only recorded from two - 90 -Table XI. The recorded incidence of Cladosporium tax. sp. 1 on Cron artium comandrae cankers observed at various locations in southwest Alberta during the years I965 to 1968. Year 1965 1966 1967 1968 No. of locations observed 3 21 23 21 No. of locations with no Cladosporium 1 7 10 16 Total no. of cankers observed 26 341 424 313 Cladosporium infected cankers 47 40 .14 Per cent Cladosporium infected 30.8 13-8 9-5 4.5 locations in all three years. The low incidence of 1968 may have been due to most of the observations being made by another observer and also be cause many locations were visited at a slightly earlier date than in other years. Generally a higher percentage of the Cladosporium infections was associated with active sporulating aecial zones than occurred with Tuberculina and only a few occurred on non-sporulating cankers which didn't also have Tuberculina. Cladosporium was recorded on a few trees for four years and on many, aeciospores were produced following two or even three years of Cladosporium infection, thus Cladosporium did not appear to act as a biological control agent of the rust in the same way as Tuberculina. Despite the low annual incidence of Cladosporium, 80$ of the cankers at Marmot Creek were infected by Cladosporium during the three yecar:. recording period, and 68$ of the cankers at The Wedge during a four year period, Cladosporium was not observed at five locations in any year, and at a further five locations only one canker was observed infected in three years of observing. Other organisms observed infecting the active zone of the - 91 -aecial canker were white mycelial growths, which were later identified as Fusarium sp., MonociIlium sp., Coniothyrium olivaceum Bonard., and Sclerophoma pityophila (Cda.) Htihn. Their incidence in any year was very low, less than 1% in the years I966 to 1968. At one location they were present on 9% of the cankers in 1967, DUT-- were not observed at this location in 1968. Penicillium spp., notably P. tardum Thorn, P. godlewskii Zaleski, P. rugulosum Thorn, P. funiculosum Thorn series, P. brevi-compactum Dierckx. and P. nr. rolfsii Thorn, were all isolated on one or several occasions from the aecial zone, but all were probably incidental and not parasitic. Alternaria tenuis auct. and Epicoccum  nigrum Link were also occasionally isolated from the current year aecial zone, along with Sporobolomyces sp., Rhodotorula spp., and bacteria belonging to Arthrobacter sp., Pseudomonas and a number of unidentified species. A number of secondary organisms, not parasitic on the rust, were found to fruit on non-current year aecial zones of old cankers, such as Lachnellula arida (Phill.) Dennis, but they were only observed infrequently on C. comandrae, being more plentiful on P. stalactiforme can kers . Microfauna Insect damage was widespread on C. comandrae cankers. Table XII records the incidence of obvious insect damage on the cankers ob served during the years 1964 to 1968 at 3 to 22 locations. Data for 1964 and 1965 are based on only a small sample of observed cankers and are therefore less reliable, although the percentages are close to the range of the other years. Insect damage was recorded at each location on at least one occasion. At four locations more than 60% of the cankers - 92 -Table XII. The incidence of insect damage on Cronartium comandrae cankers on lodgepole pine and the number of cankers ob served at various locations in southwest Alberta during the years 1964 to 1968. Year Wo. of locations observed No. of locations with no damage Total no. of cankers observed Cankers with insect damage Per cent • insect damage 1964 4 2 47 15 31.9 1965 3 1 22 13 59-1 1966 20 4 299 137 45.8 1967 22 2 4o4 154 38.8 I968 22 5 317 132 ' 41.6 had some evidence of damage in each of the years I966 to 1968, and at one location over 88% were damaged each year. Much of the fluctuation in incidence was related to associated rodent damage, as with the re moval of infected bark there was little suitable habitat remaining for the insects. Table XII records only the obvious insect damage, often recorded on only one annual visit to a canker, and must be considered an underestimate of the incidence of damage. Often signs cf insect damage were not visible on the canker surface for aeciospores may cover the boring holes, or insects may mine their way into the canker through bark crevices. Insects may also have been present in the aecia and were not visible without investigation, this was especially so of young larval instars which only became visible at a later stage. Some larvae also took on the color of the spores when feeding on them and even a trained eye could easily miss them in a mass of aeciospores. Insect-damage took several forms, often the rust infected bark was heavily chewed and there was much boring with its associated frass (Fig. 44). Fig. kk. Basal canker of Cronartium comandrae with typical rough bark in aecial zone and showing evidence of insect damage. Note exit holes and Lepidoptera frass at top of canker, and further frass in lower rough zone. Fig. 45. Pupal chambers of Pissodes schwarzi scored into the sap-wood throughout the Cronartium comandrae canker area on small Pinus contorta stem. - 93 -- 9*+ -At the time of aecial sporulation the insects, both larvae and adult forms, fed extensively on the aeciospores, reducing the number available for dispersal. They similarly fed on the pycniospores and the pycnial drops which have a high concentration of sugars. Often the aeciospores became matted together and gave a mealy, bleached appearance, having lost their viability'. Much of the canker became a mass of insect faecal pellets composed almost exclusively of aeciospores, or with cer tain species of insects, a mixture of infected bark and spores. The phloem infected tissue often became a mass of feeding galleries and pupal chambers (Fig. 45). The insect frass often hung together on the surface of the canker as large aggregates held together by insect silk. The different forms of insect damage were largely dictated by the species of insect inhabiting the canker. In the spring and early summer the insects occurred mainly in the current aecial zone and past aecial zones towards the center of the canker. As the season advanced the larval forms and visiting adults were attracted more to the sporulating pycnial zone. By late summer and early fall the whole pycnial zone was heavily damaged. Many, insects overwintered in the canker, thus insects were often present in an aecial pustule before the peridium of the pustule broke open. On some occasions cankers were so heavily infested with larvae that very few aeciospores were dispersed from the canker despite being produced in great abundance. Many of these aeciospores were eaten by larvae, but many others were aggregated together and held by fine silk threads. All the identified microfauna — insects, mites and spiders, that were collected from the exposed surface of the canker or were reared from the infected cankers are given in Appendix II. The list contains a - 95 -total of 117 species. A few of the organisms only identified to genus, many being immatures or of the nondiagnostic sex, may be identical with others identified to species. The list includes 98 insects, 17 mites and 2 spiders. The most important insect:, orders represented were the Coleoptera with 25 species belonging to 12 familes, the Diptera with 17 species in 9 families, the Hymenoptera with 2k species in 7 families, the Lepidoptera with 10 species in k families, and the Collembola with 8 species in 2 families. The number of specimens of each species collected or reared from the cankers is recorded in Appendix II. In many cases not all specimens of a species were removed from a canker, also on numerous occasions the more common or frequently occurring species were not collected from a canker. This group of common species are usually indicated in the list by a plus (+) following the number, which indicates that they were far more common than the number of specimens collected would indicate. Mycetocoles, or animals associated with fungi, have.been classified in three main categories: mycetobionts, mycetophiles and mycetoxenes (Benick 1952). Animals which cannot complete their develop ment without utilizing the fungus as food are known as mycetobionts. Mycetophiles are those organisms that are not absolutely dependent upon fungi for development, while organisms that are simply chance visitors or users of the fungus are termed mycetoxenes. Many of the species collected from the canker in the present study are probably mycetoxenes — using the canker as shelter or food, or seeking out species upon which they feed which happen to be using the canker. Many of the Hymenoptera probably fall into this last category as they are largely - 96 -parasitic on other insects. A few of the species collected appear to be true mycetobionts — the Coleoptera, Epuraea obliquus Hatch, and two diptera, an unidentified Cecidomyiidae and Paracacoxenus guttatus Hardy and Wheeler. These three were very important in reducing the production of aeciospores and will be discussed in some detail. Among the others collected there appeared to be a number of mycetophiles which were utili zing the fungus but known more commonly to develop in other habitats. The most commonly observed species on the cankers were the oribatid mite, Ceratozetes sp., the flat brown nitidulid beetle, Epuraea  obliquus in the larval and adult stages, various diptera larvae, includ ing Paracacoxenus guttatus and an unidentified Cecidomyiidae, and larvae of a number of Lepidoptera. Epuraea obliquus was prominent on the canker throughout the aecial and pycnial sporulation period. Adult beetles were observed in unruptured pustules, and on two occasions a beetle was observed chewing through the peridium and effectively rupturing the pustule. Evidence from rearings and from finding specimens at the bottom of cages indi cated that the larvae probably pupate and overwinter in the soil or duff layer, from which the adult seeks out the canker on emerging in the spring. Some may pupate in the canker zone, but beetles were only reared once directly from logs. Adult beetles were fairly common from' mid May until August, and a few were present on the cankers through to September" or later (latest collected October 17, 1966). Generally from one to six'were observed on an individual canker. Epuraea were observed on 80$ of the cankers at location 2 in 1965, and at least 50$ of the cankers in 1966 and 1968. A similar high incidence was recorded at - 97 -location 3 in the years 1966 to 1968. The first larvae were found from mid June onwards, and some were collected as late as October 9 in 1968, although in other years few were observed after August. The adults and larvae became covered with aeciospores; the average number of spores on the body of 8 beetles was 580, and on 3 larvae was 1198. At the time of aeciospore production the Epuraea existed entirely on spores. A number of faecal pellets was examined from both adults and larvae, and only spores and peridial cells were present. The number of spores in l6 pellets was counted from adult beetles and gave an average of 586 spores, very few of which retained their form. These spores were checked for viability but none germinated. The number of aecio spores made inviable through being eaten by the Epuraea beetles and larvae •was enormous. Counts of adult Epuraea pellets from C_. comptoniae, which has a smaller aeciospore, gave lk^8 spores per pellet. The larvae fed extensively on the pycnial drops containing the pycniospores, and were observed crawling from one drop to another. Generally the larvae oc curred in larger numbers on an individual canker than the adults, but rarely more than 12 were counted. Epuraea also fed extensively on Tuber culina and to a lesser extent on Cladosporium. The larvae and beetles often carried a large number of spores on their bodies, which made them appear purple, and probably assisted in dispersing the spores over the canker zone to areas not infected by Tuberculina. In all thar- travels, the larvae tended to leave slime trails to which aeciospores became at tached. Other spores became attached to faecal pellets, and these attached spores were effectively prevented from being dispersed. On one occasion three beetles were allowed to crawl over a malt agar plate to ascertain what spores they might be carrying. On this occasion colonies of - 98 -Cladosporium herbarum and Cladosporium tax. sp. 1 developed. Epuraea and other insects no doubt act as carriers of many fungi and bacteria, and may well be responsible for a number of soil organisms being present among the aeciospores. Epuraea obliquus were collected or reared from six different locations (#1, 2, 3, 5, 7 and 9) in the study area on C. comandrae, and were reared from two canker collections on P. banksiana, one from a point 33 miles southwest of Rae, N.W.T., and the other from Twatinaw in northern Alberta. It was probably much more widespread than the collections would indicate, for it was collected on a number of other stem rusts in the region. Additional collection data is given for some of my E. obliquus collections on pine stem rusts in the paper by Parsons (1967). Diptera larvae were commonly observed eating among the aecio spores and often-caused the aeciospores to become-aggregated together in a mass of fine silk and gave a mealy bleached appearance to the spores. These larvae were often present in very large numbers on an individual canker and as many as fifty were counted. Between 25 and 70$ of the cankers at locations 2 and 3 were observed with dipterous larvae in the various years. An average of 253 aeciospores were counted attached to the bodies of two young larvae. The larvae took on the characteristic orange-yellow color of the spores. Inffact, similar dipterous larvae feeding on the white form of P. stalactiforme (Powell 1966), were all white in color. These dipterous larvae belong to at least two species and probably more, although the larvae of only two species were con sistently collected. One is an unidentified Cecidomyiidae which usually occurred in larger numbers. Cecidomyiidae larvae on C. comandrae cankers - 99 -were collected from locations 1, 2, 3, and Altrude Creek, Baril Creek and Saskatchewan River Crossing. Dr. J. R. Vockeroth (personal com munication 1966) considered the larvae collected in large numbers from the uredial and telial states of C. comandrae on Comandra plants, in widely scattered localities, and from alternate host plants of P. stalactiforme to be the same species. Unfortunately, only one Cecido-myiidae adult fly was collected from the cankers during the study period and could only be identified to genus, belonging to the tribe Lestremiini. This was obtained from a canker cage in June. Larvae of one Cecidomyiidae collection were identified as belonging to sub-family Lestremiinae. Cecidomyiidae larvae were first collected in the latter half of June on C. comandrae cankers, and an odd one could still be found in September and October. The other common dipterous larvae were tentatively identi fied in I965 as Drosophilidae or Lauxaniidae, but were later reared to adult and identified as the drosophilid Paracacoxenus guttatus (McAlpine I968), adults of which had been reared or collected earlier. This was the first information on the habitat of this species and on the larval and pupal stages. Paracacoxenus had only been reported once previously from sweepings over muddy ground along streams or lakes in a forested area. P. guttatus were collected from cankers at locations 1, 2, 3 and 5 in the study area, and adults were collected when visiting the pycnial drops of exposed cankers. Additional data is given for some of my Paracacoxenus collections in the paper by McAlpine (1968). The Paracaco xenus larvae, like the Cecidomyiidae larvae, fed extensively on the aecio spores and pycnial drops. Adults were taken in different years between June lk and October 17, but larvae were not observed until mid July in - 100 -most years. Specimens of another important diptera genus, which largely occurred in overwinter rearings, belonged to one or more species of Bradysia. These emerged over a long period of time in the insectary and the larvae presumably fed extensively on the infected bark, although they were not observed doing so in the field. Several adult Sciaridae, all of which were identified as Bradysia. were collected in the field. A Staphylinidae beetle, of the genus Atheta was also very prominent in rearings, but was not observed in the field. Coleoptera that were observed in the field, other than the Epuraea, and were ap parently feeding in the infected canker tissues and possibly on the spores, belonged to some of the following genera: Fissodes, Cylindro-copturus, Corticaria, Melanopthalma and Ernobius. The Curculionidae, Pissodes schwarzi Hopkins and Cylindrocopturus deleoni Buchanan were observed feeding on the'aecial zone of the canker, and their feeding galleries and pupal chambers (Fig. h^) could be found in the infected rust tissues. These weevils probably hasten the death of the tree as their larval galleries usually scored the outer sapwood. P. schwarzi and C. deleoni were collected or reared from material from locations 1, 3, 5, and from Cline River, Saskatchewan River Crossing, and an area 3 mies northeast of Robb, while P. schwarzi was additionally collected from the Yukon, and C. deleoni from location 2. The adult Corticaria,-Melanopthalma and Ernobius observed in the field, were all collected between late May and the end of July at locations 1, 2 and 3. Three genera of Lepidoptera were frequently encountered in the canker, their damage was characterized by aggregates of frass at the entrances to their feeding galleries and over their pupal chambers - 101 -which were often found in the fissures of the rough central portion of the canker. The greatest damage was done by the larger larvae of Dioryctria zimmermani Grt., or specimens identified only as Dioryctria sp., which could often be seen momentarily on the surface of the canker, and which pupated in a silk-lined chamber in the infected tissue. These larvae mine extensively into the phloem tissue and destroy large areas of the aecial and pycnial zone of the canker. Dioryctria were collected or reared from locations 1, 2, 3 and 55 Saskatchewan River Crossing and Cline River, as well as a number of other points in Alberta, the North west Territories, and the Cypress Hills, Saskatchewan. The larvae of the Laspreyresia and Coleotechnites (Recurvaria) were much smaller, but still caused some damage in the current or old aecial zone. On occasion an empty cocoon could, be found sticking out from the canker surface, or fromtthe associated resin crust. -The orbatid mite, Ceratozetes sp., was present at several lo cations in the study area-and could be found on most infected trees. However, it was not restricted to the canker area but could be found on many areas of the stem and apparently lived in the soil or duff layer. Large numbers were often found in the aecial zone of the canker from the beginning of spore production, but were encountered less often as aecio spore sporulation ended. A few of the other mites could have been phyto phagous, but most are probably 'classed as mycetoxenes. The Eupodidae and Tydeid mites taken from aeciospore collections were approximately the same color as the spores, and were probably, feeding on them. Aphids of the genus Cinara were observed covering the whole aecial and pycnial zone of a few cankers, and were attended by species - 102 -of ant belonging to Lasius and Camponatus. The aphids fed extensively on the aeciospores and later on the pycnial drops, they probably also obtain nutrients from the infected bark tissues. In one case the ants associated with the Cihara had built up earth to enclose the canker and aphids, effectively preventing aeciospore dispersal. The aphids were often found sucking the aecial area of the canker and had worked in un der the bark. Of the other orders of insects, some of the Collembola and Thysanoptera probably scavenged on the aecial zone of the canker, and ate spores. The Hymenoptera, excluding the Formicidae, were largely parasitic on Coleoptera, Lepidoptera or Diptera species inhabiting the canker and therefore were incidental. All others given in the list may also fall into this incidental or mycetoxenes category, and were certainly not mycetobionts and probably not mycetophiles. Rodents Rodent damage was widespread at most locations. The Richardson red squirrel, Tamiasciurus hudsonicus richardsoni (Bachman), British Columbia varying hare, Lepus americanus columbiensis Rhoads, and the dusky porcupine, Erethizon dorsat um nigrescens Allen, were all observed chewing on the aecia bearing bark at one time or another, and judging from teeth marks, chipmunks and mice were also responsible for removing some bark. Rodents generally removed all the bark down to the sapwood, but usually restricted their activity, except in the case of porcupines, to the infected bark. They showed a preference for the pycnial and aecial zones, but often ate the infected bark outside the pycnial zone. The gnawing rarely extended into non-infected healthy bark. In most - 103 -cases the rodents failed to eat the entire diseased area, enabling the rust to produce aecia in limited bark areas, or to continue to grow beyond the chewed area where pycnial and aecial sporulation would again take place in future years. In some cases the rodents hastened the death of the tree by completely girdling the stem or by removing the one remaining live infected strip of bark. Such trees would generally succumb to the attack of the rust in a year or two. In many cases the rodent-chewing helped to suppress or destroy the canker for around the edge of the chewed areas callus tissue developed. In several 20 to 30 year old stands a fair proportion of the trees showed old chewed areas where presumably a rust canker was .once active, but had been inactivated-by rodent chewing. In these cases the whole canker area was chewed out and presumably no infected bark remained from which the rust could continue to infect the tree and produce aecia. One of the most striking cases of this form of biological control encountered in the present study, was the case of the P. sylvestris plantation near Beaver Mines, already referred to in the section on damage. Rodent damage was characteristic of infected trees, as in in fected stands only rust infected trees showed evidence of rodent damage. The one exception to this was an area of porcupine damage where the por cupine had been less selective although still showing a preference for rust infected trees. Rodents appeared to only attack trees which had actively producing pycnial and aecial zones, and prefered the pycnial zones. Most of the damage probably occurred during the winter or early spring, but.some damage was recorded sporadically throughout the summer and early fall. In many cases, cankers showed evidence of annual rodent visits. - 10k Each year the rodents removed the succulent pycnial zone near the limits of the canker often leaving a thin strip of callus tissue adjacent to the previous year's chewed area. An example of such annual chewing is shown in Fig. 36, which had probably been visited for at least 10 conse cutive years. Often the rodents removed all the pycnial zone and some of the aecial zone, completely ringing the canker, but left untouched the center of the canker composed of dead bark tissues. At locations 2 and 3, where many of the cankers were protected by insect cages, it was not uncommon to discover, in the spring, that the rodents had chewed the newly infected bark beyond the top and bottom of the cage. There were also a few cases where the rodents damaged the insect cage and were able to chew the canker beneath. During the period 1966 to 1968, kkY cankers were observed for rodent damage, and of these 320 or 71.7$ received some damage in one of the years or in many cases each year. This percentage was slightly lower than would have occurred normally as some 20 of the cankers were protected by insect cages during two or three years of the period. Fresh chewing was only recorded on 6 of these "protected" cankers, far lower than the average rodent damage incidence at these locations. Table XIII Table XIII. The incidence of new rodent damage on Cronartium comandrae cankers, and the number of cankers observed at various lo cations in southwest Alberta during the years 1966 to 1968. Year Wo. of locations Total no. of No. of can- Per cent observed cankers kens with new new rodent observed rodent damage damage 1966 19 333 133 " 39-9 1967 21 385 166 43.1 1968 21 307 158 51.5 - 105 -shows the incidence of fresh rodent damage recorded each year and the number of cankers observed, excluding the cankers wholly or partly pro tected by insect cages. At two locations no rodent damage was recorded in any of the years. These, were areas of young open grown infected trees. Generally the greatest incidence of damage occurred in the older stands where damage by squirrels was predominant. At some locations the rodent population must remain high, but at others considerable fluctuation must occur. Table XIV shows the percentage incidence of fresh canker chewing at a number of selected' locations. At Mist Creek the stand was very open which may have accounted for the lower incidence, also less cankers were active in I968 which perhaps reduced their natural attractiveness to the rodents. The latter factor may also be responsible for the lower inci dence of rodent damage at Honeymoon and Ribbon Creeks in 1968, as much of the active portions of cankers had been removed in the two previous years. The last three locaions given in Table XIV were about half a mile apart, but Saskatchewan River Crossing 2 was an older stand and probably more suitable as permanent squirrel territory. In the older stands it was not uncommon to find squirrel nests in rust infected trees. At lo cation 1 a nest occurred at the level of a canker, but aecia were never observed during the 5 years, although fresh chewing occurred each year. The main result of the rodent-feeding was the-removal of the pycnial zone, and the future aecial zone, which brought about an enormous reduction in the volume of potential aeciospores for the following season. Often cankers were prevented from producing any aecia for a number of years through the rodent activity. A secondary factor was that the death of a diseased tree was often hastened, which further reduced the r Table XIV. The percentage incidence of fresh rodent chewing of Cronartium comandrae cankers at selected locations during the years 1966 to 1968. Location Vicary Honeymoon Baril Mist Ribbon Marmot Robb Saskatchewan River Crossing Creek Creek Creek Creek Creek Creek Road burn #1 #2 #3 1 H Year £ 1966 78 81 25 25 31 83 67 0 77 0 1967 . 55 90 26 13 68 74 32 17 80 9 1968 77 53 82 0 33 63 26' 83 100 1+6 - 107 -potential of a canker to produce aeciospores. Also, associated with the rodent chewing damage there was generally copious resin flow, which often contaminated the aecial zone preventing aeciospore dispersal. Dur ing the winter-spring season of 1966-67 at k locations, an average of 25$ (range for individual locations 10 to 3*+$) of the canker bark was removed from all cankers showing any rodent damage, and in many cases the bark not removed was no longer productive. Similar values for the 1967-68 winter-spring season, from cankers chewed at 17 locations, showed that 30$ (range for individual locations 7 to 73$) of the canker bark was removed, most being from the potential aecia-bearing zone of the canker. There were numerous cases where 70 to 100$ of the potential aecia-bearing bark on a canker was removed. Rodents tended to remove the infected bark nearest the center of the canker, the portion on which the first aecia would develop, thus even if some infected bark remained, aecial sporula tion was still delayed on these cankers. Often the few remaining aecia would not sporulate until two weeks after the main aeciospore sporula tion period. It was common to find rodent gnawing all round a branch stub which had served as the mode of entry of the rust into the main stem: Response of the tree The main response of the tree to the canker was the production of resin. The most noticeable resin production occurred in the pycnial zone, although considerable amounts were also produced in the tissues of the aecial zone. Characteristically the resin flowed down over the canker and the stem area below, causing most damage when originating in the upper pycnial zone. The resin often prevented further dispersal of aeciospores during the season, and formed a crust over the bark which may have - 108 -effectively prevented, aecia from pushing through the crust in future years to disperse aeciospores. It was not uncommon for the resin flow to take on an orange-yellow color where pycnial drops or aeciospores were incor porated into the resin. Most resin flow occurred towards the end of the main aeciospore sporulation period, continuing until at least late Septem ber. Resin was probably produced because the resin canal cells were rup tured by the aecial zone cracking and drying out, and through pressure produced on the resin canal cells by the increased amount of intercellular rust mycelium. Further resin flow occurred through the insect and rodent-damage, especially that of the latter, and helped substantially to pro duce a crust- of dried resin over much of the canker. Often resin flow extended for several feet below cankers occurring higher upon trees. Resin production often occurred annually on active cankers. TableXV re cords the incidence of fresh resinosis on cankers in the years I966 to 1968 at various locations. Most of those cankers not recording fresh resinosis were inactive cankers, although some cankers showing no active pycnial or aecial sporulation zones or the presence of Tuberculina or other fungi, still produced some surface resin.flow presumably in res ponse to the extension of the rust- mycelium around the periphery of the canker,.- Resin production was more extensive in association with Tuber culina than with the rust alone. In the three main years of observation only 37 cankers (7.9$) were observed that did not show evidence of some resinosis during their history. The majority of these occurred on very young trees (ll), or on trees that were observed for only one or two years before dying (20). Practically every tree with an inactive canker showed evidence of some period of resinosis, and often the vast majority of the - 109 -canker and area below was covered in dried resin. The resin often im pregnated large areas of the bark and outer sapwood. Table XV. The incidence- of fresh resinosis on Cronartium comandrae cankers on lodgepole pine and the number of cankers obser ved at a number of locations in southwest Alberta during the years 1966 to 1968. Year Wo. of lo cations -" observed No. of cankers observed No. of cankers with fresh resin Per cent of cankers with fresh resin 1966 20 356 237 . 66.6 1967 22 398 282 71.1 1968 22 320 212 66.2 Discussion Tuberculina maxima is reported to be a hyperparasite of rusts, including those of the Cronartium (Peridermium) group (Hedgcock 1935; Hubert 1935a; Mielke 1933; Tubeuf 1901, 19lk; Weir and Hubert 1917). However, present observations, and the study of Wicker on C_. ribicola in Idaho (Leaphart and Wicker 1968) indicate that T. maxima is not a true hyperparasite of the rust, but is parasitic on the rust canker area where the mycelium invades the bark tissue. This is in agreement with the early work of Lechmere (1914) who found that T. maxima did not attack or destroy the rust mycelium. All earlier reports are in agreement that the cankered bark dies after invasion by T. maxima and that it greatly reduces or-• entirely'inhibits the production of aeciospores where it occurs. T. maxima was described by Rostrup (1890) as attacking C. ribicola (Puccinia  klebahni) on Pinus strobus in Europe, and was first reported on C. comandrae - 110 -on P. ponderosa and P. contorta "by Weir and Hubert (1917) from Montana. Tuberculina was not found on this group of rusts in Alberta until 1964 (Powell and Morf 1965), but has been known since I926 on C.' ribicola in British Columbia (Mielke 1933). In 1964, Tuberculina was found at 11 locations on C_. comandrae on P. contorta and one on P. sylvestris between Robb and Beaver Mines, Alberta. Since that time Tuberculina has been found at a further 11 locations in Alberta, and one location in Kootenay National Park, British Columbia, on P. contorta, the latter•record being the first on C. comandrae in that Province. Several Tuberculina collec tions were also made on C. comandrae on P. banksiana from one location on the Mackenzie Highway, 110 miles south of Rae, Northwest Territories (Baranyay 1968), but Tuberculina has not been recorded on this pine host in Alberta, Saskatchewan or Manitoba. Powell and Morf (1965) reported the occurrence, of Tuberculina on Peridermium stalactiforme in Alberta, but not on C. ribicola or P. harknessii. However in 1965 Tuberculina was found on P. harknessii on P. contorta (CFB 6895) in Kootenay National Park, British Columbia, the first record on this rust in western Canada. In 1966 Tuberculina was collected on C_. comptoniae on P. contorta x P. banksiana hybrid (CFB 7736), 40 miles north of Nahanni Butte, Northwest Territories, which was recently reported by Baranyay (1968). Tuberculina was previously only known on this rust host from British Columbia (Mielke 1933). Its occurrence on C. comandrae and other Cronartium (Peridermium) spp. in western North American is probably more widespread than the col lections indicate. On C. ribicola and C. comptoniae, T. maxima is re ported to sporulate most abundantly on the pycnial zones of the canker (Hubert 1935a,b; Mielke 1933; Wicker, personal communication 1968), but on C_. comandrae it rarely sporulates on the pycnial zone but abundantly - Ill -on the aecial zone. It was found in the present study that only current season pycnial and aecial zones appear to be susceptible to T. maxima:, no fresh infection occurred on inactive cankers. Inoculation tests, re ported by Wicker and Kimmey (1967) similarly showed that nonfruiting cankers were not susceptible to infection, proving that infection entry was only gained through sporulating pycnia or aecia. Wicker and Wells (1968) reported that T. maxima was able to overwinter as spores, sporo-dochia, or in the mycelial stage within the pine cortex. Some spore viability was retained after 19 months storage at -31°C. T. maxima was always found in direct association with the rust, and from the evidence that small areas remained uninfected and were able to sporulate, it-appeared that T. maxima did not extend to or beyond the limit of the rust mycelium in the infected bark and therefore did not give complete biologi cal control of the rust. A number of people in Europe and America have used T. maxima as a control of C. ribicola with limited or no success (Hubert 1935a,b; Mielke 1933, Quick and Lamoureaux 1967, Tubeuf 1914, 1917, 1930). Mielke (1933) and Hubert (1935a,b) were of the opinion that its possibilities as a control agent were remote, but the more re cent work of Quick and Lamoureaux (1967), Wicker and Wells (1968) and Wicker (personal communication 1968) indicated greater potential. T. maxima was certainly able to cause marked reduction in aeciospore pro duction (Lechmere 191k; Spaulding 1929, Tubeuf 19l4; and the present study). Leaphart and Wicker (1968) reported a marked increase in the percentage of C. ribicola cankers parasitized by T. maxima. Ranges of incidence at the start were 37 to 80%, but by 1966 were 83 to 100%, al though some may not have been inoculated naturally. In the present study, - 112 -average annual incidence of 13 to 24$vas recorded for all locations. The high level of incidence, plus the fact that it was rarely present on the same canker for more than one or two years and that it usually inactivated the canker, indicates that Tuberculina was playing an important role in controlling C. comandrae and presumably other Cronartium rusts. A Cladosporium sp. has not been reported as an associated fungus of Cronartium stem, rusts, although Keener (1964) found Cladosporium spp. on the aecial sori of the cone rust Cronartium conigenum Hedge. & Hunt.. He also stated that Cladosporium aecidiicola Thum was the most frequently encountered fungus on all types of rust sori, and reported several cases on rusts in Worth America (Keener 1954, 1956, 1964). Cladosporium exoasci Lindau, C. exobasidii Jaap. (C. cladosporioides Thum), and C. hemileiae Steyaert were also reported to be hyperparasites on other parasitic fungi (de Vries 1952). According to Dr. Ellis (see above, page 83) Cladosporium tax. sp. 1. does not fit any of the descriptions for Cladosporium spp. associated with rusts. Cladosporium tax. sp. 1 was fairly widespread as in addition to the locations reported above, it was found on C. comandrae on P. banksiana (CFB 7447), 60 miles northeast of Fort Providence, W.W.T., and is quite common (9 CFB collections) on P. harknessii galls, on P. contorta in Alberta and the Wational Parks in British Columbia. Dr. Sutton (personal communication 1966) also reported it on C. comandrae and P. harknessii on P. banksiana in the Saskatchewan and Manitoba area. Some evidence was gained that Cladosporium tax. sp. 1 might be parasitic on C. comandrae aeciospores. de Vries (1952) doubted whether all of the described rust associated Cladosporium species were actually "hyperpara sites" as he was unable to detect connections between Cladosporium hyphae - 113 -and uredospores of Melampsora larici-epitea Kleb. on Salix sp., the sori of which were covered by C. cladosporioides, but the report of Steyaert (1930) for C_. hemileiae on Hemileia vastatrix Berk. & Br. is convincing. Keener (1964) was of the opinion that species such as Cladosporium ap parently lacked the capacity to destroy rust spores and probably only impeded spore dissemination. Smith (1905) reported that a species of Cladosporium on Asparagus rust in California, which he believed to be C. herbarum Link, was able to destroy rust spores and illustrated a Clados porium germinating from a dead rust spore. C. herbarum was observed less frequently in the present study from spore collections and may be playing a role in reducing the viability of aeciospores, along with C_. tax. sp. 1. C. herbarum was recently isolated from deteriorating telial galls of of Gymnosporangium juvenescens Kern (Eslyn i960). Cladosporium spp. are common among the primary fungal colonizers on plant material (Greene 1952; Leben .1965). A number of people have indicated that there is a succession of fungal colonizers on plant materials (Hudson 1962; Last 1955), and that Cladosporium often overruns the other colonizers, especially under higb humidity conditions. When insect cages were maintained on cankers in the present study, often causing a higher humidity level, Cladosporium, when present, rapidly.covered the whole canker. Of the other microfloral species isolated in the present study a few, or species of the same genus, have been reported as playing a pos sible role in reducing the effect of other plant pathogens. Wollenweber (1934) described Fusarium bactridioides which was found to parasitize the cone blister rust, Cronartium conigenum. Goodding was also able to demon strate the ability of F. bactridioides to attack C. ribicola, P. (C.) - 114 -harknessii and P. (C.) filamentosum from successful inoculations in Oregon and Idaho (Goodding, in Wollenweber 1934). Goodding (1932) also briefly discussed another Fusarium sp. which occurred on C. ribicola cankers in dependent of the aecia and was associated with a Hectria sp. There are a number of reports of the secondary fungi associated with C. ribicola can kers (Bingham 1942; Goodding 1932; Rhoads 1920; Snell 1929a,b; Spaulding 1922, 1929; Stouffer 1932) which are not parasitic on the rust. Rhoads (1920), Spaulding (1922) and Hubert (1935b), believed that the secondary fungi accelerated the girdling of the rust and hastened the death of the canker tissues. Posey and Gravatt (cited in Spaulding 1929) reported that 15$ of the rust infected trees recovered through the action of the secondary fungi in killing the infected pine bark, and through suppres sion of lower branches before the rust spread to the stem. Bingham (1942) and Bingham and Ehrlich (1943) believed that the secondary fungi reduce the aecial sporulation of the rust. Also that one of these, a Dasyscypha sp., once established killed the bark much more rapidly than C. ribicola (Bingham and Ehrlich 1943). Among the secondary fungi reported by others . on C. ribicola and identified in the present study were Lachnellula (Dasyscypha) spp. (Bingham 1942; Bingham and Ehrlich:.l943; Goodding 1932; Snell 1929a,b; Stillinger 1929), Tympanis spp. (Bingham 1942; Hubert 1931), Phoma sp. (Snell 1929a) and Phomopsis sp. (Goodding 1932). Keener (1964) considered that Verticillium along with Darluca and Tuberculina were species capable of destroying rust spores. Castellani and Graniti (1949) reported a Verticillium that was parasitic on Cronartium  asclepiadeum aeciospores causing them to become hyaline. Sukapure and Thirumalachar (1966) have reported a Cephalosporium parasitic on the uredia - 115 -of three rusts in India. Greene (1952) reports a Coniothyrium sp. on the telia of Puccinia anemones-virginianae Schw. in Wisconsin, and a Phoma sp. on Taphrina mirabilis (Atk.) Giesenhag. Myren (1964) isolated various imperfect fungi from all of the Cronartium fusiforme cankers col lected from three locations. Thirty per cent of the 132 cankers also yielded blue-stain fungi, and 13%-basidiomycetes. He thought that the associated insect galleries served as infection courts for various wood-inhabiting fungi. Microorganisms have also been reported associated with Hypoxylon pruinatum (Klotzsch) Cke. cankers. Bier and Rowat (1962a) re ported Pullularia sp. and Epicoccum nigrum to be common, and Wood and French (1965) added species of Alternaria, Chaetomium, Cytospora and bacteria. Several of these fungi inhibited growth of Hypoxylon in dual culture, and prevented canker formation (Bier and Rowat 1962a, b, 1963). Bier (1963) later reported that healthy bark-water suspensions containing fungi and bacteria controlled a leaf rust and a number of canker and decay diseases. Wood and French (1965) found that the bacteria were capable of reducing or preventing ascospore germination. Penicillium spp. are very common on a number of surfaces. P. brevi-compactum Dierckx. has been isolated from decaying fleshy fungi, P. funiculosum has been reported from lumber and is one of the world's com mon soil fungi, and P. cyclopium causes a bulb rot (Raper and Thorn 19^9). In associated studies various Penicillium spp. were isolated from Peri dermium harknessii galls, where they appeared to be far more common than on C. comandrae cankers. Paecilomyces farinosus (Dicks, ex Fr.) Brown & Smith is a common insect parasite (Brown & Smith 1957). Seimatosporium  discosioides (Ell. & Ev.) Shoemaker has been found on stems, and leaves of - n6 -Rosa (Shoemaker 1964). The Monocillium sp. in the present study may be an undescribed species as it differs from the type species of the genus, M. indicum Saksena, isolated from soil (Saksena 1955), in having much smaller conidia and different cultural features (B. C. Sutton, personal communication 1966). Barron (1961) recently added the species M. humicola, isolated from various forest soils in Ontario. The conidia of M. humicola are similar in size to those found in the present study. The bacterium Pseudomonas fluorescens (Fliigge) Migula was found, to inhibit a number of bacterial pathogens (Teliz-Ortiz and Burkholder i960). A number of other epiphytic bacteria are reported to reduce or control other diseases (Leben 1965). Pullularia pullulans and Sporobolo-myces sp. are widespread epiphytic yeasts or yeast-like fungi (Leben 1965). Voznyakovskaya (1963) discusses the widespread occurrence of 19 epiphytic yeasts on plant materials and reported that they sometimes constitute k-0 to 100$ of the total number of epiphytic microorganisms, and that the mycelial yeast-like forms often predominate on the leaves of trees, es pecially Pullularia pullulans. Of the non-mycelial forms, species of Rhodotorula (including R. aurantica), Sporobolomyces, and Cryptococcus were most common. Kais (1963) found Sporobolomyces on the telial columns of Cronartium f us if or me. Levine e_t al. (1936) described a Bacillus which inhibited cereal rust development in the field, and Poh et al. (195*0 described a Xanthomonas parasitic on uredia of cereal rusts. Similarly, Morgan.(1963) and French et al. (1964) discussed the effects of Pseudomonas  fluorescens and Bacillus spp. on cereal rust spore germination and infec tion. Further study may indicate that bacteria and 'yeasts' play a role . in reducing the production and viability of C. comandrae aeciospores. - 117 -Leben (1965) comments that the epiphytic bacteria, which seem to colonize actively growing tissue first, should be investigated intensively, as many have been shown to be inhibitory to test bacteria or fungi in dual culture. Insects have been reported from Cronartium stem rusts, but there is little reference to the damage caused by them. Peterson (i960) refers to insects eating aeciospores of P. harknessii, and Gravatt and Posey (1918), and Snell (1929a) noted insects feeding on aeciospores and pycnial drops of C. ribicola. A number of insects are reported to carry Cronartium aeciospores externally on their bodies which aids in spore dissemination (Gravatt and Marshall 1917; Gravatt and Posey 1918; Peterson i960; Snell 19193 1929a). Myren (1964) reported from surveys on slash and loblolly pine plantations, that insect attack occurred in kk to 100% of the C. fusiforme cankers on 6 to 15 year old pine, and that the percent age was higher in older plantations. The most common invaders were Dioryctria amatella (Hulst), a pitch moth, Eurytoma sciromatis Bugbee, a chalcid wasp, and Pissodes nemorensis Germar., the deodar weevil. A few other writers have mentioned insects associated with Cronartium cankers. These cankers obviously provide a microhabitat in which many microfaunal organisms spend a part or all of their lives. Snell (1919) listed Ik species of insects, as well as a crustacean and spider, which he collected on or near cankers of C. ribicola and which were later found to be carry ing aeciospores on their bodies. Eleven of these species were beetles, and included three genera represented in the present study, Melanophthalriia  gibbosa Hbst., Pissodes strobi Peck, and Dendroctonus valens Lec. He reported specimens with as many as 16,500 spores on their bodies, al though few carried more than 1,000. The only Lepidoptera collected was - 118 -a Liparidae, Porthetria dispar L., and Gravatt and Posey (1918) reported that the larvae of this moth rapidly destroyed the aecia and the under laying C. ribicola infected tissues. They noted that the larvae often destroyed a large percentage of the next year's sporulating zone and that aeciospore production was prematurely arrested in 25 'to 100% of the pustules. They observed an average of 18,100 spores on the surface of each small larva, and counted a further 26,000 spores in the alimentary tract. During a 13 hour period.20 larvae feeding on aeciospores produced k23 faecal pellets,•each of which contained an average of 8,l6o spores, but from tests they found that many spores were still viable. From this they estimated that each Porthetria larvae was capable of eating 318,6l6 spores a day. The Dioryctria spp., found in the present study, may well cause similar damage to that caused by the Porthetria larvae. Heinrich (1956), in his monograph on the moths of the subfamily Phycitinae, mentioned that D. zimmermani was the most economically important Dioryctria and that the larvae of this species bore into the cambium causing considerable damage. The Dioryctria as a group, may well be important.in reducing the quantity of Cronartium aeciospores produced. As already mentioned, Myren (1964) found D. amatella was one of the three most important insect invaders. Ebel (1965b) also found this species was common in C. fusiforme cankers, that D. abietella (Denis & Schiffermuller) occurred sporadically, and D. clarioralis (Walker) rarely. Ebel noted that the larvae of D. amatella, fed first among the spore masses upon the canker from January to March, and then entered the cankered tissue to complete development. Heikkenen (1964) found that 35% of the C. fusiforme cankers in plantations of - 119 -loblolly pine were infected by D. amatella, and a similar percentage oc curred in cankers on slash pine. D. amatella were also prevalent in first-year cones infected by cone rust, Cronartium strobilinum (Arth.) Hedge. & Hahn, on slash pine (Ebel 1965b; Merkel 1958). Ebel (1965b) reared several hymenopterous parasites from Dioryctria, including species of Agathis and Apanteles, which were both reared from cankers containing Dioryctria in the present study. In British Columbia, Ross and Evans (1957) reared D. abietella and D. zimmermani from Cronartium galls on Pinus contorta, and D. sp. nr. zimmermani from the area surrounding a patch of rodent-damaged bark. Anderson and French (1964) reported that D. zimmermani was a common inhabitant of C. comptoniae cankers, causing considerable resin flow. Specimens of D. zimmermani were also reared from P. harknessii and P. stalactiforme cankers in Alberta. The Las-preyresia spp. are often seed and cone feeders, and the Recurvaria are largely needle miners (Prentice et al. 1965), thus the occurrence of both in the cankers was probably due to the larvae seeking shelter for pupation. A Laspreyresia sp. gp. 2 was also obtained from a P. stalacti forme canker. This species does not fit the description of the others placed in Laspreyresia Htm., group 2 (MacKay 1959> and personal communi cation 1968). The Pulicalvaria sp. represents new host and geographical records, according to T. N. Freeman (personal communication 1968). Lar vae of the family Blastobasidae are often scavangers, although one species (Holcocera immaculella McD.) has been found boring in galls on pines (Prentice et al. I965). Nothing was known about the habitat of Epuraea obliquus until the present study, as it was only recently described (Hatch 1962) from two Oregon specimens. Parsons (1967) gives further detailed descriptions - 120 -based largely on my material. In addition to the eight reports of inci dence on C. comandrae cankers, it was reared or collected from one P. stalactiforme canker,.and k P. harknessii galls in Alberta, and from one P. harknessii gall from British Columbia and one from Quebec. Parsons (1967) reported seeing a specimen collected from near Oliver, British Columbia, and one reared from a C. quercuum gall from New Brunswick. There are many reports of other members of the genus collected from fungi. Hubert (1935b) mentions E. orata (probably E. ovata Horn) causing damage on C. ribicola cankers. BoVing and Rozen (1962) mention larvae of E. avara (Randall) from a fungus on Pinus taeda, which could have been C. fusiforme. Hubbard (1892) and Leech (1947) reported E. monogama Cr. found covered with fungus spores. Benick (1952), Hatch (1962) and others, men tioned several other Epuraea species found in fungi. Parsons (1967) re ported that E. avara, E. corticina Er., and E. terminalis Mann, were reared from oak wilt fungus mats, and Yount et al. (1955) reported the iso lation of viable Endoconidiophora fagacearum Bretz. conidia from faecal material of Epuraea spp. The aeciospores from E. obliquus faecal material were not viable. Of the other Coleoptera genera represented in the present study, both larvae and adults of the tribe Corticariini (Lathridiidae) feed ex tensively on fungi (Beaver 1966; Hatch 1962). Members of the genera Atheta are one of the most common in fungi (Benick 1952; Donisthorpe 1935; Graves i960), and were termed mycetocoles by Benick (1952). The Tenebri-onidae, except for a few species, have little connection with fungi (Graves i960). Benick (1952) stated that Aphodinus fimetarius (L.) was rarely found in fungi, but that certain members of the family Anobiidae, to which - 121 -Ernobius belongs, were strongly mycetophilous, almost mycetobionts, living and breeding on fungi. The Curculionidae and the Scolytidae were probably secondary species, attracted to the canker zone of the rust as they do not normally infest healthy trees. Other species of the same genera are primary species. In the description of the species Cylindrocopturus  deleoni Buchanan, Buchanan (Vjko) mentioned that some specimens from Idaho were collected from a "fungus gall infested with coleopterous larvae on yellow pine". This could have been a gall of P. harknessii which is common on Pinus ponderosa. Wood (1964) reported larvae, pupae or teneral adults of Pissodes schwarzi and P. curriei Hopk. in the root collars of dead or dying Pinus monticola saplings in British Columbia, which had been infected with C. ribicola or root rot. Wood's distribution map for P. schwarzi showed that it ranged throughout'..the interior from the United States border and north into the Yukon Territory. As mentioned above, P. nemorensis was a common invader, of C. fusiforme cankers (Myren 1964). Snell (1919) found P. strobi associated with C. ribicola. Snell also mentioned Dendroctonus valens Lec. at the base of an infected C. ribicola tree. The specimen of D. murrayanae Hopkins of the present study, was found burrowing into a basal canker. Rhoads (1920) reported that the beetle Pityogenes hopkinsi Swaine attacked trees weakened by C. ribicola, and accelerated"their death. Many of these were thrifty young trees which would not have died for at least one to three years from the activity of the rust. A similar relationship was noted in the pre sent study, trees from which C_. deleoni and P. schwarzi were collected were often dying or dead by the following year. Kais (1963) reported that larvae of the. diptera species Mycophila  fungicola Felt fed on the spores of 0. fusiforme, and Snell (1919) found - 122 -a species Rhagio (Leptis) mystaceus (Macquart), collected from the base of an infected C_. ribicola tree, with 500 aeciospores on its body. . At present no other type of habitat is known for Paracacoxenus guttatus and it seems that this species may be a true mycetobiont. Adult flies were reared from cankers, and adults were collected when visitng the aecial and especially the pycnial zone. The larvae of this diptera and other larvae on the cankers may play a role in the exchange of (+) and (-) pycniospores between the haploid pustules of the rust (Craigie 1931)5 as they move from one pycnial drop to another, thus helping the rust to com plete its life cycle. Eradysia were reported in moist places wherever:; fungi grew (Stone et al. I965). A few of the Cecidomyiidae cause primary damage to fungi (Stone et al. 1965), and Pielou (1966) reported many speimens collected from Polyporus betulinus (Bulliard) Fries, which in cluded the most abundant species found. Many Phoridae have been reared from fleshy and woody fungi, and the Piophilidae are scavengers in fungi (Stone et al. 1965). Members of the genus Medetera are common predators on immature stages of bark beetles (Beaver 1966). Jackson and Parker (1958), as well as Myren (196U) have noted the occurrence of chalcid wasp larvae in C_. fusiforme cankers. They re ported that 2 to 30 larvae could be found in the canker and tentatively assigned the insect to the genus Bephratoides. They were of the opinion that the canker afforded an ideal habitat for breeding. Most of the hymenoptera reared in the present study were probably parasitic on other species and were attracted by the numerous larvae that developed in the cankers. A few may be phytophagous like Eurytoma sciromatis on C. fusiforme.(Krombein and Burks 1967, Myren 1964). Coelichneumon - 123 -brunneri Rohw. has been reported as a parasite of Dioryctria aurantice11a Grt. (Krombein and Burks 1967), and was reared from a collection of D. zimmermani in the present study. Agathis binominata M.&W., Apanteles, Phaeogenes, Glypta, Microchelonus, and Copidosoma are probably parasites of Lepidoptera larvae (Krombein and Burks 1967; Muesbeck et_ al. 1951). Phygadeunon are usually parasites of muscoid Diptera (Muesbeck et al. 1951) and were reared from a canker•that produced P^iacacoxenus guttatus. Triaspis are parasitic on Coleoptera larvae (Muesbeck et al. 195l), and Dolichomitus terebrans nublipennis has been reported from Pissodes (Krom bein and Burks 1967). Pissodes and Cylindrocopturus were both reared from the same collections in this study. Many collembolan species are herbivorous but have often been found associated with fungi, and it is not unreasonable to assume that the cortical species live on fungal hyphae and spores (Graves i960). Pielou and Matthewman (1966) recorded several Collembola species from bracket fungi in Quebec. The Correntia were mostly from rearings and form an important group of mycetophages, as Graves (i960) found in fungal conks.. Gnophothrips fuscus (Morgan) has been reported damaging pine species in a number of areas (Ebel 1961,- 1965a; Lindquist and Harn-den 1957; O'Neill I965). G. fuscus were also reared from a P. stalacti-forme canker in a companion study of insects associated with other Cronartium cankers. Stannard (1957) pointed out that thrips of the fam ily Phloeothripidae, to which G. fuscus belongs, feed extensively on spores of fungi, thus cankers of Cronartium are a likely habitat for this species. The Laelaptidae mites are usually parasites on vertebrates and - 12k -invertebrates, and the Bdellidae are predaceous on mites and small in sects such as Collembola (Graves i960). The Eupopidae and Tydeidae ap peared to be fungus feeders. The Anystidae are predaceous on mites and small insects, and the Erythraeidae are parasites of insects, although the adults are free-living predators (Baker and Wharton 1952). The Acaridae live on all kinds;-of organic material, some fairly exclusively on fungi (Baker and Wharton 1952). The oribatid mites are largely found in soil debris where they feed upon organic matter, including fungi (Graves i960), although little is known of the individual species habitats. A number of different species of rodents are believed to eat the infected living bark from Cronartium rust cankers. They-have not „ all been observed, as many rodents are strictly nocturnal, and most ob servations are made during the day. Squirrels are regarded as the most important infected bark removers, and have been reported by a large num ber of people (Hedgcock and Hunt 1920; Hubert 1935b, Mielke 1935, 1956; Peterson i960; Rhoads 1920; Snell 1929a; Spaulding 1918, 1922, 1929; Stillinger l^kk). Mielke (1935) listed many examples of squirrel damage that occurred on C. ribicola cankers, and were largely reported in "The Blister Rust News" from 1928 to 1933- Damage by porcupines was mentioned by Pennington (cited in Spaulding 1922), Hedgcock and Hunt (l<920\ Hubert (1935b) and Mielke (1935, 1957), and several of these authors referred to damage resulting from mice and rabbits. Mielke (1935) mentioned chip munks and the possibility of a pika. Rodent feeding on cankers generally occurred during winter and early spring and was usually restricted to the living infected tissues of the canker. There are several reports of the incidence of rodent damage on C. ribicola cankers from both eastern and - 125 -western North America. Gravatt and Posey (cited in Mielke 1935) reported that 17% of the cankers had been partially eaten off at Kittery Point, Maine, and 75% of these had kO to 100% of the bark removed. Pennington (cited in Spaulding 1922) estimated that the production of aeciospores in the Adirondacks was reduced about 15% by the eating of infected bark. Snell (1929a) in New York State, found that 1+1% of 11,000 cankers had been gnawed by rodents and that the entire crop of pycnia was consumed in some locations. Perry (cited in Mielke 1935) estimated that the fruit ing area of cankers was reduced by 95% at Pembroke, Massachusetts in 1928. At Cheekye, British Columbia, Lachmund (cited in Mielke 1935) found 1+5.2% of the cankers were gnawed to some degree in 192k, and that 25% of the aecia-bearing bark was removed. He also noted that older cankers, which had sporulated for more than one year, had a greater incidence of chewing. In a light infected stand near Revelstoke the number of cankers gnawed in creased from 13.5% in 1931, to 25% in 1933- Mielke (1935) estimated that rodents removed 10 to 35% of the aecia-bearing bark in the older infec tion areas of the west. Stillinger (l9kk) reported that the Richardson red squirrel chewed 28% of 1575 cankers and 38% of 10,360 C. ribicola can kers observed at two locations in Idaho. Mielke (1956) found only a few small cankers of P. stalactiforme, in many thousands observed, that had not had zones of infected bark removed annually by rodents. He reported a margin or ridge of dry and hardened bark on the P. stalactiforme can kers which was not usually fed upon, as observed on many C. comandrae can kers. Anderson et al. (1967) reported incidences where 37 to 50% of the P. stalactiforme cankers had rodent feeding, and in many instances the feeding appeared to have prevented further development of the cankers. - 126 -Mielke (1957) reported rodent damage on C. comandrae cankers on lodgepole pine, and some reports of severe porcupine damage associated with this rust. He further observed that the various rodents causing the damage tended to congregate more or less within the diseased stands. Childs (1968) referred to oval gnawed areas centered at the branch where C. comandrae infection started on Pinus ponderosa. Krebill (1965), in a study in 12 National Forests in the Rocky Mountain States area, found over 90$ °f the sampled C. comandrae cankers had been scarred by ro dent chewing. Cordell et al. (1967) found that over 75$ of the fruiting _C. comandrae cankers in a P. taeda plantation had been chewed, and in many cases the cankers were completely gnawed off. Peterson (i960) re ported the gnawing of P. harknessii galls, and their removal from small branches by squirrels, chipmunks and porcupines. In the study area, and especially near locations 2 and 5, large numbers of P. stalactiforme can kers were extensively gnawed by rodents each year, and lighter damage was observed on P. harknessii cankers. A number of references mention resin exuding abundantly from cankers of C. comandrae (Mielke 1957, I96I; Mielke et al. 1968; Peterson and Krebill 1967), especially on larger stem infections. Some refer to the additional resin flow associated with rodent damage on the canker, but there is no quantitative data on the incidence. Krebill (1968a) found some small patches of resinosis in the bark and wood infected by C_. com andrae but did not believe this was sufficient to impair the conducting ability of the xylem. Mielke (1961) noted that resinosis was less marked on C. comandrae cankers on Pinus ponderosa than on P. contorta. Mielke (I956) mentioned that impregnation of the wood by resin usually occurred in P. stalactiforme cankers. Spaulding (1929) reported that abundant - 127 -resin production is a good symptom of the disease caused by C_. ribicola. Colley (1918) described the manner in which resin canals were broken with the consequent exudation of resin in large quantities. He indi cated that the resin impregnated the whole cortex and phloem of the cracked C. ribicola canker area, resulting in stoppage of the conducting elements in the phloem hastening the girdling of the tree. Hirt (1964) suggested that the increase in number and size of the epithelial cells adjacent to vegelative hyphae in infected bark, which tend to fill the resin canals, may partially account for the abundance of resin associated with C_. ribi cola cankers. The impregnation of the bark and wood around the canker prevents desiccation of the sapwood, and hinders the growth of the fungus, so that complete girdling of the tree is greatly delayed. -128 -AECIOSPORE DISPERSAL This section of the study aims: (l) to establish the daily and seasonal periodicity and concentration of aeciospore release, from indi vidual cankers, and to relate this to various meteorological parameters; (2) to establish the natural pattern and gradient of aeciospore dispersal and deposition around individual cankers, and the distance of dispersal in the forest; (3) to establish from experimental tests the aeciospore concentration pattern and distance of dispersal in the open from two source release heights under different wind velocities and to relate this to diffusion theory; and (k) to establish the rate of fall of aeciospores in still air, a factor which is important in understanding dispersal dis tance and spore deposition. METEOROLOGICAL FACTORS AFFECTING DISPERSAL Methods and Materials Two areas were instrumented with a number of meteorological instruments to record the various meteorological parameters close to the ground within the forest stands. A variety of spore collectors were set up in the same areas to record the numbers, periods and distance of spore dispersal. Experimental Sites In I96U an area was instrumented at 5,000 feet in a dense 25 year old lodgepole pine stand on the north slope of The Wedge and on the south side of Evans-Thomas Creek, 50°53' N, U5°09' W, (location 2, Fig. - 129 -38). This site was relatively level as it formed part of an old river terrace, but on the west and north the terrace sloped steeply towards the Kananaskis River, and Evans-Thomas Creek. The site was largely abandoned in I965 for this aspect of the study, except for the mainten ance of a few weekly recording meteorological instruments and weekly ob servations on the development of the rust cankers, which was continued through to I968. Operation of spore collectors on this site in 1964 in volved the use of six and 12.volt batteries, which for satisfactory op eration were changed daily, and in the case of the Hirst spore trap changed twice daily for continuous operation. In 1964 a few instruments were run near the shore of Barrier Lake, 51°02' W, 115°Q2' W, (location 1, Fig. 38), in an area of minor un dulating relief at about 4,550 feet, with fairly open grown lodgepole pine of uneven age. In the period 1965 to 1968 this site became the main experi mental area and was fully instrumented (Figs. 46 to 53). Many of the spore collectors, and the wind and some dew recording instruments, were opera ted from a temporary 110-volt power line from the Kananaskis•Forest Ex periment Station. A sketch showing the area and the positions of the instruments and infected trees used in the study is shown in Fig. 39- In addition to the trees shown in Fig. 39? "two other trees (#2710 and 271l) were observed near location 1. These were on a knoll, 1600 feet to the east of tree #2720, where a hygrothermograph and rain-gauge were maintained in all the summers, and a spore collector was operated next to #2710 in 1964, I965 and I967. The number of individual instruments and collectors used varied from year to year depending on the number of cankers used to observe daily Fig. 46. View of site no. 2 at study location 1, showing a 24-hour impaction spore collector by Cronartium comandrae canker no. 2721 on a small Pinus contorta, and instru ments for recording the weather. Instruments are, from left to right, black porous disc atmometer, instrument shelter containing hygrothermograph, Wallin-Polhemus dew duration recorder on ground, mast with anemometer cups for wind speed recorder and spore collector. (Rain gauge is out of the picture) Fig. 47. Standard Stevenson screen in an opening, containing hy grothermograph and thermometers used as reference weather . station at study location 1, with a bi-metal actinograph for recording incoming radiation seen at the back. A recording rain gauge and a wind direction recorder (out of the picture) were maintained at this open site. Fig. 48. A Hirst spore trap with sampling orifice one foot above ground, close to a Comandra umbellata plot. Fig. 49. Seven-day pollen sampler of the Sarvas type, used to collect aeciospores at set distances from sporulating Cronartium comandrae cankers. Spores, pass through the sampler orifice and are deposited on a vaseline-coated sampling band placed around a clock-driven drum housed within the intake cylinder. - 131 -spore dispersal, and the requirements for satisfaction of the various aspects of the study. Spore collectors Five types of spore collectors were used over a period of five seasons. The efficiency and usefulness of each collector type varied considerably. One "Casella" model of the Hirst spore trap (Hirst 1952) and five weekly pollen collectors of a design similar to Sarvas (1952) proved un satisfactory for obtaining detailed information on the diurnal dispersal of aeciospores. This type of wind directional recorder cannot be placed close to a sporulating.rust canker, consequently they were only used to gain information on distance of spore dispersal. In 1964 the Hirst spore trap was operated in a small clearing at location 2 approximately k2 feet from the nearest sporulating cankers. From 1965 to 1967 the trap was operated at location 1 throughout the sporulation period near a patch of Comandra plants and 3**- feet from the nearest infected pine tree (Fig. k8). The trap sampled air at the rate of 10 liters/minute through the operation of a small separate electric air suction pump. The orifice was one foot above ground and was contin uously directed into the wind. The microscope slides were prepared for exposure according to the method described by Hirst (1953), and were changed daily between 0845 and 0900 M.S.T. The slide was moved past an orifice, 2 mm wide, at a rate of 2 mm per hour, thus hourly deposits of spores throughout a 2k hour period, were obtained. The slides were scanned under a stereomicroscope with a 50x power, and the number of spores de posited on each 2 mm interval or band were counted. - 132 -The clock-driven pollen collectors (Fig. 1+9) were identical to those used by Ebell and Schmidt (1964, p. 4-5) in their study of pollen dispersal on Vancouver Island. The spores were deposited on a vaseline-coated celluloid sampling band which was placed round the drum of an ;8-day clock enclosed in a removable cylinder which allowed for the passage of air through the orifice around the drum and out of a hole in the rear of the cylinder. The sampler is wind directional and in the study was run one foot above ground. The celluloid band had vertical lines printed on it at 2 hour intervals. Approximately 2 hours of the band were ex posed at one time. These pollen samplers were operated at various intervals during the sporulation period.of I965, 1966 and I967, and at various points removed from sporulating cankers, in an effort to obtain distance of spore dispersal data. Impaction spore collectors of a design similar to Panzer e_t al. (1957) were used in 1964. They could be placed close to a sporulating canker, but were not wind directional. These collectors had some disad vantages; a low spore collection efficiency, a non-continuous record of collection for a 24 hour period, and were time consuming in changing 24 slides daily, espeically during inclement weather when the chances of contamination were increased. Some features of the basic design of the Panzer e_t al. model were modified in the winters of 1964 and 1965, and a model was developed which proved entirely satisfactory for.the daily collection of dispersed spores (Powell and Morf 1967? and Figs. 50 and 5l). The fan of the impaction spore collector was operated at a control led 12 volts in 1965, 1966 and 1967, through a transformer rectifier unit powered by a 110 volt power line. To improve on the collection and Fig. 50. A 24-hour impaction spore collector, with the sealed lid removed to show the acrylic plastic disc holding 24 mi croscope slides upon which spores are deposited, and the fan for drawing air through the box at a controlled rate Fig. 51- A 24-hour impaction spore collector in operation, with its sampling orifice placed close to the sporulating surface of a Cronartium comandrae canker. Fig. 52. Central instrument power and recording box, with Thorn-thwaite four-unit wind speed register recorder (left), wind direction recorder, and digital printout recorder with polaroid camera for wind speed recorder system (right). . Behind are the transformers and rectifiers for reducing the 110 volt power supply and for controlling an output of 12 volts for operation of wind instruments, dew instruments and 24-hour impaction spore collectors. A standby 12 volt battery is also present. Fig. 53. Set of anemometer cups of.the Thornthwaite wind speed register recorder system operated at canker height near a sporulating Cronartium comandrae canker. - 133 -- 13h -retention efficiencies of the rPanzer et al. collector our model was designed to increase the volume of air sampled to 13 tl liter per minute (O.78 m /hr.) through the use of a larger diameter sampling orifice (0.8 cm diameter, compared to a O.k cm diameter). Also the gaps between the 2k microscope slides of the Panzer et al. model were eliminated by ar ranging the slides side by side, held together along the bottom by a strip of masking tape with an overlap of tape at each end for attaching to a plastic holder. The slides were prepared in the laboratory and' attached to spare holders which facilitated the daily change at approxi mately 0900 hours; holders were changed in the same order each day. The prepared slides and holders were carried to the field in dust proof boxes which further minimized the chance of contamination. In 196I+ two col lectors of this type were operated at location 1 and four at location 2. From 1965 to 1967, three to six collectors.were used each year at loca tion 1 for the duration of aeciospore sporulation. The same four cankers at location 1 were used in three of the four years, although they were not the same years. The same trees could not be used each year because the tree had died or because of lack of sporulation in some seasons. Counts were made of the total number of spores deposited on 5.^ cm of each slide, and no correction was applied for the efficiency of the trap which depended on wind speed. The efficiency of the trap was un-• known, although it was apparently good. It was impractical to scan all slides daily but during initial and cessation spore periods slides were examined soon after- exposure. After long storage spores tended to lose color and counts at these times were probably underestimated, although the overall picture was probably not changed. Spore trapping usually - 135 -started just before rupture of the aecia which was established through frequent visits to the infected tree, and continued for two to three months until no spores were found on slides exposed on three consecutive days and when macroscopic examination of the canker revealed no active sporulating aecia. Occasionally spore trapping was stopped because of mechanical failure, or for placement of the trap on a different sporula ting canker. Another simple wind directional plastic collector was designed in 1966 for the dispersal experiments involving the release of spores from a fixed point (see page 175), and were also used in 1966, 1967 and I968 to catch spores at set distances (5 to 100 feet) from sporulating cankers. These collectors were mounted at one foot above ground, or at 1, 5 and 10 feet above ground on a supporting stand. Slides were ex posed for 2k hour periods being changed between 0800 and 0830 hours. One or two (in I967 and 1968) rotorod samplers of a modified type, similar to that described by Perkins (1957) were used to catch spores on a daily basis at set distances from sporulating cankers and during the dispersal experiments. Rotorod samplers collected spores by impaction on a pair of small clear plastic collector rods which were held in a specially designed U-shaped holder and rotated at a constant speed of 2k00 r.p.m. by means of a'miniature battery-operated electric motor. The lightly vaseline coated surfaces of the collector rods sam pled at an average rate of 60 liters of air per.iminute. This impaction collector had a high retention and efficiency level, in excess of 90% (Metronics Associates, 1966). The rotorod samplers were operated on a small tripod at approximately 3 feet above the ground. In most cases the collector rods were exposed for 2k hours being changed between 0800 - 136 -and 0830 hours. Meteorological instruments Temperature and relative humidity were recorded at both sites by weekly recording Fuess hygrothermographs operated in standard Steven son screens or modified instrument shelters (Figs, 46 and 47). In both areas one was operated at standard height (4-1/2 feet) and up to five others were operated close to and at the height of the cankers (between 6 inches and 1 foot above ground). The instruments at location 1 were checked daily, and standard certified maximum and minimum thermometers . .r were maintained at some stations as a check against the performance of the temperature sensor of the hygrothermograph. The relative humidity sen sor of the recorder was checked at regular intervals by readings from a sling psychrometer. All instruments, including clocks, were calibrated at the beginning and end of the field season over a range of temperature and moisture conditions as a further check on the accuracy of their op eration. Precipitation intensity and duration were recorded by two Casella natural siphon weekly recording rain-gauges, one at each location. A series of plastic wedge-shaped and MSC (Meteorological Service of Canada) standard non-recording rain-gauges were also placed throughout the study area at location 1 and attended on a daily basis. Huff (1955) showed that the plastic wedge-shaped gauge, with a 2.5 by 2.3 inch rectangular orifice, was satisfactory in its performance when compared with an 8 inch U.S. Weather Bureau standard gauge. Four types of dew duration or leaf-wetness recorders were used at location 1 for varying periods, none of which were 100$ satisfactory, - 137 -but yielded sufficient dew and leaf wetness information for short period analysis. One Wallin-Polhemus dew duration recorder (Wallin and Polhemus 195*1-) employing a lambs-gut or cellophane strip sensor was operated on the ground at one canker site (#2721 - Fig. k6) over the period 1965 to I967. Two types of ground glass dew duration recorders were used in 1965, one, a 7-day recorder of a type similar to that described by Theis and Calpouzos (1957), and the other, a daily recorder modified from Taylor (1956) by Dr. A. K. Parker, Department of Fisheries and Forestry, Victoria, British Columbia, and used in his studies of the Rhabdocline needle cast on Douglas fir. An electrical grid wetness recorder system with four sensors was developed, employing two parallel strips of metal set in a plastic disc, and attached to a Rustrak four-channel on-off recorder. The sensors were placed close to sporulating cankers and up to 500 feet from the recorder. When moisture was deposited on the disc the circuit between the two metal strips was completed and the recorder activated to the "on" position. The period of contact denoted the length of the wet period. A grass minimum thermometer was operated in the open at 6 inches above the ground which gave additional information on the daily occurrence of dew and minimum temperatures at this height. Five atmometer assemblies were run in 1965, 1966 and 1967 at location';.l, to obtain estimates of the daily amounts of evaporation. Two were the "shielded plastic mount" type, which employed the black Bellani plate as an evaporating surface, and three were the "black porous disc" type, an assembly which functions on the principle of the Piche atmometer. Carder (i960) compared both, finding that the "black porous disc" had several advantages over the other. Two "black porous disc" atmometers were operated in the open, and the others within the - 138 -stand, near infected trees. The amount .of water evaporated by each at-memeter (in c.c), was recorded daily between 0815 and 08^5 hours. A Casella model of the improved Robitzsch bi-metal actinograph (Fig. 1+7) was run daily in 19&5, 1966 and 1967 in an open clearing at location 1, and in 196k at location 2, to give estimates of the global (sun plus sky) radiation, measured in cal/cm^/min, and information on the periods of daytime cloud cover over the study area. Wind speeds were recorded by a Thornthwaite Four-unit Wind Register System with a Digital Printout Recorder (Fig. 52) in 1965 and I966 and intermittently in I967 at location 1. The four anemometers (Fig. 53) were placed close to and at the height of four sporulating cankers where there were spore collectors, to obtain information on wind speed at the height of spore release. Wind direction was recorded at one point central to the study area at a height of k feet by a Thornthwaite Wind Direction Recording System, (Fig. 52) to obtain the pattern of air move ment over the study area. This wind direction recorder was situated on the southwest side of a slight knoll exposed to the prevailing southwest winds, and was 60 feet from the nearest anemometer. Wind direction at the canker height was probably similar to that at the h foot recording height. Supplementary wind speed and direction information during I965 to 1968 was obtained from a Meteorological Branch, Department of Trans port, recorder operated at k8 feet at the Kananaskis Station and at a distance of approximately 330 feet from the anemometer near canker #2721 on the study area. Time of day reference was made throughout to Mountain Standard Time (M.S.T.). - 139 -Results Diurnal periodicity of meteorological factors and microclimate of study  location 1 Most meteorological factors show marked diurnal rhythms, namely light, air temperature and humidity, and wind velocity (cf. Figs. 58-60, Table XVI). Of these, light is the most constant but its Inten sity varies through the presence of haze or cloud. The diurnal tem perature pattern rarely departs from a daytime maximum and a nighttime minimum. Similarly, relative humidity is low during the day and high at night, but this rhythm is often upset when high humidities persist dur ing the daytime, or low humidities during the night (MacHattie 1966). The diurnal periodicity of wind is often less obvious, but when averaged over relatively long.periods there is greater wind velocity during the daytime than during the nighttime. Table XVI summarizes the diurnal pattern of hourly temperature, relative humidity and wind speed for two of the stations situated near sporulating cankers, 0.6 to 1 ft above the ground, for the period May 17 to July 11, 1966. The temperature and re lative humidity data are compared with records from the standard Bay station situated in an opening at 4.5 ft above ground and 100 ft from #2712. The wind speed records are compared with those taken at 48 ft at the Kananaskis station. The two stations (#2712 and 2721) represent the extremes of the conditions recorded 0.6 to 1 ft above the ground at location 1. Meteorological conditions at station #2713 were very simi lar to conditions recorded at #2712, although those at the latter were slightly more extreme as this station was situated on the edge of an open area. Conditions at #2724 were intermediate between #2713 and #2721, Table XVI. The hourly mean temperature, relative humidity and wind speed at two canker stations, compared with'records from the Bay station for temperature and humidity, and from Kananaskis for wind speed, for the period May 17 to July 11, 1966. Bay #2712 #2721 Ht. of sensor Hour of day Daily (ft) 1 2 3 U 5 6 7 8 9 10 11 12 13 ll* 15 16 17 18 19 20 21 22 23 2U aver TEMPERATURE °C 10.5 9-9 9-7 9.1* 9-2 9-7 11.I* 13.5 15.3 16.8 17. •* 18.U 18.8 19.2 19.1 18.8 18.3 17.6 16.3 15.2 13.8 12.8 11.9 11.3 1U.3 0.6 9.6 9-1 9.0 8.6 8.3 8.2 9-8 12.1 H*.3 l"l.9 16.1 17.0 17.9 18.3 19.1 18.7 17.7 17.2 16.5 15.1 13.** 12.3 11.3 10.1* 13.5 0.6 8.5 8.2 7.8 7.1+ 7.3 7.6 9-7 11.5 ll*.2 16.5 17.5 18.2 18.6 18.5 18.8 18.3 17.6 17.0 15.7 13.9 12.5 11.1 10.3 9-2 13.2 RELATIVE HUMIDITY %, Bay 1*.5 77 80 81 82 82 78 70 63 57 51* 51 1*8 1*8 1*7 1*7 1.8 50 51 55 58 61* 69 73 75 63 #2712 1.0 77 80 82 83 85 86 79 68 58 52 1*8 1*5 1*2 1*1 1*0 1*1 U3 1*7 51 53 61 61* 70 71* 61 #2721 1.0 81 82 83 8U 85 86 81 70 59 5"* 1*9 1*7 1*6 1*1* 1*5 1*5 1*7 1*8 52 59 65 71 75 78 61* WIHD SPEED m.p.h. -Kananaskis 1*8.0 3.8 3.9 U.O 3.7 3.5 3.3 3.1* 3.6 5.3 6.1* 7.2 8.0 7.9 8.3 8.7 8.9 8.1* 8.0 7.8 6.8 6.1 5.1 "*.5 3.9 5.8 #2712 1.0 1.1* 1.1* 1.1* 1.1* 1.3 1.3 1.5 1.7 2.0 2.1* 2.5 2.6 2.6 2.7 2.8 2.7 2.6 2.3 2.1 1.9 1.7 1.6 l.l* l.l* 1.9 #2721 1.0 1.2 1.2 1.2 1.1 1.1 l.l 1-3 l.l* 1.7 1.8 2.0 2.2 2.3 2.3 2.2 2.1 2.2 1.9 1.8 1.6 1.5 l.l* 1.2 1.1 1.6 H -F-O - l4<l -the latter being representative of a less open stand area. Temperatures recorded at the canker stations were lower than those at 4.5 ft, especially at night, when a temperature inversion developed through * cooling at the ground on clear nights. Relative humidities were very similar at all stations, with only a slight tendency at the Bay station for humidities to average a little lower at night, and station #2712 to have lower mid-afternoon humidities. Minimum wind speeds occurred around sunrise at all stations, and maximum velocities in mid-afterhoon. Maximum wind speeds occurred at Kananaskis an hour or so after the near ground stations, and wind speeds at all hours averaged three to four times greater at Kananas kis than at #2721. Winds at #2712 and the other canker stations were a little higher than at #2721, a less exposed site. The wind direction at Kananaskis was predominantly from the east (45%) or southeast (l4%) dur ing the hours 2400 to 0700, and from the southwest (54%) or west (22%) during the hours 0900 to 2200, during the same period of observation. Even during the hours 2400 to 0700, 34% of the winds were still from the southwest (27%) and west (7%). At the wind direction recorder near the Bay station, the same dominant southwest pattern of winds existed during the daytime, but at night the winds were predominantly from the' southeast rather than the east. The wind shift in the morning from the southeast to the southwest tended to occur between 0700 and 0800 hours, but the time of the reverse shift in the evening was more variable. A very simi lar diurnal pattern of meteorological conditions existed during the main sporulation period in other years, in this study area. Diurnal spore periodicity One of the most striking results from this study was the marked - 1-2: -diurnal periodicity in the dispersal of aeciospores. The diurnal dis tribution of aeciospores dispersed into the air, expressed as percentages of the peak hourly geometric mean concentrations for three cankers in the four seasons are shown in Fig. ^k. The observation period - number of days - for the mean diurnal periodicity curves for each of the three cankers in each year varies, which accounts for some of the intra-canker variation in a season. Maximum concentrations were dispersed during the period 0800 to 1800 hours. Low concentrations of spores generally occurred between 2200 and 0500 hours when numbers sometimes dropped to zero. The rise from low to maximum concentrations was normally rapid, but the fsill was relatively slow. In 1966 peak concentrations occurred around 1500 hours, in 1967 between 0900 and 1100 hours, and in 1964 and I965 there was greater spread in the period of maximum between the three cankers. There was some evidence that individual cankers retained a similar maxi mum concentration hour between seasons. The maximum concentration for canker no. 2721 occurred at 1500 hours during the three years of obser vation, while the maximum of canker no. 2710 occurred between 0800 and 1100 hours in the three years. This variation may have been related to the aspect and exposure of the 'canker. No. 2721 occurred on a south facing slope and no. 2710 on a steep northeast facing slope which would receive the drying effect of the sun considerably earlier than no. 2721. However, cankers nos. 2712 and 2713 which were only 4 5 feet apart had different maximum hours in the two comparable years of record. There was some evidence, from most of the cankers, of a double or secondary peak, either around 1000 or 1500 hours, there being a considerable re duction in the spore concentration around 1200 to 1300 hours. In an effort to understand some of the other peaks'in the Fig. 54. Mean diurnal aeciospore periodicity curves for three cankers over varying periods of trapping in the years I96U to 1967, expressed as a percentage of the peak geometric mean hourly concentration. - 143 -TIME (HOURS) - M -curves the data for 1966 and I967 were analysed on the basis of hourly frequency of maximum concentration for each of the three cankers, and separated into days with rain, and days with no rain, or with a light shower which did not effect spore concentration (Fig. 55). From this analysis it is apparent that on days without rain the maximum concentra tion characteristically occurred around 1000 hours. Days with rain showed a greater dispersion of the maximum with a larger number likely to occur at any hour, though there was still a concentration around 1000 to 1200. hours. The highest hourly concentration observed during the trapping period in 1966 was 128,200/m3 of air at 1500 hours on May 31 from canker no. 2721, which coincided with the beginning of heavy rain, and in 1967 a concentration of 21,600/m^ of air at 1100 hours on June 17 from canker no. 2712. Peak concentration observed during 1964 and I965, when the less efficient traps were in use, was 21,200/m^ of air at 1500 hours on June 19, 1964, from canker no. 2721. These high concentrations were unusual, there being only 53 occasions during 457 canker days of observation when •3 hourly concentrations exceeded 10,000 spores/m of air, and 23 of these occurred on the three days, May 30 to. June 1, I966. The average daily concentration for the three cankers in 1966 for the 84- days when at least 100 spores/m^ per day were recorded was 1,138/m^ per day, and in 1967, 1,314 spores/m^ per day for 97 canker days. Fewer than 100 spores/m-per day were collected on l43 canker days in 1966 and 112 canker days in I967 during the spore release period, or for 63 and 54 per cent of the days respectively. There was considerable variation between cankers and between years in daily spore concentration, ranging from 2, 543/m3 per day for canker no. 2712 in 1967, to 5l4/m3 per day for the same canker Fig. 55- Frequency of hourly maximum aeciospore concentrations for three Cronartium comandrae cankers on 91 d-rv and k-0 rainy days in 1966 and 95 dry and 29 rainy days in 19673 during the main spore production periods. - 145 -TIME (HOURS) - 146 -in 1966. The individual daily maximum concentration recorded was 17,217 spores/m^ per day on May 31, 1966, for canker no. 2713. To further il lustrate the variation between cankers, Figs. 56 and 57 show the average hourly concentration for the month of June in 1966 and for 20 days (June 6-25) in 1967 for three cankers. Although this gives some idea of the variation it must be remembered that the individual cankers were at dif ferent stages, in their spore production and that the area of aecial pro duction also varied between cankers, therefore eliminating detailed comparison. Any difference in wind direction would have a large effect on the spore concentration reaching the orifice, although all traps were placed on the northeast sector of the trees to take advantage of the prevailing southwest winds. The minor peak in Fig. 56 around 0400 hours is largely a result of high spore concentrations occurring on June k dur ing heavy rain. If concentrations for this day were eliminated for canker no. 2721, concentrations for the hours 0300, 0400 and 0500 would be res pectively 38, 164 and 239 spores per hour. All these concentrations re presented the sampling of air at a distance of only 5 cm from the sporu lating canker, from other evidence presented later this concentration was reduced very rapidly with increasing distance from the source. Effects of relative humidity and temperature - Fig. 58 shows that the daily initiation of aeciospore dispersal generally coincided with a fall in humidity from the high nighttime level (above 80$) around 0600 hours and that aeciospore dispersal continued un til this high humidity level was again reached around 2300 hours. Simi larly, there was little or no dispersal at temperatures below 7°C, but a peak dispersal when temperatures were above 20°C, at which time air humidity Fig. 56. Average hourly aeciospore concentrations collected from three Cronartium comandrae cankers in June 1966.;. - lk7 -2000-, HOUR OF DAY Fig. 57. Average' hourly aeciospore concentrations collected from three Cronartium comandrae cankers from June 6 to 25, 1967. HOUR OF DAY Fig. 58. Average hourly aeciospore concentration collected from Cronartium comandrae canker no. 2713 during periods without rain (average 19 to 2k days) in June I966, com pared with the average hourly air temperature, relative humidity and wind speed. - ii+9 -- 150 -levels were at their lowest. It is evident from the graph curves that there is a strong positive correlation between dispersed spore concen tration and temperature and an equally strong negative correlation be tween dispersed spore concentration and relative humidity. However, the average values (Fig. 58) are misleading as on the few nights when relative humidity did not exceed 80%, there was no marked change in the normal diurnal pattern, suggesting that relative humidity is not the sole controlling factor for spore dispersal. Similarly with tempera ture, during some hours when air temperatures remained below 5°0 near the normal time of initiation of daily dispersal, spore counts amounted to I67O to 2800/m3/hr, concentrations which were little exceeded later in the day at much higher temperatures (>20°C). Unfortunately, on the night when relative humidities did not exceed horfo and temperatures one foot above the ground did not fall below l8°C (July 9/l0, I966), total numbers of available spores were very low. Effects of rainfall Fig. 59 would tend to confirm that there is little or no spore dispersal when humidities remain high or when temperatures remain low. The two main rain periods, June 12-19 and June 25-29, 19^5, when humid ities remained high were periods with no dispersal, but during dry con ditions when the diurnal temperature and humidity pattern was maintained there was a daytime dispersal peak. However,.Fig. 60 which illustrates two wet periods in 1966, does not show the same relationship. In these periods there was considerable dispersal during rain. The onset of heavy rain on May 30 and 31 increased the spore concentration which on May 31 Fig. 59- Hourly number of aeciospores collected from Cronartium  comandrae canker no. 2516 during the period June 7 to July 3, 1965, related to hourly air temperature, rela tive humidity, rainfall, incoming radiation and wind speed. - 151 -Fig. 60. Hourly number of aeciospores collected from Cronartium  comandrae canker no. 2721 during the period May 29 to June 9> 1966, related to air temperature, relative humidity, rainfall, incoming radiation and wind speed. - 153 -was maintained for several hours. In these cases the effects of rain splash creating radial air shock waves with turbulent currents and the composite effect of rain droplets hitting the dry spores were important for dispersal. Hirst and Stedman (1963) term them "rain puff" and "rain tap" processes. The resulting peak concentrations were often several times those occurring in dry weather, 6 of the 9 largest daily-hourly maximum canker concentrations in 1966 were associated with rain, usually the onset. The heavy 0.2k inch rainfall that occurred at:1800 hours on May 30, 1966, a time outside the normal peak spore dispersal period, brought about a 23 and 25. times increase in the hourly dispersed spore concentration from two cankers, and this was k and 5 times larger than the 'normal' peak which occurred around Q900 hours on that day. Wight rain caused similar high spore dispersal concentrations on some occasions, at a time when infection conditions were more favourable and had more chance of persisting than during the day. Light showers or steady driz zle did not have the same effect on dispersal as heavy rainstorms, pro bably because there is less turbulence and shaking of the infected tis sues, and because the aecia and their aeciospores gradually become wet thus preventing dispersal. A light shower tended to interfere with the daily dispersal rhythm to a certain degree (eg. June 9? 1966, at 0700 and at..1200 hours. Fig. 60), but often the normal diurnal pattern was resumed within two or three hours of the shower. With prolonged rain the spore concentration in the air was removed, as probably occurred on June 1, 1966 (Fig. 60), when the diurnal pattern was abruptly halted at 0600 hours, but was resumed around 150.0 hours, 7 to 8 hours after the rain ceased. - 154 -Generally a steady or light rain shower reduced dispersal, and dry daytime conditions promoted dispersal, however, a heavy rain, especially with large droplets, at a time when dry mature aeciospores were available for dispersal from the aecia increased initially, the dis persal. A prolonged rain reduced the dispersal of spores as the aecia became too wet. Characteristically, aeciospores appeared to! hang from the aecia in large aggregated masses after exposure to long humid condi tions or following long calm periods when few winds were available to disperse spores from the canker. Effect of dew Dew occurred-on the ground every night of recording in I965 to 1967, at one site in a relatively open stand, although dew only lasted for 2 or 3 hours on a few nights. Around mid-summer, on days without rain, dew commenced on the average at this site just after 2200 hours and dissipated by 07*1-0 hours. In 1965 dew duration was recorded at four sites, two in open grown forest and two in relatively open exposed sites. The duration of dew varied considerably from one site to another within the forest. For example, on June 6, the initiation of dew varied be tween I855 and 2245 hours and the dissipation on June 7 between 0730 and 0905 hours. Generally dew lasted for a shorter period in an open grown forest than in the open exposed areas. Despite the individual site variation there was no corresponding variation of the diurnal dispersal pattern of nearby cankers. On the nights when the initiation of dew was appreciably delayed there was no marked tendency for spore dispersal to continue, suggesting that other factors were more important. A heavy dew tended to delay the daily initiation of dispersal, although how much - 155 -this effect is independent of relative humidity and other factors is unknown as on many of these days the relative humidity remained high until the same hour, although on other occasions, as is to be expected, the relative humidity of the air was decreasing well before dew dis appeared. Effect of wind Fig. 58 shows that the main diurnal spore dispersal period on dry days was closely associated with the higher daytime wind speeds. . On about 80% of the dry days the maximum daily spore concentration occurred during the period of high daily wind speeds; on the other occasions the maximum wind speeds occurred later in the day, presumably at a time when the available mature spores had already been dispersed. Generally wind speeds of 1.1 mph (1.61 ft/sec) were required for considerable initial spore dispersal to take place from the canker, but spores continued to be dispersed at lower wind speeds. At times during the night when higher wind speeds occurred there was no noticeable- rise in spore dispersal from the canker, thus wind speed is not the sole requirement for disper sal. There was also no close correlation between high wind speeds and high spore concentrations, as equally high concentrations occurred at much lower wind speeds, and high wind speeds rapidly dispersed the spore clouds. Availability of spores In addition to the effects of the various meteorological fac tors on spore dispersal, the availability of spores for dispersal must-be considered. Availability depends on the stage of aeciospore production - 15.6 -of the individual aecial pustules and on the magnitude of previous dis persals. More spores are available for dispersal following a few days of non-dispersal, also very high concentrations of spore dispersal early in the day are often followed by only low concentrations in otherwise suitable dispersal conditions. Large numbers of spores are probably available for dispersal for a period of only about four weeks, but the period of spore production lasts two, and in some years three times this duration (cf. next section). Seasonal spore periodicity At the beginning of the spore dispersal season, towards the end of May, there is a very sudden rise in the number of spores released. During this initial dispersal period most of the aecia of a canker rup ture within a few days, releasing the mature spores. This high level of spore release is maintained for two to three weeks to be followed by a gradual decrease (Fig. 6l). By mid-July the daily spore release is at. a low level but a few spores continue to be released on most days until around mid- or late August by which time spore production has ceased and all spores have been dislodged from the aecia by wind and rain action. Minor peaks occur throughout the spore production season after the initial high peak due to the rupture of new aecia, but these only last a few days and spore concentrations rarely reach the level attained during the ini tial spore release period. The day to day release of spores is affected by the seasonal weather conditions, which may bring about minor dispersal peaks independent of the major seasonal peak. In some seasons the spore release is more affected by weather conditions than in others. Fig. 62 shows the seasonal spore release for one tree with two basal cankers in Fig. 6l. Daily number of aeciospores collected from two Cronartium  comandrae cankers during the I966 spore production period, plotted on semi-log scale. 1 - 15? -1,000,000 Fig. 62. Daily number of aeciospores collected from Cronartium  comandrae canker no. 2713 during 1965, I966 and 1967, plotted on semi-log scale. - 158 -- 159 -three different years. In 1965 the lower efficiency spore collector was in use, which accounts for much of the lower level of spore concentration in that year, as the canker appeared to have abundant spores available. The beginning of sporulation was obtained in two years and the end of sporulation in all three years. In I967 sporulation probably began around June 2nd, as on May 31st prominent swelling of the aecial area of the cankers was observed but no aecia were visible. In 1965 spore production ceased much earlier. By July 12 sporulation from the lower of the two cankers had completely ceased and sporulation from the upper canker ceased a few days later. The very high concentrations of May 30 - June 1, I966, and August 5 - 7, 19^7 were associated with heavy rainfall which released large numbers of spores from the aecia. Often, prior to the occurrence of rain, aeciospores would be hanging from the aecia in large masses but rain rapidly dispersed these, and there was often a reduction of the dis persed spore concentration following the rainy period. Spore production from this tree was not followed in I968 but it would have been very low as only two aecial pustules were produced, and the tree was dead through the girdling action of the fungus by mid-July. AECIOSPORE DISPERSAL FROM A NATURAL POINT SOURCE Methods and Material Various methods were used to establish the pattern of aeciospore dispersal around a natural point source. In I967 and 1968, 15 microscope slides lightly coated with vaseline were placed on the ground along eight-radii at 45° .intervals around smalllodgepole pine with abundantly sporu lating cankers. On June 23, 1967, a test was run around a tree (#2713 at - i6o -location l) with two basal cankers. The lower canker extended from the ground to 0.6 feet, the upper canker extended from 0.6 to 1.2 feet. The aspect of the lower canker was predominantly south and the upper east, although both completely girdled the tree and produced spores on all as pects. The circumference of the tree at canker height was 0.55 feet. From June 20 to June 28, 1968, six tests were run around a tree (#2689 at location 3) with a single canker extending from 2.1 to 3.8 feet above ground, which girdled the tree except on the northwest aspect. Spores were produced abundantly from all areas of the canker, which had a cir cumference at canker height of 0.95 feet. On the eight radii in I967, slides were centered at the following distances from the tree: l/4, l/2 1, 1 1/2, 2, 2 1/2, 3, k, 5, 6, 7, 8, 9, 10 and 15 feet. In 1968 the same sampling points were used, except that the 7 and 9 foot points were replaced by 12 and 20 foot distances. The slides were exposed for vari ous periods of time (k to 10-g- hours) during the main daily aeciospore dispersal period, commencing at 0800 hours. Duration of exposure was governed by the threat of rain, which limited the tests on certain days to less than 8 hours (which was considered to include the peak spore dis persal period — 1000 - 1600 hours). After removal, 13-5 sq cm of each slide were scanned under the microscope for deposited spores. Aerial shoots of ground vegetation, which would intercept dispersed spores, were removed from the test areas. Wind speed, air temperature, relative humi dity and rainfall were recorded rear the cankers for the period of the test. Wind direction (and wind speed) was obtained at location 1 from a recorder kQ feet above ground near the study area, and at location 3 from one at 150 feet (operated by the Meteorological Branch, Department - l6l -of Transport, in the Marmot Creek Watershed Research Basin). An impac tion spore collector was run on the test area, 9 inches from the tree on the east-southeast aspect in the 1967 test, and collections.from a net work of plastic slide holder spore collectors (see page 175) supplemented the I968 tests. At location 1 the nearest sources of background spores were from poorly sporulating cankers 100 feet to the southwest and k8 feet to the north. At location 3, the nearest sporulating cankers were 63 feet to the west-southwest and 106 feet to the northwest. Therefore, the number of spores from cankers other then the test trees were probably small and could be ignored. On July 7, 19&8, exposed leaves of ground vegetation growing around a lodgepole pine with a sporulating canker, centered 3 feet above ground, were collected at location 3- Leaves were taken near the four main cardinal points at the following distances from the tree: 1, 2, 55 10, 15, 20 and 30 feet. The area of each leaf was measured, and the num ber of spores present on the surface of the leaves was counted, to give a gradient of spore deposition. Leaves from a number of species were collected, thus the capabilities of leaves to collect and retain spores on their surfaces during different environmental conditions was not uni form. The leaves generally had the same horizontal orientation thus the surface plane of "deposition was approximately the same. Chances for back ground dispersal from other cankers was small as the tree was 63 feet from the.nearest sporulating canker. During the period June 9-, to July 1, 1967, plastic slide holder spore collectors were placed at a height of- one foot and at a distance of 10 feet on the eight cardinal radii around two fairly isolated trees, at - 162 -or near location 1. One or two collectors were placed at a distance of 20 feet. The canker on tree #2719 extended from 3.2 to 7.9 feet above ground and was sporulating on the north and south aspects, and on tree #2710 the canker extended from 0.9 to 1.3 feet above ground. In I968, tree #2710 was again used for 10 days between June 17 and July k, and tree #2689, at location 3, for 9 days between June 2k and July 15. In I968 slide holders were placed at the 5 and 10 foot distances, and around tree #2689 also at the 15 foot distance. The slides were changed daily between 0800 and 0900 hours, and 13.5 sq cm of each slide were scanned for deposited spores. Chances for background dispersal from other sporulating cankers was small. Wind direction and speed, air temperature, relative humidity and rainfall, were recorded near the test trees. Results Figure 63 shows the concentration pattern of spore dispersal around the infected tree #2713 during a 10^ hour/period June 23, 1967. There was a very marked concentration of spores within a few inches of the tree with a very rapid decrease away from the source in all direc tions. The mean hourly wind direction during the period was from the southwest, except for one hour when it was from the northwest. The mean wind speed at one foot above ground was I.7 mph, and at k8 feet 7 mph. From the mean.wind direction during the period one would expect a larger concentration of spores to be deposited in the northeast sector, but this was not apparent. The higher concentration in the southeast sector may have been due to a greater amount of sporulating aecia on this aspect, Fig. 63. Pattern of aeciospore deposition around Cronartium comandrae canker no.' 2713 during a 10-§- hour period on June 23, I967. - 163 -SCALE 0 - 164 -rather than to some wind factor. The predominance of winds from the southwest may well have accounted for a lack of spores beyond the 5 foot sampling point from the tree in the southwest sector. The. areas along the east-west axis, with slightly higher concentrations around the 10 to 15 foot distance points, may have resulted from a wave pattern of dispersal due to gusty or turbulent conditions. Local turbulence in the air, especially around the tree, was likely to have caused some dis persion of spores in all directions from the center, but at greater dis tances the dispersal pattern was likely to correspond to the prevailing directions of air movement during the dispersal period. To further indicate the rapid decrease in concentration, the values for each of the sampling points beyond the 3-inch point were con verted to a percentage of the concentration at the 3-inch sampling dis tance on the same radii. These were then averaged for the eight radii and the average change of concentration with distance is shown in Fig. 64 on a semi-log scale, along with the change for the southeast radii, which received the maximum concentration of spore deposition close to the tree. The decrease in concentration, much of which is due to diffusion, amounted to nearly 80% in a distance of 3 inches and to nearly 99% at a distance of 3 feet. The southeast radii showed a slightly less rapid change of concentration as would be expected from the radii receiving the highest number of deposited spores. Fig. 65 shows the average concentration pattern of spore dis persal around tree #2689 during 8 hour periods on June 24, 26 and 28, I96.8, when the mean wind speed was 1.5 mph, 4 feet above ground. The pattern was similar to Fig. 63 except that there was less variation due to the Fig. 6k. The average percentage change of aeciospore deposition concentration with distance oh eight radii around Cronartium comandrae canker no. 2713 on June 23, 19&7, and the percentage change on the southeast radii, plotted on semi-log scale. DISTANCE IN FEET Fig. 65. Pattern of aeciospore deposition around Cronartium comandae canker no. 2689, on three days (080O to 1600 hours) in June 1968. - 166 -East SCALE IN FEET 1 0 5 - 167 -averaging of three days of spore deposition. Again there was a very marked concentration of spores close to the source, even more marked than shown in the figure, where the inner circle shows spore deposi tions greater than 2500.spores. Depositions at the 3-inch distance on the radii from northwest to northeast all averaged more than 13,000, and on the north deposition was over 25,000 spores. These high con centrations were largely the result of south and southeast upvalley winds on June 2k, and the availability of spores. On the other days fewer spores were available for dispersal, although the wind speeds and direction were similar. Fig. 66 shows the rapid decrease in the con centration of spores for the individual days plotted on semi-log scale, with the spore concentration at the 3-inch sampling distance taken as 100$. All show the .same steep gradient of deposition near the source, giving a hollow curve, similar to that shown in Fig. Gk for the I967 test. The decrease in concentration much of which is due to diffusion of the spore cloud, amounted to nearly 9°$ in a distance of 1 3/k feet on each day, and .to nearly 99$ in a distance of "6 feet on two of the three days. On June 26, 1968, the decrease was more gradual, although there was more than a 98$ decrease at 10 feet. The average wind speed on this day was slightly higher than on the other two days and could account for the flatter curve.-Number of spores deposited per sq. in of leaf at various dis tances from a sporulating canker showed a similar rapid decrease with increasing distance (Table XVII). Many of the spores were deposited close to the veins of the leaves, and it was noticeable that more spores were retained on leaves with abundant epidermal hairs. Disregarding the non-Fig. 66. The average percentage change of aeciospore deposition with distance around Cronartium comandrae canker no. 2689, on three days (0800 to 1600 hours) in June 1968, plotted on semi-log scale. - 168 -DISTANCE IN FEET - 169 -uniformity of leaf sampling surfaces, the spore catch at 10 feet was nearly 99% less than at 1 foot, and no spores were observed at 20 or 30 feet, although the sample leaves were collected towards the end of the I968 sporulation period. One would expect some deposition at distances beyond 15 feet due to natural diffusion. The total deposition probably represented spores dispersed over several days and indicates that such surfaces did not retain many spores. Counts of daily deposits of spores on the plastic slide holder collectors 10 feet from the cankers gave very low values for spores de posited in 1967. Around one tree (#2710), during the nine days of re.-cording, only 8 of the 72 slides had 6 or more spores deposited on them, and 20% of them had none. There was abundant dispersal from this canker as a spore impaction collector placed a few inches from the canker, with a high collection efficiency, collected a daily average of 13,653 spores during the same period. Around the other tree (#2719;), during 20 days of recording, spores were collected on 9*+ of the l6o daily slides at the 10-foot distance, but only 15 slides had 6 or more spores. Trapping of spores on these.collectors was much higher in I968, especially around tree #2689, probably because of a higher availability of spores from the cankers, and in the case of #2689 a release point averaging 3 feet above ground, which would allow greater distance dispersal. Table XVIII shows the average number of aeciospores deposited per sq cm at various dis tances around #2689 during 9 days of collecting. Again there was a very steep gradient in the number of spores deposited in all directions, with a reduction of between 81.9 and 9*+. 3% from the 5-foot sampling point to the 15 foot point. Most winds during the main daily dispersal period came - 170 -Table XVII. Number of aeciospores deposited on natural leaf surfaces collected along the four cardinal, radii at various dis tances from a sporulating Cronartium comandrae canker on July 7, 1968, at location 3-Distance Leaf area Total No. of from canker examined no. of spores (feet) (sq in) spores per sq in 1 15.78 13,169 834.5 2 8.99 5,117 569.0 5 11.45 1,083 94.1 10 15.10 153 10.1 15 10.35 58 5-6 20 11.42 0 0 30 14.02 0 0 Table XVIII. Average number of aeciospores deposited per square centi meter on spore collector coated slides at distances of 55 10 and 15 feet along eight radii around a sporulating Cronartium comandrae canker at location 3, cm 9 days be tween June 24 and July 15, 1968. Distance Average number of spores per sq cm deposited along each from canker radii (feet) N NE E SE S sw w NW 5 21.8 29.6 19. • 9 21.2 13.9 9.2 8.0 14.8 10 12.0 8.9 5. • 9 ' %h 4.4 2.1 2.8 2.1 15 3.1 1.7 3. .6 1-9 2.5 0.8 0.7 1.4 from the west and southwest, which accounted for the lowest spore values in these direction, and the highest in the north to southeast sector (Table XVIII). - 171 -OTHER. DATA ON DISTANCE OF AECIOSPORE DISPERSAL In an effort to collect some further information on distance of aeciospore dispersal under natural conditions, various types of spore collectors were run at varying distances downwind (for southwest winds) from known sources of spores. The number of spore collectors employed at different times depended on their not being required for other aspects of the study, and on the availability of sporulating cankers sufficiently removed from other sporulating cankers, so that the possibilities of spore deposition from background cankers were kept to a minimum. During ten days in July I966, plastic slide holder spore col lectors were placed at three heights (l, 5 and. 10 feet) and at four dis tances (15, 25, 50 and 100 feet) downwind from a sporulating canker one foot above ground. At the same time a rotorod sampler was placed at a distance of 150 feet from the canker. A few spores were caught at most distances each day, and at most heighte, but on no occasion were more than 11 spores deposited, indicating that only small numbers of spores were carried any distance. Collections at the one foot height compared with the 5 and 10 foot heights showed a ratio.jof approximately 1.5:1-0 indi cating that spores can be transported to considerable heights above the release point. Similar data obtained from impaction spore collectors, equip ped with a single slide instead of 2h slides, placed downwind from a basal canker at 25 and 50 foot distances at the one foot level, showed a decrease of deposited spores between these points. During 15 days when both collectors were operating spores were only deposited at 50 feet on one occasion. Catches at 25 feet ranged, on these days, between 0 and - 172 -12 spores. The 25 foot collector was operated on 7 other days and the catch on these days ranged from 1 to 122 spores. The average catch for the 22 days operation at the 25 foot distance was 10 spores, but a simi lar collector placed within a few inches of the same canker collected an average of 7,993 spores a day during the same period. During the spore dispersal period in 1965 to 1967 inclusive a Hirst spore trap was run 34 feet north of a sporulating canker. In 1966 spores were only collected on 31 days of the period May 19 to August 17, with 15 the highest number of spores collected in one day. Similarly, in 1967 spores were only collected on 10 days of the 43 day sampling period between June 26 and August 18, but on several of these days over 100 spores were collected. Rotorod samplers were run in 1967, 50 and 200 feet north of a sporulating canker for l6 days. Spores were only collected on the 50 foot sampler on 3 days, and on the 200 foot sampler on 2 days, despite the fact that this type of sampler samples air at an average rate of 60 liters per'minute. In 1968 a rotorod sampler was run at a point 100 feet away from several sporulating cankers, and another at a distance of 725 feet from the nearest canker. The samplers were run for 13 days between June 17 and July 4. Spores were only collected on the 100 foot sampler on 2 days, and none were collected at a distance of 725 feet. Weekly recording pollen samplers were operated in 1965 and 1967 at various distances from sporulating cankers during the spore dispersal period. A few spores were collected in 1965 when they were operated at distances no greater than 25 feet, but in 1967 no spores were collected when operated at distances of 150 and 725 feet from the nearest cankers. - 173 -AECIOSPORE DISPERSAL EXPERIMENTS FROM POINT SOURCES Materials and Methods Ten experimental tests were carried out with aeciospores liber ated from artificial spore-ejectors positioned one and five feet above ground level, to establish the concentration pattern and distance, of dis persal under the process of diffusion. In eight experiments spores were released at five feet and in two experiments at one foot above ground. Vaseline coated slides were set up in plastic holders at varying distances from the release point to collect spores. All experiments were carried out over a flat short grass area, surrounded by small trees and buildings at the Kananaskis Forest Experiment Station. These experiments were carried out over a plane surface in the open rather than within a forest stand, to simplify the conditions under which the process of spore dif fusion takes place. In most experiments 3 gms of fresh or stored aecio spores were liberated when winds came from the southwest. For liberation the spores were placed in a spore-ejector (Fig. 67) of similar, design.to that described by Gregory et al. (1961), except that the diaphragm was held between two pieces of brass gauge, and the lower funnel stem was attached directly by some k-0 feet of rubber tubing to the air flow regula ting unit (Fig. 68), situated away from the experimental site. When air was forced through the diaphragm, single spores were carried out through the exit tube of the spore-ejector in the form of spore clouds. The exit tube of the spore-ejector was directed towards the midline of the trapping slide network, so that the spores would be released into the normal flow of wind during the experiment. The rate of spore release was controlled by the air compressor valve, and usually spores were released Fig. 67. Spore-ejector used for liberating aeciospores in dispersal experiments from point source. Fig. 68. Compressor unit with connecting rubber tubing, for regula ting air flow for release of aeciospores in dispersal experiments. Fig. 69. Wind-directional ^plastic slide holder spore collectors (without microscope slides) used for collecting aecio spores in dispersal experiments and from natural sources. - 174 -- 175 -under a pressure of 10 pounds per square inch. To collect spores from the spore cloud at a large number of points downwind from the release source, a simple inexpensive wind-directional spore collector was developed, and 113 of these were used to collect spores during each experiment. Each collector consisted of a standard 2.5 x 7-5 cm microscope slide, with one side lightly coated with vaseline, held in a plastic holder. The plastic holder (Fig. 69) consisted of four pieces of acrylic plastic. A piece to hold the slide, a stem attached to a piece of plastic tubing and a vane to maintain the holder facing into the wind. The collector was placed on a small aluminum rod which had a brass tip and washer upon which the collector turned freely in response to wind. The holder was inclined at 1+5° to the horizontal as Gregory and Stedman (1953) had found that the collection efficiency of a slide inclined at 45° varied less than if a slide was oriented vertically or horizontally. The inclined slide was least efficient at low wind speeds, but was better for larger than for smaller spores.. Others (Har rington et al. 1959> Hyre 1950; Ogden and Raynor i960) have also shown that the catch is greater with slides at a. "+5° angle. Also the uniform' orientation of the wind-directional holders with respect to air flow en sures comparability of samples (Ogden and Raynor i960). Experimental arrangement The spore collectors were set .out in a network at three heights, over an angle of 90° and a radius of 150 feet, with samplers on the mid line up to 200 feet from the release point (Figs. 70, 71 and 72). Also on the mid-line a rotorod sampler (Perkins 1957) was set out at 200 or Fig. 70. Portion of spore collector network, with collectors on different radii at various distances from release point. Wind direction vane and wind speed anemometer can be seen to the left of the stand for the spore-ejector. Note spore collectors at 5 and 10 feet on the masts. Fig. 71. Portion of spore collector network with collectors on different radii at various distances from spore-ejector at release point. Wind speed anemometer to left of spore-ejector. - 176 -Fig. 72. Diagram of point source aeciospore dispersal experiment sampling grid, with 113 sampling collectors laid out for southwest winds. Spore collectors were also located at the 300 and k-00 foot distances along the mid-line. - 177 -© © © © © © © © © © 0 • © # © * © • •. • © *• • • • ':>©•© • © 0 25 50 I I I FEET • Spore release point • Single spore collector at 1 foot ©Mast with spore collectors at 1, 5 and 10 feet - 178 -300 feet, and an impaction spore collector (Powell and Morf 1967) at k-00 feet, to augment the plastic spore collector network. The spore col lectors were set out for a southwest wind- (the prevailing daytime wind direction during the period of the experiments) so that the arc of 90° was centered on a northeast line from the release point. Collectors were placed on nine radii from the release point set at 11.25° intervals, with a collector on each radii at 5, 10, 20, 35, 50, 75, 100 and 150 feet from the release point at 1 foot above ground (Fig. 72). On the 20, 50, 100 and 150 foot arcs collectors were also placed at the 5 and 10 foot levels above ground on alternate radii for collection of spores at higher levels. Maximum collection distances from the release point were limited to 150 feet except on the mid-line by the size of the experimental area. Wind speed was recorded at the 1, 5 and 10 foot levels in the center of the experimental layout, and at 1 or 5 feet at the release point, by Casella totalizing cup anemometers of the Sheppard type. Wind direction, air temperature and humidity were also recorded during the release period. Although the experimental site was selected as the most suit able area available, the presence of buildings and trees to the windward of the release point undoubtedly added the further complication of down-draught and downwash effects, of the type described by Hawkins and Non-hebel (1955), both effects tended to bring spores to ground level more quickly than in unobstructed flow. However, the effects from the re lease heights of the tests were probably small, and similar effects would be experienced in forest clearings or -along forest margins, thus this error was of little importance. - 179 -Experimental procedure In conducting the experiment the prepared microscope slides were placed in the holders of the collectors when favourable southwest winds were indicated. The spores were placed in the spore-ejector, and the compressor was run with the valve closed. At a selected time, when winds were favourable for the collector layout, spore liberation was commenced by slowly opening the compressor valve until the desired rate of liberation was obtained. At the same time the four totalizing wind anemometers were started. After making certain that all spores inside the spore-ejector had been liberated the time was again noted, and the totalizing wind anemometers were stopped and read to give the mean wind speed during the period of the dispersal experiment. The records of the continuously recording wind direction and hygrothermograph recorders were also noted. Table XIX summarizes the time and meteorological conditions for each of the ten experiments, numbers I to X, and indicates the height of spore release. Some spore release inevitably took place in gusty conditions which caused dispersal away from the mid-line. With a shift of wind direction the release of spores was halted until the wind was once again predominatly from the southwest. The experiments were only run on clear days during the hours around noon which corresponds to the optimum conditions and natural period of spore dispersal. After completion of the dispersal experiment the labelled slides were carefully removed from their respective holders and kept in a dust-proof slide box until counted. The chance of naturally dispersed spores being carried into the experimental area during the experiments was very small, as the nearest sporulating infected trees were over 600 feet away from the Table XIX. Particulars of experiments on dispersion of Cronartium  comandrae aeciospores from a point source. Expt. no. Date (June 1966) Time of start (MST) Duration (min.) Spores fresh/ dry Height of liberation (ft.) Mean wind speed at release height (ft/sec.) Dominant wind direction Temper ature (°c) Relative humidity I 8 11.1+9 4 D 1 5.6 SW 16 55 II 21 13.47. 30 D 5 7-9 SSW 21 30 III 22 15.25 30 D 5 11.1 SW(S) 19 36 IV 27 13.47 13 D 5 9-8 SSW 21 . 37 V 27 14.32 13 F/D 5 20.6 SSW 21 37 VI 27 15.52 11 D 5 13.3 SSW 20 40 VII 28 14.22 4 D 1 8.3 SSW 20 46 VIII 28 16.00 7 D 5 12.7 SSW 20 4l IX 29 13.57 12 D 5 6.5 ssw-wsw 22 30 X 30 14.10 30 F 5 5.3 ssw-wsw 22 32 Site: Grass square, Kananaskis Forest Experiment Station. Trap height and position for all experiments: 1 ft above ground on 9 radii, 11.25 degrees apart, at 5, 10,. 20, 35, 50, 75, 100, 150 ft from source. 5 and 10 ft above ground on 5 radii, 22.5 degrees apart, at 20, 50, 100 and 150 ft from source. Spore quantity: 3 gms. in all experiments, except Expt. 1 (l gm) and Expt. V (5 gms). - 181 -release point, and considerable forest growth intervened which would intercept most spores carried by air currents close to the ground. Analysis of data Spore deposition from the spore cloud was counted on a 13-5 sq cm area of the exposed portion of the slide by making 2 mm wide traverses of the slide under the microscope, or on a third of this area by counting every third traverse. Comparison of the two methods of spore counting indicated that the method of scanning only a third of the area gave a fair estimate of the total catch on the total exposed area of the slide. All counts were adjusted for a 3 gm spore release quantity. The spore collectors undoubtedly had a relatively small impaction efficiency, and therefore offered an underestimate of spore cloud concentration. They were not representative of natural surfaces, but there was little evi dence of spore blow-off because of the sticky coating. The total counts were plotted on scale diagrams of the sampling area and isopleths of spore concentration drawn for the three levels of trapping. Patterns of dispersal do not always show a smooth unimodal distribution because changes in wind direction associated with a few large scale eddies during the short sampling period result in multimodal patterns covering different angular sectors. For this reason mid-line data alone are often inadequate and therefore averages for various sector sizes (22*5, ^5 and 90°) have been used which represent more accurately the typical horizontal plume dispersal pattern. Vertical profiles of mid-line and sector data were constructed to show change of concentration with distance under the dif ferent wind speed conditions and levels of spore release. Vertical com parisons between different heights introduces a possible error since - 182 -efficiency of spore catch is a function of height. Also the use of the mean wind speed during the release period was a possible source of error particularly at the one foot level, since the efficiency of spore collec tion on the collectors would vary rapidly with wind speed and especially at low speeds. Measurements were made of sizes of spores deposited at different distances and heights from the release point to establish whether there was any correlation between spore size and distance or height of dispersal, and between fresh and stored spores used in the tests. As simultaneous measurements of cloud concentration and spore deposition on the. ground were not taken, and the volume of air sampled by the spore collectors was unknown, it was not possible to estimate accurately the.amount of deposition that occurred in the experimental area. However, a numerical ratio of spores collected at intervals from the source was calculated geometrically, assuming that all other factors affecting dispersal and deposition were constant. A geometrical approach is not too satisfactory (Gregory 1961), but under the conditions of the present experiment series this did afford an estimate of the number of spores expected at greater distances from the source. If the chances for spore collection were the same for all distances within the sampling area, one would expect the concentration of the spore cloud to decrease in inverse proportion to the square of the distance from the release point. Geometrically the cone shape best describes the expansion of a cloud of spores. In cross section the cone is not circular, but is greater in the horizontal or lateral dimension. If a pyramid base is assumed instead of a cone base the proportions of the basal areas at any distance from the - 183 -release point will remain the same, thus little error is introduced. To find the cross sectional area of the pyramid at each sampling distance through which the spore cloud passed it was assumed that in each test the vertical dispersion was limited to 15° from the horizontal, and lat eral dispersion was limited to 30° on either side of. the mid-line. The vertical dispersion of 15° was chosen as spores were collected at 5 feet above the release point at a 20 foot distance in all tests from a 5 foot-release point. Changing the angle of lateral dispersion to other limits would not change the proportions of the areas at each distance, these would remain constant, therefore the angle of lateral dispersion was not important. The cross sectional areas at 20, 50, 100 and 150 feet from the release point, were calculated from the equation: Spore release height + the arc for 15° x the arc for 60° = area sq ft, e.g. at 20 feet the cross sectional area = (5 + 40rf) x 40rt = 214.53 sq ft. The areas at 24 6 the four sampling distances were reduced to rough proportions as follows: 1 : 4.4 : 15.2 : 32.4. These distances were chosen as they included spore collections at three heights on five radii 22.5° apart. Spore collections on the other four radii at 1 foot above ground were not included in the analysis. Two methods of estimating the expected catches at the various distances were employed, using the observed spore counts: A) by accepting the total number of spores caught at 20 feet as a reference point the expected catch at the further distances could be calculated by dividing the total 20 feet catch by the proper ratio fraction, or B) distribute the total number of spores caught- at the four distances in inverse proportion to the cross sectional areas at the various distances. These methods were used to analyse the catches for all releases at 5 feet. - 184 -Results Results of the ten experiments are summarized in Tables XX and XXI, giving the total catch.at each distance and height. Complete results for one typical experiment are given for the spore catch at the one foot height at each sampling point, in Table XXII. Concentration patterns Figures 73 and 74 illustrate the concentration patterns at three heights for releases from the one and five'foot levels in two separate experiments. The decrease in spore concentration was very evi dent being most marked in the one foot releases. The point of maximum concentration one foot above ground was further from the source with a five foot release than a one foot. Large differences were evident in concentration patterns from one experiment to another which could be related to wind speed and turbulence differences. Wind direction shifts or eddies of varying sizes during the experiment tended to dis perse the spores by diffusion unevenly in the trapping sector. Mean wind direction was always close to the mid-line, but gustiness could distort the average picture. Wilson and Baker (1946) found a similar variation, but when the release period was increased to 30 minutes or more, the resulting distributions were found to resemble more nearly that of normal probability. A composite concentration pattern for all the experiments released from the five foot level is shown in Fig. 75 for the three heights. Change of concentration with distance Vertical profiles of concentration are shown in Fig. 76 for - 185 -Table XX. Total number of spores trapped at one foot above ground at each distance on all radii. Expt. Distance from source (feet) no. 5 10 20 35 50 75 100 150 1* 17,7^8 II 1, 377 III 438 IV 33 V 46 VI 3 VII* 8,010 VIII " 15 IX 108 X 194 *Spores released 2,736 324 27 1,566 612 159 552 586 210 57 231 96 60 ' 59 58 58 99 54 6,894 1,392 252 82 216 221 627 207 87 246 536 383 at 1 foot, all others 0 0 0 0 84 5 4 37 70 89 81 9 27 18 0 0 47. 7 5 25 51 18 3 0 126 18 9 9 131 82 44 • 33 66 ' 51 21 12 198 77 45 19 at 5 feet. Table XXI. Total number of spores trapped at 5 and 10 feet above ground at 20, 50, 100 and 150 feet on five radii. Expt. 5 feet above ground 10 feet above ground no. Distance from source (feet) 20 50 100 150 20 50 100 • 150 I* 36 0 0 0 9 0 0 0 II 126 21 17 14 26 6 1 0 III 450 134 37 10 237 49 10 5 IV 75 21 0 0 21 21 3 0 V 232:! 34 2 0 79 38 14 2 VI . 63 33 6 0 36 3 6 3 VII* 132 45 15 9 30 18 15 v 9 VIII 239 53 10 5 4o 22 18 4 IX 213 21 3 15 171 45 15 3 X 300 101 8 14 122 113 32 9 *Spores released . at 1 foot, all others at 5 feet. Fig. 73- Aeciospore concentration patterns at three sampling heights for a spore release one foot above ground (Ex periment VII). 1 FOOT 5 FEET 10 FEET i i i i i i 0 10 20 30 40 50 FEET Fig. 7k. Aeciospore concentration patterns at three sampling heights for a spore release five feet above ground (Experiment X). Fig. 75- Composite aeciospore concentration patterns at three sampling heights for all eight spore releases five feet above ground (Experiments II-VI, VIII-X). 0 Fig. 76. Vertical profiles of aeciospore concentration patterns along the mid-line of the sampling network for two spore releases at one foot (Experiments I and VII), and two spore releases five feet above ground (Experiments V and X). - 189 -Exp. I lO-i DISTANCE FROM SOURCE (FEET) - 190 -Table XXII. Results of Experiment X, showing number of spores trapped on an area of 13.5 sq cm at the one foot level at each sampling point. Angle to wind Distance from source (feet) (°) 5 10 20 35 50 75 100 150 200 -45.0 8 6 14 19 12 3 0 0 -33.75 22 20 48 30 11 4 3 0 -22.5 13 31 41 34 26 4 4 4 -11.25 8 36 l8l ill 53 19 11 0 0 11 22 120 73 39 24 8 5 0 +11.25 69 28 4o 69 26 7 4 2 +22. 5 15 36 44. 24 20 5 9 1 +33-75 26 51 10 9 6 8 6 5 +45.0 22 16 42 14 5 3 0 2 Total 194 246 536 383 198 77 45 19 -.-line profiles for four experiments, and in Fig. 77 for vertical pro-files encompassing various sector angles of the experimental area for Experiment VIII. Experiments I and VII, the two tests with release of spores at one foot, both show a maximum concentration at the five foot distance, the closest sampling distance, with a very rapid reduction at greater distances from the release source (Fig. 76). In Experiment I, no spores were collected more than 35 feet from the source and upward dispersion was less than in Experiment VII, when spores were dispersed to a height of 10 feet within a distance of 20 feet. The other experi ments also show considerable variation of the concentration along the mid-line. Generally the maximum concentration sampled occurred at a distance of 20 feet and at a height of 5 feet with a five foot release point (Fig. 78). In Experiments II and X the maximum concentration at the 20 foot distance was at the one foot height. In both these tests the wind speed was less and fresh spores were dispersed in Experiment X. Fig. 77. Vertical profiles of aeciospore concentration patterns along the mid-line and for H-5° and 90° sectors of the sampling network, for one spore release at five feet above ground (Experiment VIII). - 191 -EXPERIMENT VIII 0 5 10 20 35 50 75 100 DISTANCE FROM SOURCE (FEET) Fig. 78. Composite vertical profiles of aeciospore concentration patterns along the mid-line and for and 90° sectors of the sampling network, for all eight spore releases five feet above ground (Experiments II-VI, VIII-X). DISTANCE FROM SOURCE (FEET) - 193 -In Experiment II, when the mean wind speed was only 7.9 ft/sec, higher concentrations were deposited at the 5 and 10 foot distances than at the 20 foot, whereas all other 5 foot release tests gave higher con centrations at the 20 foot distances and the 1 foot height, as shown by the composite vertical profile of Fig. 78. Relatively high concentrations were obtained in Experiments III, V, IX and X, at the 10 foot height at the 20 foot distance, illustrating that spores may be carried upwards by air turbulence, and transported beyond the experimental area before being deposited. Experiment V, which had the highest mean wind speed (20.6 ft/ sec), shows this possibility best. Size of spore apparently had little effect on dispersal patterns, measurements of 10 spores collected at 5, 20 and 50 feet from the source in each experiment gave no evidence that larger spores were deposited closer to the source or that smaller spores were deposited further away. Variation in spore size and its effect on rate of spore fall, therefore, could not be said to have any effect at the windspeeds of the experiments. Lateral spread of spores extended to, 1+5° on either side of the mean wind direction at distances of 10 and 20 feet in all experiments except Experiments TV and VI, which showed only 22.5° or 33-75° lateral spread on one or both sides of the mean. This was probably a result of the influence of wind fluctuations. Sreeramulu and Ramalingam (1961) found that lateral spread varied depending on time of day, mean wind speed, and duration of spore liberation. In their cases lateralj'spore spread extended to 30° on either side of the mean wind axis, and around noon and early afternoon, the time of my experiments, extended to 50° and 70°. Lateral spread in some of my experiments extended well beyond - 19k -k5°, as was illustrated by some of the high spore deposits on the 45° radii from the mid-line, e.g., 753 spores at 10 feet in Experiment VII, 105 spores at 10 feet in Experiment IX. Wilson and Baker (1946) found that at low-wind velocities dispersion at a given distance from the source was frequently much greater than at medium velocities, probably because of more variability in wind direction at lower than at higher velocities. In the ten experiments spores were only collected at the 200 foot distance three times with the plastic slide holder collector, and twice more by the more efficient rotorod sampler, which was placed at this distance for three tests. With the rotorod sampler placed at 300 feet spores were collected during three of five tests.- An impaction spore collector placed at 400 feet collected spores (2) only during one test. The lack of spores collected at greater distances from the source and the rapid decreases of concentration were misleading. Each sampling point only collected spores whose flight path lay through the small vertical area- taken up by the inclined slide. Due to diffusion in the vertical and horizontal planes, the concentration of the- spore cloud must have decreased approximately, with the square of the distance from the source. Therefore, the observed rapid decrease of the spore concen tration may have been due largely to diffusion and not have represented depletion of the spore cloud by deposition. To obtain an estimate of the expected decrease of the sampled spore concentration with distance, due to diffusion, two methods were employed, accepting the total spore catch at four distances, or the catch at 20 feet, as reference points (Table XXIIl). - 195 -Table XXIII. Comparison of the observed and expected concentra tion of spores at four distances from the 5 foot release point during eight tests. Experiment Type of Spore concentration at each no. data^'^ sampling distance (feet) 20 50 100 150 II Observed 764 111 51 Expected A 764 174 50 24 Expected B 580 272 79 18 III Observed 1273 253 128 24 Expected A 1273 289- 84 39 Expected B 1023 481 i4o 32 IV Observed 327 79 3 0 Expected A 327 74 22 10 Expected B 250 117 34 8 V Observed 618 198 36 45 Expected A 618 141 41 19 Expected B 548 257 74 17 VI Observed 198 87 15 0 Expected A 198 45 13 6 Expected B 183 86 25 6 VIII Observed 495 206 72 42 Expected A 495 113 33 15 Expected B 498 234 68 15 IX Observed 477 132 39 30 Expected A 477 108 31 . 14 Expected B 4l4 195 56 13 X Observed 683 316 85 35 Expected A 683 155 45 21 Expected B 684 321 93 21 Average Observed 6o4 173 50 28 Expected A 6o4 137 4o 19 Expected B 523 245 71 16 -^Observed values are totals of spores collected at three heights and on five radii at each distance. ^Expected values are estimates of spores collected at each distance, A) accepting catch at 20 feet and assuming ratio of 1:4.4:15.2:32.4, and B) taking total catch at all distances and distributing the total spores in inverse proportion to the cross sectional areas at each distance. - 196 -Inl.Experiments V, VIII, IX and X, the observed counts at several distances were higher than those estimated by method A, which assumes an accurate measure of the spore concentration at 20 feet. Experiments III, IV and VI, all showed a lower total at the 150 foot distance than the expected, which may have been due to greater vertical dispersion. Experiments II and V showed higher observed totals at 15O than 100 feet. Based on method B the expected values at 20 feet were generally much lower than the ob served values, except in Experiments VI, VIII and X. Table XXIII indicates that much of the observed decrease in spore concentration with distance was due to diffusion of the spore cloud and not to deposition. It was difficult to interpret the estimated and observed values. The difference in the two methods for obtaining expected values was largely a problem of redistributing the data, assuming accuracy for some of the observed values. Method A, on an average, showed that the observed values at 50, 100 and 150 feet were too high, but method B indicated that observed values were too low at 20, 50 and 100 feet. Probably there was considerable variation and error in the observed values, making spore diffusion estimates based on these questionable. At low wind speeds the efficiency of the collector was low and the catch therefore was small, thus the error of estimation becomes great. 'There could have been momentary large variations in spore concentration between adjacent sampling areas due to changes in turbu lence and size of eddies, or fluctuations in wind velocity which affected the vertical and lateral dispersion from one distance to another. The number of spores passing through a unit area would therefore have varied greatly over the sampling network. With an increase in wind velocity the expected degree of dispersion would decrease and the amount of deposition - 197 -would be reduced. In unstable air on clear days, spore clouds may move in a series of great loops (Waggoner I965I and certain sampling points may have been skipped which could have accounted for points closer to the source having lower spore concentrations (Gregory 1968). RATE OF FALL OF AECIOSPORES IM CALM AIR The distances that spores are carried in dispersal depend on the altitude reached by spores in the air currents, the rate of spore fall under the influence of gravity, the velocity, azimuth and duration of winds, and the vertical mass exchange brought about under turbulent movement. If the rate of fall of a spore in calm air is determined, the theoretical dispersal distance for a spore liberated at a given height, or carried to a certain altitude can be calculated for any mean wind speed. More important, the velocity of fall is- an important factor in determining the flight path of a spore in atmospheric turbulence. Materials and Methods Various types of cylinders for measuring rate of.Vfall of spores or pollen grains have been employed by McCubbin (1918a),' Ukkelberg (1933), Durham (19I+6), Weinhold (1955), and others. In the present study the rate of fall of aeciospores was mea sured in a plastic cylinder, 3m long and 23 cm in diameter. The cylinder was grounded and coated with teflon powder to reduce static effects. The upper end of the cylinder was covered with a tightly-fitting acrylic plas tic disc with a center hole for release of spores. The hole was covered - 198 -with a plastic strip before and after spore release to minimize turbulence within the cylinder. The upper end of the cylinder was attached to the ceiling allowing the cylinder to hang freely in a room held at constant temperature (22°C, 30% R.H.) with minimal air movement in the experimental room. The lower end of the cylinder consisted of a tightly-fitting shal low dish with a false bottom made from 0.64 cm thick acrylic plastic. In the base of the dish an opening 5 cm x 1.5 cm was cut to allow free fall of aeciospores on to an exposed vaseline coated slide. A slide carrier 56 cm long was constructed to hold eighteen 2.5 x 7-5 cm microscope slides. The slide carrier was inserted through the openings of the false bottom of the cylinder, passing directly through the center of the cylinder under the spore outlet. During the fall of aeciospores each slide of the series was exposed beneath the spore outlet for 15 seconds. The changing of slides required less than a second. Before each trial the cylinder was carefully washed with water to remove all spores from the wall and base. The cylinder was then dried, freshly coated with teflon powder and allowed to stabilize in the environment of the experimental room. A control slide was exposed in the cylinder before each trial to determine whether it was free of spores.. The aeciospores used in the experiments were either fresh spores or had been stored for known periods of time, the samples being carefully screened to remove extraneous material and to reduce the num ber and size of any spore clumps present. A measured amount of spores was released into the cylinder by inverting a 2 ml vial containing the spores over the open hole at the top of the cylinder. All spores col lected on the slides therefore had fallen 3 meters. Traverses of the complete exposed portion of each slide were made under the microscope to - 199 -count the number of single spores collected during each exposure period. The number of clumps of spores deposited on each slide were also noted. Larger clumps tended to occur on the earlier exposed slides, but unfor tunately estimates of the size of clumps' were not kept. The diameters of 20 spores were measured on the second and every second or third slide thereafter, of each series, to check for any relationship between the size of spore and rate of fall. One test was also carried out with each slide exposed for 30 seconds to check whether all spores were being de posited in the 4 min 15 sec duration of the other tests. No spores were collected from K min 30 sec onwards, thus the exposure duration was satis factory. Trials were also carried out with wet and dry spores to check for any marked effect of moisture content on the rate of fall. A col lection of spores was divided into two equal weight units. One unit was released immediately in the cylinder. The other was allowed to take up moisture:'.in a saturated atmosphere for h8 hours, was reweighed and re leased in the cylinder. An increase in weight, following exposure to a saturated atmosphere, was recorded. Diameter measurements of spores were also made from the wet and dry samples immediately after release in the cylinder. The formula and method used to calculate the rate of fall fol lowed that used by Ukkelberg (1933)- This formula calculated the average number of seconds required for single spores to fall. The formula was Zfx/N, where Z= summation, f = frequency in a class (number of spores deposited on each slide during each exposure interval), x =. class center (mean of each exposure period) and N = total number of spores deposited. - 200 -By using the class center, the assumption was made that the spores were deposited uniformly, throughout each exposure period, a situation which probably did not occur. The slight error introduced, however, would have little effect on the mean as the variations would be in both direc tions. The rate of fall as expressed in centimeters per second was ob tained from the following formula: R = D/T, where D = distance fallen ( and T = average time in seconds required to fall the given distance. Results Considerable variation was found in the time required for sin gle aeciospores to fall the length of the cylinder. Table XXIV shows the percentage of single aeciospores deposited 3 meters from point of liberation on glass slides exposed successively for 15 second periods during ten individual rate of fall trials, and the average of these trials. The rate of fall of over 55,000 individual aeciospores was measured. In three trials a few spores fell 3 meters in 15 seconds or less, while other spores-required over k minutes to fall that, distance. On average 82% of the aeciospores fell within 2 minutes and 97% within 3 minutes. The average number of seconds required for the single spores of the ten different trials to fall 3 meters, and the average rate of fall are shown in Table XXV. By averaging the mean rates of fall one obtains an average velocity of fall for aeciospores of 3-64 cm per second, however, the range of the individual tests is considerable (2.79 to 5.10 cm per sec). Much of this variation between individual tests can be attributed to the degree of spore clumping in the test sample and to its associated mass .subsidence, which was more obvious in some tests Table XXIV. Percentages of individual tests of aeciospores of Cronartium comandrae deposited 3 meters from point of liberation on glass slides exposed successively for 15 second periods in a closed cylinder. Test Exposure periods (no. of seconds after spores were liberated) no. 0- 16- 31- 46- 61- 76- 91- 106- 121- 136- 151- 166- 181- 196- 211- 226- 24l-15 30 45 6o 75 90 105 120 135 150 I65 180 195 210 225 240 255 1 0.3 2.3 35-5 19-5 2.1 5-3 9- •9 7.1 6.2 5.8 2.6 1.8 1.1 0.3 • 0.3 0. 0 0.0 2 0.0 0.1 0.8 7.9 13.4 13.8 17. ,2 14.8 10.9 7-5 5-0 2-5 2.6 1.5 0.6 ' 0. 7 0.5 3 0.0 0.2 io.8 16.7 23.4 4.3 5. ,7 12.9 9.8 4.9 3-1 3-1 1.8 1.1 1.7 0. 2 0.2 4 0.0 1.1 7-3 10.8 28.5 20.6 12. ,1 6.3 3.8 2.0 1.6 2.7 1.2 0.8 0.4 0. 5 0.5 5 0.5 11.1 20.4 15.9 8.8 13-7 11. • 7 7.0 3.0 2.5 1.8 0.9 1.0 0.5 0.3 0. 5 0.3 6 0.0 0.2 1.2 36.2 21.5 16.0 10. .2 3-9 3.8 1.9 1.8 1.2 6.8 0.6 o.h 0. 4 0.2 7 0.0 0.0 0.7 3-5 21.1 25.8 12. ,1 10.9 6.0 7-0 5-2 2-9 1.4 1.5 1.5 0. 3 0.3 8 0.1 0.3 1.4 19.1 1.8 19.2 22. • 7 15.8 .9-9 3-6 2.3 1.1 0.7 1.2 0.6 0. 1 0.1 9 0.0 18.3 36.4 9-4 6.6 7-4 9-• 3 4.9 4.0 2.1 0.5 0.7 0.2 0.0 0.1 0. 1 0.1 10 0.0 7-9 18.8 4.2 12.8 21.3 14. ,2 6.5 4.9 2.1 3.1 1.6 1.7 0.6 0.2 0. 0 0.1 Average 0.1 4.1 13-3 14.3 i4.o 14.7 12. • 5 9.0 6.2 3-9 2.7 1.8 1.3 0.8 0.6 0. 3 0.2 - 202 -Table XXV. Average rate of fall in still air of ten aeciospore tests Test Total no. Aver. no. seconds Aver, rate of fall no. of spores to fall 3 m. (cm/sec.) 1 3,278 74.12 4.04 2 6,909 107.39 2.79 3 M78 92.34 3.24 1+ 5,509 85.60 3.50 5 11,910 72.51 M3 6 4,674 78.46 3.82 7 3,789 102.53 2.92 8 3,144 96.43 3-11 9 6,947 58.81 5.10 10 ^,575 80.4i 3-73 Mean — 84.90 .3.64 • than in others. On being released at the top of the cylinder it was noticeable that some spores fell more rapidly than others initially, but as the spores fell further the mass of spores became more dispersed and the fall speed of individual spores was more uniform throughout the test. By the time this fall rate became more uniform some spores were a third or more of the way down the cylinder. The number of spore clumps de posited on the individual slides in the ten tests is shown in Table XXVI. No clumps occurred on any slides exposed after 2 min 15 sec. There was no way of knowing how many individual spores had broken away from clumps during the fall period and therefore had an excessive fall rate for a portion of the period. However, some samples had very few clumps and these tests, e.g. 7 and 8 (Table XXV), probably give a better estimate of the true rate of fall. Test numbers 1, 5 and 9, all had numerous clumps, with most being deposited on the early exposed slides, which indicated that they were probably larger clumps which may well have created a higher degree of mass subsidence. These tests all had double maxima in their - 203 -Table XXVI. Number of aeciospore clumps deposited 3 meters from point of liberation on glass slides exposed successively for 15 second periods in a closed cylinder. Test no. 0- 16- 31- - 46- 61- 76- 91- 106- 121- Total no. 15 30 45 60 75 90 105 120 135 of clumps 1 0 4 22 9 2 l 0 0 0 38 .2 0 0 2 6 4 5 4 4 1 26 3 0 1 0 6 8 4 2 0 0 21 4 0 0 5 5 5 0 0 0 0 15 5 1 17 15 13 4 2 1 0 0 53 6 0 0 3 0 1 0 0 0 0 4 7 0 0 0 1 2 0 0 0 0 :3 8 0 0 0 7 0 ' 0 1 0 0 8 9 0 7 29 8 3 3 0 0 0 50-. 10 0 4 7 4 3 0 0 0 0 18 rate ! of fall (Table XXIV), and had the highest avera^ le rate of fall of the test series (Table XXV). . Test no. 10 also had a doub le maxima wit! the first peak occurring within 45 sec of spore : liberation (Table XXV) Two other tests had double maxima but these occurred much later after the release. If the average rate of fall for tests with early double maxima (nos. 1, 53 9 and 10) is omitted, then the remaining- tests show an average fall rate of 3-23 cm/sec (range 2.79 -' 3-82), which is pro bably much closer to the true rate, although initial mass subsidence probably still caused this to be an overestimate. The one test with slides exposed for 30 sec intervals gave an average fall rate of 3-06 cm/ sec. Other factors may cause variations between tests. There may 1 •: have been slight differences in the room environmental conditions, un avoidable air currents within the cylinder, •-or variations in spore size or spore moisture content. Measurements of length and width of spores deposited on the second, fourth, sixth and seventh slides of each trial - 204 -series showed little overall variation, although smaller spores generally took longer to fall. The spores for the ten tests were collected from in fected trees, maintained in the greenhouse, and kept for periods from one hour to two days before being released, thus varying degrees of spore drying were experienced which explained some of the test variation. To explain the effect of this variation a separate experiment was run with dry and saturated wet spores, originating from the same source. Table XXVII gives the average rate of fall of the dry and wet spores in two tests. The rate of fall of the dry spores (average 3.46 cm/sec) was probably an overestimate, but the wet spores fell about 2.7 times faster. There was considerable clumping of the wet spores and mass subsidence was evident, but the change of spore fall rate with variation in spore moisture content was illustrated. Measurement of the length and width of 10-15 spores de posited on the second, sixth, tenth and fourteenth slide during the first dry-wet test showed some variation (Table XXVIII), with smaller spores generally taking longer to fall. Table XXVII. Average rate of fall in still air of dry and wet aeciospores during two tests. Test Spore Aver. no. seconds Aver, rate of no. condition to fall 3 m. fall (cm/sec.) 1 Dry 80.91 3.52 2 Dry 88.41 3-39 1 Wet 28.84 10.02 Wet 32.60 9,20 - 205 -Table XXVIII. Average length and width (u) of aeciospores deposited on slides during the dry and wet spore tests of rate of fall. Slide Dry Wet no. length width length width 2 62.4 21.1 55.9 28.9 6 61.5 22.4 55.7 28.1 10 55-5 19-4 49.1 28.5 14 55-5 19-6 45.0 29.1 AECIOSPORE DISPERSAL DISCUSSION Meteorological factors and spore periodicity Diurnal periodicity in airborne organisms was first observed by Pierre Miquel in the last quarter of the nineteenth century in France, from his daily counts of bacteria and molds (Miquel 1878-1899). Since 1950 there has been renewed interest due largely to the development of the Hirst spore trap (Hirst 1952) and other continuous air samplers. The diurnal periodicity of various components of the air fungal spora have now been reported by many workers (Adams 1964; Cammack 1955; Carter and Banyer 1964; Cole 1966; Gregory 1952, I96I; Gregory and Hirst 1957; Gregory and Sreeramulu 1958; Hamilton 1959; Harvey 1967; Hirst 1953; Kramer e_t al. 1963, 1964; Lukezic and Kaiser 1966; Mills I967; Pady et al. 1962, 1965; Panzer et al. 1957; Pathak and Pady I965; Pawsey 1964; Rich and Waggoner 1962; Shanmuganathan and Arulpragasam 1966; Sreeramulu 1959, 1962, 1963; Sreeramulu and Seshavatarum 1962; Sreeramulu and Vittal 1966; Van Arsdel 1967; Waggoner and Taylor I958). Gregory (1961) divided the spore diurnal patterns into five types: bacterial (with two maxima and - 206 -minima), nocturnal, forenoon, afternoon and evening. Generally the pat terns with C. comandrae aeciospores fall into a"forenoon" or "afternoon" pattern. The latter pattern is typical of the majority of the daytime dry spore forms, and the former group includes a few crop-pathogenic fungi (Gregory 1961). These patterns show a peak spore concentration between 1000 and 1600 hours, although the weather of a particular day may disturb the normal rhythm. A review of the literature showed little comparable information on the diurnal periodicity of rust spores of individual spe cies, usually the data were grouped under 'uredospores' and 'basidio-spores', sometimes with an indication of possible sources. Wo information was noted on the dirunal periodicity of rust aeciospores. Hirst (1953)5 in his reports of the air-spora at Rothamsted Experimental Station in 1951 and 1952, stated that "no aecidiospores were recognized". The pre sent study is therefore the first for this rust spore state. Information on rust uredospores generally show an afternoon pattern (Adams I96U-; Cammack 1955; Hamilton 1959; Hirst 1953; Pady et al. 1965; Sreeramulu 1959; an(i Powell, unpublished data for Cronartium comandrae) . Basidio-spores, as ,a group, show a predominantly nocturnal pattern (Adams 1964; Gregory I96I; Gregory and Hirst 1957; Hirst 1953; Kramer et al. 1963; Pawsey 1964; Shanmuganathan and Arulpragasam I966; Sreeramulu and Seshavataram 1962; Van Arsdel 196?), although few reports have been ap plied to individual species of Basidiomycetes., and fewer still to rust species, eg. Puccinia malvacearum Mont. (Carter and Banyer 1964), Cronartium ribicola (Van Arsdel 1967)• Other studies also found evidence of a double or secondary peak in the air spora concentration (Gregory and Stedman 1958; Harvey - 207 -1967; Pady et al. 1965; Pathak and Pady 1965; Pawsey 1964; Rich and Waggoner 1962). In some cases this was a persistent feature, but in others occurred sporadically or through some combination of environ mental factors. Most of this work concerned the diurnal periodicity of Cladosporium spores, where a peak was recorded in the forenoon, and a second smaller peak in the afternoon or early evening. This peculiar diurnal cycle was first discussed by Rich and Waggoner (1962) who ascribed the phenomenon to the cycle of atmospheric turbulence, whereby mature spores become airborne with the increasing morning turbulence to produce a forenoon peak. During the afternoon or early evening the turbulence lessens allowing the spores to settle and give rise to a secondary peak. The midday minor minimum was explained by the spore source being depleted as there was only a single daily spore crop and because the concentration of spores in the atmosphere was lowered. It is not known if rusts produce a single daily crop of aeciospores, but the atmospheric turbulence cycle may help to explain the occasional evidence of a secondary peak with aeciospores of C_. comandrae. It is more likely that, because the spores were collected close to the source, a combination of interrelated factors was involved in disrupting the typical diurnal pattern. The apparent close relationship of spore dispersal period icity with temperature and relative humidity is not unexpected, in that it is typical of most components of dry weather air spora. The latter show an increase in spore concentration generally coinciding with the start of atmospheric turbulence (a rapid change in temperature, humidity and wind conditions). Turbulence generally increases from sunrise to - 208 -around noon and then falls again, and it seems likely that turbulence is a more important factor than temperature and humidity acting separately for daily spore dispersal, although the latter are contri buting factors to turbulence. Turbulent conditions were associated with all peak aeciospore concentrations whether they occurred in the forenoon or afternoon, on dry sunny days or associated with the onset of heavy rain showers or storms. Lukezic and Kaiser (1966) showed a highly significant correlation between Fusarium spore concentration and atmospheric turbulence. Holmes and Bassett. (1963) showed a similar relationship with diurnal ragweed pollen dispersal studies at Ottawa, and observed an evening rise in pollen count on about half the days, which may have been associated with pollen settling following a decrease in atmospheric turbulence. Mills (1967) showed that turbulence was the main disseminating agent for Ustilago avenae (Pers.) Rostr. when wind velocities were low. The varying effect of rain, has been noted by other workers. Hirst (1953) observed that light rains reduced the concentration of Cladosporium spores, but that a thunderstorm increased concentration. Ainsworth (1952) and Gregory (195*0 also found a rapid increase in Cladosporium spore concentration during rain showers, but Hamilton (1959) found an appreciable decrease. Rich and Waggoner (1962) found an increase in the Cladosporium spore concentration with the heavy rain showers of unstable conditions, but the concentration was not increased by light rain and drizzle associated with stability in the atmosphere. Hirst and Stedman (1963) reported that the transient increase in spore concentration of many spore types associated with - 209 -onset of rain was discernible in about half of the rain periods. Shanmu-ganathan and Arulpragasam (1966) noted a sharp increase in basidiospores of Exobasidium vexans Massee, either during, or immediately after heavy afternoon thunderstorms. If the rain was prolonged the concentration de creased after the initial onset due to removal of spores from the air, however, a steady drizzle had little effect. Jarvis ('1962) found that rain showers were frequently associated with high spore concentrations of Botrytis cinerea Fr. in otherwise unsuitable conditions at night and dur ing the day. Mills (1967) reported that with some rain periods accom panied by increased winds there was a marked increase in numbers of Ustilago avenae spores caught, but not with other rain showers which were accompanied by a smaller increase in wind. Sreeramulu and Vittal (1966) found that prolonged rainfall reduced the incidence of Ustilaginoidea  virens (Cke.) Tak. spores. From the varying evidence, and that of the present study, it would appear that gusty heavy rainstorms with their as sociated turbulence tend to increase spore concentrations by carrying more spores aloft through updrafts, after jarring the infected tissues which releases the spores, or through rain splash. On the-.other.liand, light gentle rains tend to wash the spores out of the atmosphere and wet the infected surfaces, which effectively prevents removal of spores- into the air. Weickmann (1963) pointed out that heavy storms with high rain fall rates are associated with atmospheric circulations of a high degree of organization and persistent updrafts, whereas weak storms lack the organized circulation and their updrafts are more transient. Davies (1959) demonstrated that wind-driven water droplets detach 6 to 335 times more Cladosporium spores than are detached by dry air moving at the same speeds, - 210 -and that the greatest effect was noted from wind speeds of 0.7 to 2.0 m/sec- Hirst and Stedman (1963) showed that the larger the rain drop size the larger the number of spores liberated into the air through me chanical shaking and by radial air movements. Chamberlain (1967) repor ted that the efficiency of capture by raindrops is high for particles of about 20u or higher, also he calculated (1953) that a rainfall intensity of 2 mm/hr will effectively reduce a 30M- particle concentration by about 65% after 30 minutes and by nearly 90% after one hour. The field results of Hirst (l953j 1959) indicated an even more rapid removal of spores from the air during a 30 minute 0.95 mm rainfall. Counts of pollen and.spores of Ustilago, Cladosporium, Alternaria and Erysiphe, before, during and after the thunderstorm, indicated the spore counts after rain were re duced to l/25, l/4, l/5, l/4 and l/ll respectively of the counts before rain, and that during the 30 minute rain, counts were increased between 9 and hh times (Hirst 1959). McDonald (1962) calculated that the "wash out probabilities" for pollen with a diameter of 20-26u, and a rainfall of 1 mm, was 99% f°r raindrop diameters of 0.2 mm, and 72% for diameters of 1.0 mm. The corresponding collection efficiences varied between 65-71% for 0.2 mm and 84-87% for 1 mm raindrops because of differences in pollen densities. Dingle and Gatz (1966) reported that heavy convective rain brought about a rapid decrease in the concentration of pollen. The effect of dew was minimal as it only tended to delay the daily initial dispersal of aeciospores. During periods of dew formation, wind velocity and turbulence are likely to be very low, thus if any spores are released, and evidence suggests that they are not generally dispersed under these conditions, deposition is likely to take place - 211 -within a meter of their point of release by sedimentation. Studies of effect of dew have usually teen undertaken in association with spores requiring wet periods for dispersal (Moore 1958; Hirst and Stedman I962; Carter 1963). For most dry-spore fungi, wind is an important factor in spore release, as they have no special active- mechanism by which they are li berated. Once the spores become exposed the number of spores shed at any time should depend on the varying ease with which spores are detached from preformed spore masses (Hirst 1953> 1959)- Particles of 30n diameter are not easily removed by air movements along (Chamberlain 1967), as any surface is surrounded by an air layer known as the laminar or surface boundary layer in which there is no turbulence, and in which air flows in streamlines parallel to the nearest surface (Gregory I96I; Tyldesley 1967). Any spore liberated into this surface boundary layer will sink to the surface by sedimentation. The major effect of wind is that of getting the passively liberated spores away from the surface where they are formed across the boundary layer of non-turbulent flowing air into the general circulation of turbulent air. The thickness of the laminar boundary layer varies with the wind speed and with the roughness of the adjacent surface, thus the higher the wind speed the more easily spores can be picked up or blown off the surface by eddies which force their way to the surface dis rupting the laminar layer. Rich and Waggoner (1962) showed that repeated jarring rapidly depletes a spore source, thus any increase of wind velocity which mechanically moves the source will bring about a rapid depletion of the mature detached spores. An initial gusty or higher wind speed pro bably produces a greater spore concentration than an initial slower wind - 212 -velocity, as a higher percentage of the spores would be passively liberated from the aecia at the outset of stronger winds than by weaker winds. Hirst (1959) stated that "there is little evidence that spore concentrations are increased in proportion to increasing wind speed, probably because spore clouds are rapidly dispersed by high winds". Both Zoberi (1961) and Smith (1966) showed a direct relationship between wind speed and spore liberation from experimental tests, and Zoberi also showed that dry air was more effective than damp air in dispersing spores. As mentioned ear lier, Mills (1967) indicated that rain accompanied by high winds brought about higher concentrations but low wind velocities did not, and the data of other workers suggests the same relationship of this combination of" factors. Sreeramulu (1962) was able to show that high concentrations of Ustilago nuda (Jens.) Rostr. spores often coincided with high wind veloci ties. However, as spores can also be dispersed by light winds, considera bly stronger winds occurring later are unlikely to appreciably increase the concentrations of airborne spores, as was noted by Jarvis (1962). Shanmuganathan and Arulpragasam (1966) indicated a negative correlation between wind speed and Exobasidium vexans spore concentration, which is the reverse of the expected physical relationship. Hamilton (1959) has also reported a decrease in concentration of some spores with increased wind. One feature that should be borne in mind is that the efficiency of trapping varies over the range of wind velocity (Hirst 1953) and it is difficult to compare adequately the performance of spores of different species in relation to wind velocity, as spores vary considerably in their impaction efficiency (Gregory 1961). - 213 -Spore dispersal In the present study no aeciospores were collected at distances greater than 400 feet from the source on pine. Samplers operated at a distance of 725 feet during two seasons failed to collect a single spore;. Wo doubt aeciospores were carried greater distances than 400 feet on some occasions due to diffusion of the spore cloud, but a suitable sampling network, free from background spore contamination, was not set up in the study area. Pennington and Snell (cited in Spaulding 1922) trapped the smaller C. ribicola aeciospore up to a distance of 550 feet, and on another occasion Pennington (cited in Spaulding 1922) caught aeciospores up to 1200 feet from pines. Tubeuf (19OI) stated that C. ribicola aeciospores spread the disease up to 500 meters or more. Snell (cited in Spaulding 1922) provided evidence that aeciospores of C. ribicola were carried at least 7 miles from pines on the mainland to infect Ribes on islands off the Wew Hampshire coast. There are several reports of Ribes infected at distances greater than a mile from the nearest known infected pine (Mielke 1943; Pennington 1924, 1925; Snell 1920; Spaulding 1922). Several workers used Lycopodium or fungal spores for experi mental dispersal studies using an array of samplers to measure the dis tribution of spore concentration from a point source (Gregory et al. I96I; Hodgson 1949; Sreeramulu and Ramalingam I96I; Stepanov 1935; Wilson and Baker' 1946), and others used ragweed pollen (Dingle et al. 1959; Hewson and Dingle 1956; Ogden et al. I963, 1964, 1966; Raynor and Ogden 1965; Raynor et al. 1966). Dispersion studies have also been carried out using gaseous or particulate releases (Allison et al. 1968; Chamberlain 1953; Hay and Pasqui11 1959; Record and Cramer I958; Singer and Smith 1966). - 21k -Some of the spore dispersion studies were conducted primarily to test dis persion theories, and these showed that the concentration decrease with distance from source resembles the decrease meteorologists have predicted from airbornematerial diffusing from a point or line source (Gregory 1°A5, I96I; Waggoner 1965; and others). When spores are released continuously from a point source the spore cloud is similar to that of a plume of smoke, which takes the form of a horizontal cone.with its apex resting at the point of spore liberation and its base oriented in the direction of the mean wind. As the spore concentration moves downwind in the turbulent air it diffuses horizontally and vertically, and at any point downwind its diameter is, on the average, proportional to the.distance it has travelled. Thus the conical form is achieved, except that in cross sec tion it has the shape of an ellipse/, since the turbulent dispersal is smaller in the vertical direction than in the horizontal one. Concentra tions remain greater around the horizontal axis of the "cone" than nearer its surface and concentrations become less and less with distance from the source. As the plume spreads vertically its lateral spread will be affected by the systematic variation of wind direction with height above ground (Pasquill I961). Generally in nature spores are released near ground level thus the base of the "cone" drags along the surface losing spores steadily by various types of deposition as it proceeds. Various formulae (Chamberlain 1953; Gregory 19"+5, I96I; Gregory and Stedman 1953; Pasquill I96I, 1962; Waggoner 1952, 1965) have been developed to describe dispersal or diffusion of spores and other particulate matter from a point or line source under a number of daytime conditions. The diffusion the ories do not apply at night because the vertical diffusivity goes to - 215 -zero under temperature inversion conditions. Evidence from eddy diffusion theory predicts rapid depostion near the source, and corresponds to ob served disease gradients- and experiments of spore deposition. In the for mulae of Gregory (1945) no account was taken of spore size, as it was assumed that the rate of fall of spores in still air was of little account in relation to the movements which occur within the eddies of a body of air in turbulent motion. Later Gregory (1952) admits that the velocity of fall of spores has some influence on the total distance of dissemina tion, however, the various diffusion formulae methods have not yet been fully adapted to include the effect of gravitational settling and de position (Tyldesley 1967). Recently this parabolic equation of diffusion was solved by an electronic analog computer which showed the effect of depletion of spore concentrations downwind from a line source (Brock 1962). Other numerical methods,"Jusing digital computers are now used in which set tling and deposition can be accounted for (Tyldesley 1967). Schrodter (i960) strongly criticizes the neglect of rate of spore fall as a factor in aerial spore dispersion formulae. He says "the velocity of fall is an extremely important factor in determining the range of flight and cannot be neglected in the problem of dissemination':' as "the1 gravitational fall is always going on" for "even in turbulent air when the net movement of the particle is upward gravitational fall continues". Hirst (1959) stated that the overall effect of sedimentation was important only in calm conditions or in sheltered places. Gregory (1961) reported that at windspeeds below 2 m/sec sedimentation under gravity plays a more important part in the spore deposition than turbulent flow over the surface. Tyldesley (1967) reasoned that the turbulence of the "frictional layer" which ascends to 500-1000 m (Sutton 1953), brings about vertical air - 216 -fluctuations of the order of 10 cm/sec which is at least as great as the fall speed of large pollen. Therefore the disorganized vertical motion of the air effectively inhibits:'settling and the air spora is essentially in suspension under the thermally unstable conditions of midday. Ludlam (1967) stated that if the "fall-speed is greater than any frequently recurrent upward air speed at the time and level of; injection, the par ticles return to the surface within a distance not many times the height at which they are injected". Waggoner (1965) concluded that extraction of spores from a spore cloud could be equal to or greater than the rate corresponding to their velocity in still air, and that extraction was most rapid when a source was embedded in a stand of plants. Velocities of fall have been calculated for spherical spores using Stokes' Law, and'Falck (1927), McCubbin (1944) and Schrodter (1954) calculated a relationship between spherical and other shaped spores. Schrodter (1954) used Falck's formulae to derive values for various groups of spores. Typical velocities for these spore.groups fall between less than 1 mm/sec to .7 cm/sec. Chamberlain (1966, 1967) gave the following values for 20, 30 and 50M- diameter spheres of unit density, 1.2, 2.7 and 7.1 cm/sec respectively. Stepanov (1935) obtained a value of 2.78 cm/sec for ellipsoidal spores of Helminthosporium sativum Pammel, King & Bakke, measuring 68 x 24u, and that a cylinder of such dimensions would approxi mately equal the volume of a sphere of 40u diameter, which would have an expected terminal velocity of about 4.8 cm/see, considerably faster than that actually recorded for the ellipsoidal spore (Gregory 1945). The aeciospore of C. comandrae is not spherical or ellipsoidal and the spore is not of equal density throughout, the "tail" being considerably less - 217 -dense. Ignoring the "tail" portion of the spore, its size would be approximately 23 x 3k\i and thus would act similarly to a sphere of 28p of equal density throughout. The average velocity of 3.23 cm/sec ob tained in the more reliable tests of the experiments is therefore probably an overestimate of the order of 0.5 cm/sec, especially when one adds the influence of the "tail", which gives a bulk density of less than unity. The overestimate is partly due to variability between tests and the pro blem of minimizing the effect of spore clumping and mass subsidence. More accurate results would probably be obtained by releasing a smaller spore sample into the cylinder. Others have explained the effect of a number of factors on velocity of fall which could account for some of the range of velocities found in the present study. Weinhold (1955) has shown that' variations in the rate of fall occur with slight changes in temperature and relative humidity. Falck (1930) stated that the speed of falling of dry spores differed from that of fresh spores because the loss of water caused changes in both weight and shape. Similarly, Buller (1922) showed that spores falling in dry air rapidly lost velocity owing to desiccation, and completely air-dried spores were seen to fall only one-third as fast as fully turgid spores, which corresponds approximately to the difference between dry and wet C. comandrae spores. Also, the storage of spores for long periods would probably affect the ability of the spores to respond to relative humidity changes, as indicated in the tests of Weinhold (1955) with uredospores of Fuccinia graminis tritici Erikss. & Henn. Harrington and Metzger (1963) have shown that ragweed pollen will take up moisture at humidities above 52%, increasing their density which affects their terminal velocity. Gregory (1945) stated that "exact prediction of the - 218 -terminal velocity of a spore is hindered by the difficulty of determining its density (which ... depends on its hydration), and by deviation from the ideal smooth spherical form". Chamberlain (1966) attributed the variation in reported values of the terminal velocity of Lycopodium spores to small variations in size of spore used. The larger the average dia meter the greater the terminal velocity. Chamberlain (1967) also mentions the phenomenon of buLk sedimentation which may take effect close to the source before turbulent dispersion has taken effect on the cloud of re leased spores and may operate to give an excessive apparent velocity, of deposition. This may affect the velocities of fall in some still air experimental tests, and he suggests it was possibly the reason for Gregory et al. (1961) finding a decreased velocity of fall with distance downwind where the released Lycopodium spores had undergone considerable diffusion. Christensen (l^kh), McCubbin (1918a) and Ukkleberg (1933) have determined the theoretical dispersal distances of spores based on the horizontal wind speed, altitude attained, and rate of velocity..of fall. Such calculations, however, have little practical value and no theoreti cal value, as they do not consider upward movements of air associated with atmospheric turbulence. Near the ground the structure of turbulence changes systematically with height (Pasquill 1961), and is composed of separate bodies — bubbles, shells or cells — of air which may move in predictable paths, and also specific local thermal updraft and downdraft patterns exist in association with local topographic or vegetational fea tures (Cone 1961, 1962;'Ludlam and Scorer 1953; Yates 1953). These struc tural atmospheric features may show marked diurnal variation, and because of this, random spore diffusion is. unlikely to occur over great distances - 219 -(Van Arsdel I967, Waggoner I965). Van Arsdel (1965, 1967) was able to show the persistence of such atmospheric features at night when conditions are more stable than under the more turbulent conditions of the day. From this he was able to predict the movement of spore clouds, and to relate this to the history of Cronartium ribicola rust infection over a period of 20 years. All the experiments and sampling of the present study supply evidence to support the findings of others (Buchanan and •Kimmey 1938; Gregory 19*+5> 1952, I96I; Gregory et al. I96I; Sreeramulu and Ramalingam I96I; Stepanov 1935; Wilson and Baker 19*+6), that under normal conditions a substantial percentage of the spores liberated near the ground will be deposited within a short distance of the source. Gregory (1952, and correction 1958) stated that "observations and theory agree that for spores or pollen liberated near the ground under normal conditions of turbulence, 90% °f "the spores will be deposited within 100 m of the source ... and under relatively calm conditions" spores would be deposi ted over "a still more limited range". Later he calculated (1961) that a spore cloud released at 0.1, 1 and 10 m above the ground would be de pleted by 90, 70 and <10% in 100 m of travel, which indicates the con sequence of elevation of source. Similarly, Waggoner (1965) calculated the maximum spore concentration on the ground would occur 33 22 and kO m downwind from sources that are 1, 6 and 10 m above the ground, as the maximum concentration is not found on the ground next to the source when the source is elevated. Also, when the source is elevated, there is less rapid depletion of the spore cloud. With release of C. comandrae aecio spores from sources 1 and 5 feet above ground, maximum spore concentrations - 220 -were respectively at the 5 and 20 foot sampling points from the source, although the actual peaks may have occurred at distances other than where the sampling points were placed. From the tests of natural disper sal patterns around a spore source one to two feet above the ground in the forest, the spore concentration in all directions from the source gave very similar steep gradients of deposition. Generally, deposition at 6 feet was less than 1% of the concentration at the first sampling point (3 inches) and reductions of 90% occurred within 2 feet. Wind speeds during the sampling averaged 2.20 to 2..4'9 ft/sec, and spore concen trations were slightly greater downwind from the source along the path of the supposed prevailing wind during the period, but the pattern of dispersal around the source indicated that the wind direction fluctuated widely. Some evidence exists of the gradient of dispersal of aeciospores of Fuccinia graminis, although these spores are considerably smaller (17 x 20u) than those of_C. comandrae. Lambert (1929) reported except ionally steep gradients of spore reduction from source. Spores were trapped for a period of 20 hours at 3, 6 and 23 feet from an infected barberry hedge, and gave counts of 160,000, 33,000 and 210 respectively. On a few days spores were trapped up to a mile from the bushes. Johnson and Dickson (1919) reported infection gradients from barberry bushes, and showed that the percentage of stems infected was 100% at 15 feet, 30% at 90 feet, 10% at 175-200 feet and 1% or less after 300 feet. These steep gradients of deposition were often interpreted as spores disseminated over only a short distance (Gregory 1968). This inference may sometimes be justified, but Gregory indicated that steep gradients may not always represent rapid deposition near the source, as diffusion of the spore cloud - 221 -should be considered. Spore dispersal in air is mainly controlled by eddy diffusion, therefore if no deposition took place the spore concentra tion would vary approximately in proportion to the distance. Spore cloud and surface deposition are inconstantly related, and unless the total effective spore emission is known, no estimates of the number of spores carried beyond the sampling area can be made (Gregory 1968). There is much evidence from aircraft and other points of collection remotely re moved from source areas, that pollen and spores of rust fungi and other plant pathogens are transported in the atmosphere over considerable dis tances, in some cases thousands of miles (Craigie 1945; Dillon Weston 1929; Hirst et al. 1967a,b; Kelly et al. 1951; MacLachlan 1935; Pady 1954; Stak-man'and Christensen 1946; Stakman et al. 1923). How can we solve the dil emma of reconciling the experimental evidence of apparent rapid deposition near the source with spore concentrations collected in the atmosphere, even over the oceans? The solution appears to rest with that fraction of spores which escapes deposition near the source and which is carried up by mechanical or thermal turbulence to greater heights in the atmosphere (Gregory 1963). Gregory (1962) suggests that this fraction may commonly be of the order of 10$. Expt. V in this study gave some evidence that spores were probably escaping from the source area. Spores liberated on warm dry days when turbulence is at a maximum, stand much.more chance of travelling long distances than those dispersed at night or in wet condi tions, and the smaller the spore the greater the chances for long distance dispersal. Waggoner (1965) indicated that spores often travel in a series of great loops on a clear turbulent day, and that towards sunset the spores would tend to be lofted. - 222 -Schrodter (1954, i960) showed that spores of approximately 20 n diameter and with a velocity of fall of approximately 1 cm/sec could be-carried 16 km under- a wind velocity of 2 m/sec and a mass exchange of 10 gm/cm sec. Gregory (1962), however, obtained values considerably less than Schrodter's using Chamberlain's (1953) modification to adapt to his. theory (Gregory 1961), only obtaining a probable flight range (distance at which 50% of spores liberated are still in suspension) of 50 m for large spores liberated in normal turbulence at a liberation height of 0.1 m, and only 4 m in low turbulence. More recently, Chamberlain (1966) calculated that 40-50% of the particles of 30n diameter released at 0.5 m above ground in neutral conditions could be transported 1 km, and 60% in unstable conditions. If the same particles are released at.10 m their median range is close to 10 km. Recently, Hirst and his colleagues (Hirst and Hurst I967, Hirst et al. 1967a, 1967b) have been able to assess long distance spore transport by measurements made from aircraft over the sea around Britain, but they have been unable to test the viability of the spores collected. Schrodter (i960) also discussed the altitude which spores of various sizes could attain in their flight. Spores of approxi mately 20u wereviunlikely to attain altitudes of more than 100 m unless extreme mass exchange or turbulence occurred. On June 30, 1965, an op portunity was taken to expose microscope slides during flights in fairly stable low turbulent conditions in an area to.the west and northwest of Edmonton at heights between 500 and 1500 feet. No C. comandrae aeciospores were observed on the slides, but smaller rust spores were present. Although there was no evidence available to indicate the presence of aeciospores in the air below the sampling height, the presence of only smaller-sized - 223 -spores at these levels may support the data of Schrodter (i960) that the larger spores are unlikely to be carried at such sampling heights under normal turbulence. However, Davidson (cited in Mielke 19^3) reported collecting C. ribicola aeciospores at each 1,000 foot level up to 5,000 feet (the highest level at which exposures were made) above a heavily infected area in British Columbia, but no information on the atmospheric conditions was given. Schrodter (i960) calculated that the flight dura tion of large spores at small mass exchange values near the ground is but a few minutes. For a spore of approximately 20u, a fall velocity of 1 cm/ sec, and a mass exchange of 10 gm/cm sec, flight duration is 2 l/k hours, and the same spore with a mass exchange of 50 gm/cm sec is 11 hours. The duration of flight and altitude attained is significant from an epidemiolo gical point of view in connection with the problem of viability of spores to be discussed in a later section. Hirst and Hurst (1967) report that the midsummer freezing level over the British Isles may be as low as 1 km and similar levels could be expected in the test area. Spores may be sub jected to considerable desiccation, also intensity of light is increased above the surface, both important factors affecting the viability. Hirst and Hurst (1967) also reported that the erosion at the base of vertical spore profiles seemed usual after long travel over both land and sea, thus spores carried highest during active convection will have the best chance of being transported further, and be subjected to-convection the next day. Spore dispersal not only involves flight but also landing or deposition. Generally speaking,, the same forces that are responsible for transportation through the air are also decisive for deposition of spores. The principle mechanisms involved are sedimentation under the influence of - 22k -gravity, impaction, including turbulent deposition, and rain wash-out, a factor discussed above. Sedimentation is only important in the absence of turbulence, or at wind speeds below 2 m/sec (Gregory 1961) where it plays a more important part in spore deposition than turbulent air flow over the surface. Deposition by impaction plays the greatest role in the landing of single spores. When air flows past an obstacle, the flow streamlines divide to pass either side, but particles in the air tend to move by their own momentum towards the object before they are in turn de flected by the wind, flowing around the obstacle before being impacted on the windward-side of the obstacle. Generally the larger the particle the greater the chance of impaction, as the efficiency of impaction in creases directly with spore size and wind velocity, and inversely with the width of the obstacle. A number of theoretical investigations of impaction efficiency on cylinders, spheres and other shapes have been summarized by Fuchs (196*+) and Green and Lane (196*+), and experimental in vestigations by Chamberlain (1966), Gregory (1961), .Gregory et al. (1961) and Gregory and Stedman (1953)- According to Gregory (1952) a spore size of 10n, the most frequent diameter of spores dispersed by wind, represents a compromise between the conflicting requirements of dispersal and depo sition. The large-spored leaf- and stem-pathogens appear as specialized impactors. Johnstone et al. (l9*+9) pointed out that the ability of a particle to penetrate among close vegetation is the inverse of its im paction efficiency. In close vegetation a high impaction efficiency, such as that of C. comandrae aeciospores, would reduce the chance of a spore getting very far from its point of release. In the forest the canopy restricts the exchange of air, and as Raynor (in Chamberlain 1967) has - 225 -pointed out, some spores will be lost by impaction in the canopy, but the effect of sedimentation in the comparatively still air below the cano py is more important. One factor associated with rain or leaf wetness, important for deposition, was that a wet surface tended to assist in the retention of spores at high wind velocities (Chamberlain 1967), but there was little difference at low velocities where sedimentation was more im portant than impaction. A wet or sticky surface probably reduces rebound or bounce-off.- Blow-off after deposition had been shown to be nil or negligible even under high wind velocities. Bagnold (i960) also showed that particles less than about 50p. diameter are not readily removed from surfaces because 'they are imbedded within the viscous boundary layer of air flow which protects them from the buffeting of turbulent eddies. How do the findings of the response of a spore released in the open relate to what happens to the spore that is released in the forest stand? The rapid depletion of the spore cloud as it travels from the source will be even faster in the forest. Below the canopy, wind speeds are greatly reduced, which will tend to reduce the distance the spore cloud can travel. Secondly, turbulence is reduced and spores will tend to settle out of the cloud, or spores will sediment directly due to ab sence of any considerable air movement. Depletion will then be very rapid. Thirdly, the relatively large spores will impact on the narrow needles and other vegetative objects which they encounter within the stand, further depleting the spore load. The narrow conifer needles will act as better impactors of the large spores than the small spores (less than 5u), and better than deciduous broad leaves in the forest stand. In a stand with an open trunk space with a high canopy the spores will - 226 -encounter less foliage and depletion should be slower'. These hypotheti cal features of the depletion of spores in a forest have been borne out by Ogden et al. (1964, 1966) and Raynor et al. (1966) in their recent stu dies of the diffusion of pollen emitted instantaneously from sources at various distances and at several heights upwind of a forest edge, and by studies by Allison et al. (1968) who released fluorescent particle tracers within a forest. Raynor et al. (1966) found that less than 1$ of the pollen released into the forest was still air-borne at 100 m, and that the sample released from a lower level was depleted first. They also found that only 50$ of the pollen released k-0 m from the forest edge ever reached the foist , but 90$ of that released 10 m away reached the forest. The pollen was depleted faster with a lower wind speed than with a higher one,vindicating that deposition was more important than impaction. Within the forest, the great decrease in wind speed slows the forward transport of the particulate cloud to the point where gravitational settling removes the bulk of the material. Allison et al. (1968) reported that a fluorescent particle cloud moved at l/lO of the outside wind speed immediately above the canopy of a dense "tropical type" deciduous forest. One interesting finding of their work was that the direction of cloud movement within the forest was, on the average, significantly to the left of the external wind by about 23°, and a similar relationship was shown by smoke puffs. Wilson and Baker (1946), who studied the spread of Sclerotinia taxa Ader. & Ruh. in orchards, showed that travel of spores through the branches of a tree increased vertical dispersion when com pared with travel across unobstructed terrain. Raynor et al. (1966) noticed that at the edge of the forest, spread occurred in the vertical - 227 -as well as the horizontal. Allison et al. (1968) also noted that verti cal diffusion was rapid within the forest in terms of downwind distance. If spores are carried from a forest stand into a clearing or out of the forest, then the reverse effects will occur, with spores be ing carried over.greater distances or depleted less rapidly because of the increase of wind velocity and turbulence. In fact, if thermals occur through heating of the clearing or open ground, spores may actually, be carried on updrafts to considerable heights before being deposited. - 228 -AECIOSPORE GERMINATION Germination is the initial stage in the development of a fungus mycelium from the spore, utilizing its stored reserves for metabolism. Spores will germinate in water or on very simple media, if other envir onmental conditions are satisfied. The germination process involves mor phological and physiological changes within the spore wall which transform the spore from a stage of low to one of high metabolic activity, and ex ternal changes involving protrusion and elongation of the germ tube from the spore wall. The process of germination is enhanced or limited by various ranges or combinations of environmental conditons. These environ mental conditions include the factors of temperature, humidity, light and form, type and hydrogen ion concentration of substrate media. Each of these factors influences germination in several important ways including the effect upon percentage germination, the time taken to germinate, the rate of germ tube elongation or growth, and the kind or form of germina tion. Spores generally germinate over a range of values for each factor. It is not always possible to state the optimum value with precision be cause of the interaction of several environmental factors together, and because of the influence of other variables. These variables include the • previous history of the spores, their age and inherent variability, as well as other factors largely connected with spore handling and germina tion techniques. Many studies have been made of the germination requirements for the urediospore and basidiospore states of rust fungi, but until very re cently, little information was available for the germination requirements - 229 -of rust aeciospores of forest pathogens and then rarely more than tem perature requirements (Doran 1919, 1922; Fergus 1959, Hirt 1937; Mains 1916; Siggers I9I+7; Spaulding 1922; Van Arsdel et al. I956). Since the commencement of this study, some information on the requirements for ger mination of Cronartium spp. aeciospores has appeared (Anderson and French I965; Nighswander and Patton I965; Peterson I968; Powell and Morf I966; Roncadori and Matthews I966; Walkinshaw I965, I968; Walkinshaw et al. 1967), but only very recently, has any information been published for C. comandrae (Krebill 1968c). The study on Peridermium stalactiforme and P. harknessii of the Cronartium coleosporioides complex (Powell and Morf I966) was a companion study to much of the present investigation on C_. comandrae, al though only involving temperature and pH requirements for aeciospore ger mination. Wo studies have followed aeciospore germination on a day to day basis, although Doran (1922) and Spaulding (1929) report on C. ribicola, and Krebill (1968c) on C_._ comandrae spore viability at infrequent inter vals in a season. This section of the study, first investigates, through labora tory tests, various environmental requirements for germination and assesses the effect of other variables on aeciospore germination, and secondly, follows the day to day variations in aeciospore germination from indivi dual aecial cankers and pustules throughout the spore production period. - 230 -FACTORS AFFECTING AECIOSPORE GERMINATION  Methods and Materials Spore material and methods of handling Dry aeciospores were collected by brushing them from indivi dual sporulating aecial pustules of the rust on lodgepole pine from a wide area in the Foothills and Rocky Mountains area of southwest Alberta be tween Robb (53°13'N, ll6°58'W) and Beaver Mines (1+9°28'N, 11H-°12'W). Where possible, spores were selected from recently ruptured aecia to give samples of uniform age, and with less contamination. Collection areas were all made in mountain valleys between 3,500 and -6,000 ft. elevation. Most of the test material was collected in the vicinity of locations 1, 2 and 3 (Fig. 38). Individual spore samples were screened through sterilized 100-and 200-mesh sieves to remove extraneous material. All collections were tested for viability by germinating on water agar before being used or placed in storage. Spores collected in I963, 1964 and I965 were stored for periods of from 1 day to 2 years in small glass vials plugged with cotton. The spore vials were kept continuously in a desiccator over a saturated calcium chloride solution, at a 5°C temperature. In later years the calcium chloride was not used for drying as it was found that spore viability was reduced by this method. Better viability was main tained from 1966 onwards by storing spores at a temperature close to 0°C with a high humidity. Good to fair germination was obtained from spores stored for two months, but germination was usually poor after six to nine months storage. Unless otherwise specified, the reported results were - 231 -obtained from fresh spores, or from spores stored for less than 10 days. Methods for exploratory experiments During 1963 and I96U various experiments were carried out of an exploratory nature, to establish techniques for handling and maintain ing germination tests, and finding suitable media and conditions for ger mination, as few guidelines were available in the literature for this rust spore state. In testing for the effect of each factor it is important that all other factors involved approach their optimum level and therefore are not limiting. To ensure that the conditions for regular tests would be satisfactory, several exploratory tests were carried out with temperature, light, humidity and pH conditions,' and further tests were carried out to find a suitable non-stimulatory- medium for germination. The time required for spores to initiate germination, and the period required for the ma jority of viable spores to germinate was also investigated, so that tests of sufficient duration could be established. Initial germination tests were carried out using a slight mo dification of Righter's (1939) germination method which enables a large number of samples to be tested under uniform conditions. Glass petri plates' were sterilized and a small amount of parafin wax was melted in the plate. On coding a series of germination chambers (holes) were pun ched in the wax surface with a sterilized cork borer. The holes were arranged in a series of eight numbered.rows each with five lettered holes. The experiment was first designed to place each spore sample replica at random in the plate but this was difficult for ease of operation and five replicas of eight samples were eventually used with each sample in a - 232 -numbered row. Distilled water was placed in each hole and spores were seeded according to a set plan. The plates were then placed in desicca tors in controlled temperature incubators. A high humidity level was maintained-by placing water in the bottom of the desiccators. This me thod became too cumbersome when large numbers were involved, or when information was required at set intervals. Counting of the ko samples sometimes took over an hour which kept the samples out of the test envir onment too long. The spores also tended to arrange themselves around the water meniscus in the hole which made counting difficult. The spores nearest the circumference of the hole germinated more rapidly than those nearer the center, thus this method was unreliable for determination of percentage germination at set intervals. In later tests small 5~cm dia meter plastic petri plates were used for germination tests in which spores were seeded onto a layer of distilled water or a layer of distilled water poured' onto a 2$ Bacto-plus 2$ Malt agar. Percentage germination was ex tremely low on such media and often the spores appeared to become saturated, lose their color and sink to the bottom of the water with no germination. Subsequently the use of a 0.3-0.5$ Difco Bacto water agar was found to improve germination. This medium gave satisfactory results and was used in all tests unless otherwise stated. Use of an agar medium obviated sink ing, post dispersal clumping of spores, and spore counting was easier on a solid surface. It was found that light and dark had no significant effect on germination, and free water or a saturated atmosphere appeared to be a germination requirement. Temperature tests showed some germination over the range 5 to 30°C with the optimum close to 15°C. A trace of germina tion was evident after 1-l/k hours, and most germination had taken place - 233 -by 2k hours, with little increase after k8 hours and nil after 96 hours. The germ tubes generally grew away from the surface of the medium dur ing the first few hours of growth, but eventually the germ tubes made contact with the medium with lysis often occurring. At this time the contents of the germ tube were concentrated towards the tip. Germination was tested at 20°C on a basic 0.3% water agar med ium (pH 6.5), the pH of which was adjusted through addition of a number of buffered and unbuffered solutions, singly or in combination, to find suitable non-toxic media for testing for effect of pH. The basic water agar medium was prepared and sterilized without the solutions, which were added in various concentrations to the melted basal medium just prior to pouring plates. The solutions used included acetic acid, calcium carbonate, calcium hydroxide, lactic acid, potassium acid phthalate, potassium hy droxide, potassium phosphate, sodium hydroxide and sodium phosphate. Wo germination occurred at pH k with any media, but some occurred with a pH approaching 9- Sodium appeared to inhibit germination, and potassium phos phate was unfavourable except at pH 6. Lactic acid was the least favour able of the acids used in the pH range 5 to 6, but no significant dif ferences were shown in the solutions used in the alkaline pH range. Cer tain media were unsuitable because they failed to maintain stability of pH over period of testing. Certain solutions were also tested for their molarity effect on germination percentage and on germ tube length. Gen erally germination improves with lower molarity (a molarity of 0.01 or less with most buffers) and germ tube elongation was greater. Best ger mination in all these tests occurred over the pH range 5>5 to 7.5, suggesting a gener.allmedium with a pH in this range would be suitable for - 234 -routine germination tests and for tests for other factors so that the hydrogen ion concentration would not be limiting. As the side effects of the buffers were unknown, a buffered solution was not used for the re gular germination tests. The technique developed for evenly dispersing a known quantity of spores over a test substrate was described earlier (Powell and Morf I966). Uniformity of spore dispersal density is recognized as a neces sity for quantitative evaluation of biological responses. The number of spores to be seeded onto the test substrate was controlled by adjusting the air flow from a pressure pump with a bleeder valve and by a timer at tached to the pump. Time tests varying between 2 and 8 seconds gave spore densities between 300 and 1600 per sq cm. The average percentage ger mination for three tests (30 fields at each time interval) was not sig nificantly different over this range of spore densities, but any effect of density on the rate of germination was not investigated. For the regular tests spores were seeded at a uniform density of 300 - 500 spores per sq cm. '. Methods for subsequent experiments A 0.3 - 0.5$ Difco Bacto water agar was used in all tests unless otherwise specified. Ten ml of agar was poured from a Cornwall continuous pipettor syringe into each 5-cm diameter plastic petri plate which was maintained, prior to use, in a walk-in refrigerator at 3 - 4°C and at 75$ relative humidity. Before seeding the spores, all water agar plates were equilibrated at the required temperature for the subsequent germination experiment, thus avoiding long equilibrium periods during the test,:: Hand ling of plates at temperatures other than those of the test conditions was - 235 -kept to a minimum. This was also the case when intermediate observations were made during a test; plates were removed from the test conditions, observed and counted, and returned immediately before another plate was counted. During peak periods, in order to obtain intermediate informa tion, often at hourly intervals, two or three people were employed in the counting, the' plates being observed in the same sequence, maintaining a constant time factor between observations.' The spores were incubated in the dark (other than in the light experiments) at various controlled tem peratures over the range 0° to 35°C (1"1°C). Usually a series at 5°C in tervals from 5 "to 30°C was used. The effect of humidity on germination was tested in an appara tus described by Scharpf (1964), where the humidity of the test chamber was maintained by passing air through a container of distilled-deionized water or a glycerine-water mixture which allowed accurate humidity adjust ment. A Honeywell relative humidity indicator (Minneapolis-Honeywell Regulator Co., Minneapolis, Minnesota), accurate to t 1% relative humidity, was used to check the humidity level of the air within the chamber, over the various glycerine-water mixtures used. The apparatus was allowed to stabilize before the experiment was carried out. The test spores were placed on a dry slide held in the chamber and were exposed for 24 to 48 hours under the test humidity and at a constant temperature of 15°C, the whole apparatus being run within an incubator. Following negative re sults the spores were seeded onto water agar inaa petri plate for 24 hours, or placed on filter paper saturated with a 1% siution of 2,3, 5~"triphenyl tetrazolium chloride (Allied Chemical Corp., New York, N. Y.) for 24 hours or longer in the dark (as tetrazolium is light-sensitive) to test 236 -for viability. As a further check on the results with the above method, a series of aqueous sucrose solutions was prepared at concentrations of 10, 30, 50 and 100$. These concentrations gave theoretical relative humidities of 99-85, 99-4-3, 99-04 and 98.03$ in small sealed humidity chambers at 20°C (Clayton 1942). Spores from two separate samples were brushed onto two cover glasses attached to a microscope slide. The slide was then held on supports just above the water or sucrose solution in small sealed glass jars, which were placed in the dark in a 20°C incuba tor for 48 hours. For a control, drops of distilled water were placed on the cover glasses and spores were brushed onto the water surface. The viability of spores giving negative results was tested as above, on com-pletiomof the test. The effect of hydration of .spores in a saturated humidity fol lowing storage was investigated prior to their use in germination tests. Hydration was accomplished by placing vials of spores or dry spores on a microscope slide, for 24 hours, on the shelf of a desiccator containing distilled-deionized water, placed in an incubator at 15°C. Won-hydrated spores served as control. Many spores are reported to swell prior to ger mination. This was investigated with both fresh and stored aeciospores. The same sample of spores were measured in the dry condition, immediately when placed' on water or Czapek-Dox agar, and after 1 - l-l/2 hours and later intervals on the media. Measurements were made with a micrometer at a.50x power. Hydrated stored spores were also measured by the same method to see whether spore swelling had occurred during hydration. To test the effect of light on germination two experiments were set up. In one, samples of spore collections were divided into three - 237 -portions, prior to being seeded on water agar plates. One plate was kept in the dark at 15°C, the second in a 15°C incubator with a 'daylight' fluorescent lamp as light source, and the third kept outside under natural light conditions, with a recording hygrothermograph nearby. In the se cond test, the"effect of various colored light wavelength bands on ger mination were compared. A series of General Electric 18 inch long, 1 inch diameter, 15-watt fluorescent lamps was set up in incubators kept at 15°C, with one incubator maintained without a light source as a control. The lamps were mounted about 12 cm above the test plates which were placed directly beneath the center portion of the lamp for 2k hours, where even light intensity was maintained. Spectral emission curves provided by General Electric Lamp Division showed that black light fluorescent lamps emit 250 and k&0 m\i with an.energy peak at 350 mu; blue between 325 and 665 mLi with peak at 525 mi-1; pink between 525 and 750+ mu with peak at 615 mp; and 'daylight' between 320 and 750+ mp with peaks at 475 and 575 mu.. The intensity of the light at the level of the test plates was measured in foot candles (ft-c.) by a Sekonic Studio Exposure Meter (Brockway Camera Corporation, New York). The respective ft-c. values were black — 5 ft-c; blue — 80 ft-c; green — 175 ft-c; pink — 75 ft-c; and 'daylight' — 125 ft-c. The intensities of light measured are not com parable as light sensors respond differently as the spectral wavelength distribution of incident light changes (Federer and Tanner I966), and no corrections were attempted. For hydrogen ion concentration tests over the range pH 3 to 10, the desired pH of the agar media was obtained by adding appropriate amounts of potassium acid phthalate, potassium phosphate, potassium dihydrogen - 238 -phosphate or boric acid buffers, and sodium hydroxide, sodium phosphate, potassium hydroxide and hydrochloric acid solutions to the melted basal agar media just prior to pouring plates. This method produced firmly set media at all pH values used. In one "test series over the pH range k to 8, two or three media of different combinations were used in the experiments for each pH value. In most tests the control medium was a 0.3$ water agar (pH 6.8 - 7-0), but in one test a 2$ Czapek-Dox agar was used (pH 7-0). Measurements of the adjusted media, before and after the experiments, with a Beckman pH meter, showed little pH change during the experiment (less that t 0.2). The resulting end pH of the media was used in all cases. In I967, when several poor germination results were obtained in the daily germination tests, ah experiment was set up to investigate the effect of some readily available commercially purified agar and broth media on germination, number of germ tubes and germ tube length. Ten media were prepared, which included Bacto water agar, potato dextrose agar, dextrose agar, malt extract agar and broth, Czapek-Dox broth, nutrient agar and yeast extract (Table XXXIV). The media were prepared according to procedures' given by the manufacturers (Difco Laboratories, Inc., Detroit, Michigan, for all media except Czapek-Dox broth — Fisher Scientific, Fair Lawn, New Jersey), except that the agar content was sometimes varied, and 1$ and in one case 3-5$ agar was added to the broths to firmly solidify the media. In 1968 the germinative responses of spores to sucrose as a carbohydrate source over the range 1 to 40$ were compared with germination on a standard Bacto water agar, following the superior results obtained with a Czapek-Dox broth agar, which itself contains 3$ sucrose. In 1966 the effect of host leaf extracts on germination was - 239 -tested by adding 5 to 25 gm of chopped Geocaulon and Comandra leaf material to 100 ml of 0.25% water agar. The leaf material was either added prior to heating of the water agar medium or added as the agar cooled but before pouring and solidification of the.plates took place. In 1967 the test was repeated by adding 5 gm of chopped leaves to 100 ml of 0.3% water agar when the agar was cool. The pH of the leaf extracts and water agar was recorded in both years. To check that the pH of the host leaves varied little throughout the season and that the values were similar to those used in the media, measurements were made of the pH of the expressed leaf juice of the plants at regular intervals, by methods similar to those used by Hurd-Karrer (1939)- Tbe effect of the presence of host leaves, was also investigated, by placing host leaves on the medium surface. In two tests 3 mm diameter holes were punched in the leaves; in another a fine spray mist was applied to lightly cover the leaves with moisture to see if this would enhance germination. Both young and old Comandra leaves were used in one test. In all tests to investigate the effect of any factor or treat ment, one or two control plates from each spore sample were employed. Samples of spores from two to ten cankers, usually five, were used for each series of treatments. In most treatments each test was repeated at least twice, and in some cases four or more times. Usually the spore samples used for comparative treatments were of the same age, collected the same hour of the day from one location, and had been subjected to identical storage and preparation conditions. When spores of different collection times or storage conditions were used, this is indicated in the text. Germination percentage for each treatment was the average of - 2k0 -all samples used (unless one was eliminated because of obvious contamina tion or because another factor affected the accuracy of the germination percentage). At least 300 spores were counted in each sample, the count being obtained by observing each spore' in a number of randomly placed 50x power microscope fields. When no germination was indicated by this method of counting, the whole area of the medium with spores was scanned as occasionally a few germinated spores could be found or a contaminant had affected germination in a definite sector of the medium. A spore was considered to have germinated when the germ tube length was equal to or longer than half the minor diameter of the spore. In as many treatments as possible, the number of germ tubes per spore, the length of the germ tubes, the frequency and type of branching, and the occurrence of plas-moptysis or other observed factors affecting germination were recorded. A compromise was necessary between the number of spores measured per sam ple and treatment and the time involved in measurements, but generally at least 10 spores selected at random in each sample were measured to obtain the average number of germ tubes or average length of longest or all germ tubes. Observations on germination percentage and germ tube growth were made at various intervals between 1 and 168 hours, but the inter vals of 6 and 2k hours were those normally used. Results Effect of temperature on germination A series of tests was carried out to establish the range and optimum temperatures for germination. In I965 three samples, all showing good-germination, were tested at 5°C intervals from 5 to 30°C (Fig. 79). Fig. 79- Effect of temperature on per cent germination of three Cronartium comandrae aeciospore samples after 2k hours on water agar (pH 6.8). Fig. 80. Effect of temperature on Cronartium comandrae aeciospore germination. Average per cent germination of five series and various numbers of replicas with three to five samples each, after 2k hours on water agar. - 2kl -- 2k2 -Germination occurred over the range 5 to 25°C, but was much reduced at 25°C. Optimum germination occurred at 15°C, and the percentage germina tion was still relatively high at 10°C. In I966 and 1967 further tests were run to confirm the optimum temperature and the upper and lower limits of germination with greater precision. Generally the later tests gave much lower- germination percentages but the pattern was similar with an optimum close to 15°C. Fig. 80 shows the average percentage germination of five separate series of.tests. Test series 1 is the average of six replicas each with k or 5 samples, for the temperatures 5, 10, 15, 17, 19, 22 and 28°C. The germination percentages were lower at temperatures above 15°C indicating that the optimum temperature for germination may be a little below 15°C as there was less reduction at 10° than at 17°C. Test series 2, with two replicas each using 5 samples over the tempera ture range 1 to 30 °C, gave some germination at 30°C, but only a trace at 1°C. Germination at 30°C was not consistent as only five samples gave any germination at this temperature. Test series 3, with two replicas, and 2 or 3 samples each, over the range 2 to 30°C, Test series k, a single test with k samples over the range 5 to 30°C, and Test series 5, a single test with k samples over the range 5 to 20°C, show the same percentage germination pattern. Although in the various series (Figs. 79 and 80) some germination was recorded over the range 1 to 30°C, germina tion was generally poor (< 10$) outside the range 5 to 22°C. The average number of germ tubes per spore which developed at the various temperatures is shown for four temperature series in Table XXIX. In all series the maximum number of germ tubes occurred at 15°C, usually with a marked reduction at temperatures above and below, with - 2h3 -Table XXIX. Average number of germ tubes per spore, and the range of the average for four test series at various temperatures. Temperature °C 5 10 15 20 25 30 Aver. no. of jerm tubes/spore 1.5 1.5 1-9 1.3 1.2 l.l* Range of the average in the individual series 1.1 - 2.7 1.1 - 2.7 1.5 - 2.7 1.1 - 1.7 1.1 - l.k 1.1* *Data from only one series fewer being produced at the higher temperatures. The high ranges, shown in Table XXIX, all occurred in one series, when there was no difference in the number of germ tubes produced at 5, 10 and 15°C, but the usual decrease at higher temperatures. The average of numerous other tests, when spores were tested only at 15°C, was 2.0 germ tubes per spore, a figure very close to that shown in Table XXIX for this temperature. One to six germ tubes were produced per spore, all emerging at the same.time through the spore wall, but the germ tubes elongated at unequal rates, and it was uncommon for more than one germ tube to develop to any length. What determines which germ tube will develop is unknown, but the two nuclei of the spore are, almost without exception, present in the developed germ tube. The average length of the longest germ tube is shown for three samples in'Fig. 8l, along with the average length of all germ tubes per spore after 2k hours at temperatures over the range 5 "to 25°C. The aver age length of the longest germ tube was greatest at 15°C, with little difference between lengths at the other temperatures, although at the Fig. 8l. Average length of germ tubes of three Cronartium comandrae aeciospore samples germinated on water agar (pH 6.8) after 2k hours at various temperatures. (a) Length of longest germ tube per spore. (b) Length of all germ tubes per spore. - ¥TS -- 24-5 -extremes (5 and 25°C) they are generally slightly shorter. When the lengths of all germ tubes are considered the relative positions are lit tle changed (Fig. 8l), because generally only one germ tube develops to any length. The degree of germ tube branching varied little with tem perature.. Usually the germ tube was irregularly branched with most of the branching occurring towards the tip of the germ tube. Branches were generally short at all temperatures, rarely more than 50M- long. There was considerable variation in the length of the longest germ tube at 15°C in the various test series, but at other temperatures growth after 2k hours varied less. The average length at 15°C ranged from 300u to 713u5 with an average value for the various temperature tests of around 400u. Germ tube elongation generally continued for 96 hours in vitro or until plasmoptysis of the germ tube occurred. Most growth occurs in the first 2k hours and growth beyond this period is much slower. Rupture of the germ tube occurred at the growing tip and the protoplasm spilled onto the agar media. On some occasions nearly half the germ tubes plasmoptyzed, and this phenomenon was more common at the optimum and supraoptimal tem peratures . The influence of temperature on the rate of germination is shown in Figs. 82 and 83 from two separate series of tests. The effect was tested over the temperature range 2 to 30°C, with counts and observations made at one or two hourly intervals up to 8 hours with a final count after 2k hours. In both series, better initial germination occurred at 20°C than at other temperatures after 1 hour, and this position was retained at 2 hours. Slowest germination occurred at 2 and 25°C with other tempera tures intermediate. The general percentage level of the final germination Fig. 82. Influence of various temperatures on rate of Cronartium  comandrae aeciospore germination after various time intervals on water agar. Fig. 83. Average rate of germination of three Cronartium comandrae aeciospore samples at different temperatures and after various time intervals on water agar. - 2hG -loo-. 60-, Temperature TIME (HOURS) - 21+7 -was usually reached after k to 5 hours, with very little increase in ger mination percentage at 2k hours. This was not the case with germ tube growth, which continued to increase up to 2k hours (Fig. 8U). Initial germ tube growth, as in other tests, was fastest at 20°C followed closely by growth at 10, 15 and 25°C, although growth at 25°C tended to slow down quickly. Growth at 2 and 5°C was appreciably slower, although final growth at 5°C was similar to that at 10 and 20°C. The growth curve tended to level off after about 6 hours, although growth at 25°C had levelled off earlier. The hourly rate of germ tube growth after the first 6 hours was less than half that of the earlier period, and remained fairly constant in this later period. Measurements of germ tubes after k8 and 72 hours growth indicated little increase in length from that at 2k hours. Often plasmoptysis of germ tubes occurred when in contact with the media, which effectively prevented any further growth. Also, the contents of the germ tube tended to collect at the tip of the tube and form a vesicle type structure (Figs. 28-30) which probably performs the function of an ap pressorium from which an infection peg would eventually be produced. Effect of humidity on germination Seven separate tests were run at a relative humidity of 100$ and one test at 97$, to establish whether germination would occur in the absence of a liquid medium. In all tests no germination was obtained. To check that viable spores were used in the tests, the spores were subse quently seeded onto a water agar petri plate and/or tested with a tetra-zolium. chloride solution. In five of the tests some germination (up to 20$) was obtained on the plates, and similar percentages of the spores turned various shades of red in the tetrazolium tests—the degree of color being Fig. 8k. Average length of the longest germ tube from three to five Cronartium comandrae aeciospore samples germinated on water agar at different temperatures and after vari ous time intervals. - 248 -360-i 0 4 8 12 16 20 24 TIME (HOURS) - 2k9 -largely due to the time the spores were in contact with the solution, proving that the test spores had been viable. Results of the tests with various sucrose solutions similarly gave no germination. However, if deposits of water were present on the test slide through condensation then germination would take place. As it is very difficult to measure and maintain with certainty a relative humidity of 100%, the conclusion reached from the above tests is that no germination took place below 100%, and it is doubtful if any germination occurred in a saturated 100% rela tive humidity, but if water was present germination occurred. Swelling of spores on liquid media Table XXX shows that all aeciospore samples swelled when placed on a Czapek-Dox agar. Similar results were obtained on water agar. The ' measurements indicated an increase in both aeciospore length and width, though the proportional increase was greater in the width dimension. The width of fresh spores in the dry state was greater than stored spores, but if stored spores remain viable, as was the case with Sample 3, the immediate increase when placed on agar was similar. However, there was no similar increase in dead stored spores (Samples h and 5)5 the increase in width only amounted to about 35% of the increase of viable stored spores, although on water agar the increase, was closer to 65%. That the width size increase took place almost immediately was shown by the little or no increase at the 1 - 1-1/2 hour measurements. Sample 3 was also measured at 2 and 3 hour intervals as germination was later in this sample, and showed virtually no size change. Samples k and 5 showed no change when measured after 3 hours nor after 3 days, thus the lower increase was not - 250 -Table XXX. Average length and width (in microns) of 25 aeciospores from 5 fresh or stored samples. Measurements were made on dry spores and when spores were placed on a Czapek-Dox agar medium. Spore sample Spore condition Dry When placed on agar After 1 - Its-hours on agar no. length (n) width (n) length (n) width (n) length (n) width M 1 fresh 56 26 65 37 72 - 37 2 fresh 6k 30 70 37 — — 3 stored 2 months 6o 20 66 36 70 36 k stored ik months 58 20 65 26 6k 28 5 stored 26 months* 59 21 63 27 *50 spores measured just a case of slow absorption. As a further check of the swelling dif ference between viable and non-viable spores, a portion of viable spores, from Sample 3 was exposed to plus 30°C temperature for a few hours and then placed on Czapek-Dox agar and measured after 1 hour and after 3 days.' Average measurements were 6k\i long and 28u wide, values typical for non viable spores, and no germination occurred. The degree of swelling was therefore associated with a spore being viable or non-viable. The in crease in length was more variable as in nature the 'tail' is quite vari able in length, and it is possible that most size increase occurred within the main spore rather than in the 'tail'. This aspect was not observed separately. If non-viable dry spores were hydrated prior to placing on a liquid medium, no swelling occurred, but viable spores did begin swelling during hydration. The taking in of moisture appeared to be a requirement - 251 -for germination to activate mechanisms within the spore, for swelling ceased when germ tube emergence began, and no germ tubes appeared with out considerable prior swelling. Following germ tube emergence, spore dimensions decreased by a few microns, indicating that some spore contents had moved out into the germ tube. Effect of hydration of spores on germination Six samples of dry spores were stored for 5, 6,17 and 38 days at -k°C. A portion of each sample was placed in a saturated humidity for 2k hours following storage to allow hydration to take place. This portion and a control sample of dry spores were then seeded onto petri plates of Czapek-Dox agar or O.jfo water agar and germinated in the dark at 15°C for 2k hours. Table XXXI gives the average germination percent- . ages of the 6 samples for the individual days of testing, followed by the germination range. Initially, hydration of spores improved the ger mination by 7 to 10% compared with stored non-hydrated spores. A similar increase (9-1% for 5 days and 6.9% for 6 days) was recorded on the water agar, although the actual germination percentages on this medium were lower. Following 17 days storage, there was no improvement in germina tion, in fact a decrease, and this became more marked after 38 days when the germination percentage of hydrated spores was less than half that of dry spores. This is shown very well by the range in percentage ger mination of the six samples. - 252 -Table XXXI. Average germination percentages for spores stored for various periods of time at -k°C, and then germinated, dry or after 24 hours hydration in a saturated humidity, on Czapek-Dox agar at 15°C. Days of Dry non- Hydrated storage hydrated spores spores 5 85.1(68.8-95.1) 94.6(91.3-98.1) 6 71.3(61.9-80.6) 78.7(61.2-95.4) 17 72.6(49.0-85.9) 65.3(32.6-82.5) 38 67.9(50.3-87.9) 30.2(16.7-38.8) Effect of light on germination Table XXXII gives the results of three series of germination tests in the dark and light. In all three series 'light' was obtained from a 15-watt 'daylight' fluorescent lamp, and in one test natural light was also used. The tests showed no significant difference between dark and 'daylight' in two series, but in Series 2, two of the 8 samples showed higher percentages in the light than the corresponding ones in the dark. The test with natural light was slightly lower in germination than the two controlled conditions. This may have resulted from the natural tem perature fluctuations, but even in this test only two of the eight samples gave germination percentages markedly different from the other tests. There was no appreciable difference in the lengths of the longest germ tube under the three conditions (Table XXXIl), and the number of germ tubes per spore and the form of branching was similar. Light and dark therefore appeared to have no significant effect on germination. In 1966 and 1967 various tests were run using a series of colored light lamps, and compared with control tests run in the dark. Sufficient - 253 -Table XXXII. Effect of light and dark conditions on percent germina tion and germ tube growth of three series of aeciospore. samples, germinated on Czapek-Dox agar at 15°C or under fluctuating outside temperatures for 2k hours. Series Wo. of samples Dark 'Daylight' Natural light Germination % 67.6 52.5 86.1 66.2 1+0.0 82.7 59-6 Average length of longest germ tube (u) 1+81 569 . 1+66 575 518 ---Temperature °C 15 15 7-26 Table XXXIII. Effect of dark and three colored light wave bands on percent germination and germ tube growth of three series of aeciospore samples, germinated at 15°C on Czapek-Dox agar for 2k hours. Series 1 2 3 1 and 3 Wo. of replicas 3 2 3 3 Total no. of samples 17 18 15 11 Germination % Average Peak wave length'of longest band germ tube (u) (millicrons) Dark 75-9 75-2 33.3 236 Green light 74.5 39-8 233 525 Blue light 81.9 89.6 1+2.9 252 kko Pink light 77-6 83.9 (365)1 615 •'-Data from only one test with 5 samples - 254 -incubators with adaptors for light lamps were not available to test the effect of the different color light wave bands together. In I967 one series was run using three colors and the control, but in all other series only two and the dark control were available. Table XXXIII summarizes the three series made in I967, Series 1 and 3 were run three times, and Series 2 twice.- In all three series the blue light gave slightly better germination than the others used, but the percentage germination of the colored lights was not markedly different from the dark .control. The results suggested that the blue light spectrum (440p) slightly favours germination, but there appeared to be no gradient from low millicron values to high, as in the one comparable series, the pink spectrum (615P) gave 3% greater germination than the green (5251-0. There was little difference in the length of the longest germ tube between dark and blue and green light. In Series 1 when pink light was compared with the others, germ tubes, averaged 75M- longer under the pink light than under blue and dark, with the greatest difference (l50u.) between, pink and green. The tests made in 1966 were less satisfactory mainly because percent germinations were low. In all these tests the same four spore samples were used and germinated on water agar at 15°C. The following colors were used — 'daylight', green, blue, pink and black —, two at a time with a dark control, and all possible combinations were repeated twice, which involved a total of 20 tests. The dark control gave the highest germination percentage half the time, similarly each of the colored lights was highest for half of their tests, except blue, which was only equal to the highest in one of its four tests. This was the opposite of the 1967 tests, but the range of the total percentages was - 255 -so small in the I966 tests that it was doubtful if this difference was significant. (in the four common tests the average percentage germina tion for blue light was within jf0 of the other light conditions). Effect of hydrogen ion concentration oh germination A 1966 test over the pH range 3 to 10 using three fresh spore collections indicated that germination after 24 hours at 15°C occurred over the range 5 to 8, but germination was very low at the extreme values (Fig. 85). Where two media were tested at one pH value (k - 8), the media giving the higher percentage germination was used. Optimum condi tions occurred at pH 6. Germination at 20°C on the same pH media series gave less than half the percentage germination recorded at 15°C. A test series was run in 1968 over the pH range 4.5 to 8 and used only combina tions of potassium acid phthalate, potassium dihydrogen phosphate and sodium hydroxide with a Czapek-Dox agar medium. Spores from two collec tions stored for k and 6 weeks at 0°C, were tested at 5, 10, 15, 20 and 25°C. Media with pH intervals of approximately 0.5 units were used to indicate better the optimum pH for germination. Fig. 86 shows the ger mination percentage over the pH range tested for spores incubated at 5, 15 and 25°C. Optimum germination occurred at pH 6.6 at 5 and 15°C, but was closer to pH 6.1 at 25°C, this was also the case at 10 and 20°C. At pH 4.5 germination only occurred at 10 and 15°C. At the other pH extreme (8.0) some germination occurred at all temperatures except 25°C which also recorded only a trace of germination at pH 7.5 and 7.7. A 1964 series of tests, using spores stored for six months germinated at 15°C on unbuffered and buffered media, gave germination over the range 5-1 to 8.8 with optimum germination between pH 6.6 and 7-3- Wo germination occurred on media with Fig. 85. Effect of hydrogen ion concentration on germination of Cronartium comandrae aeciospores after 2k hours at 15 and 20°C. Fig. 86. Effect of hydrogen ion concentration on germination of Cronartium comandrae aeciospores after 2k hours at 5, 15 and 25°C Spore samples were stored for k and 6 weeks at 0°C prior to use. - 257 -a pH between 4.0 and 5-0 in this test series. In all test series germ tube length was much reduced on the buffered pH media compared with the control unadjusted agar media (pH 6.7 - 7-0), and generally the percent age germination was lower. Similarly, the average length of germ tubes in all tests was reduced by 50%, or more, at the pH extremes for ger mination. Effect of substrate on germination a. Various agar media Table XXXIV indicates that there was considerable range in the effect of the various agar media on aeciospore germination. In the ex periment of June 21, I967, seven spore samples were used, and on July 4, three samples from the same cankers as of the earlier date. It was no ticeable that the average percentages were reduced at the later date, Table XXXIV. Average percentage and range of germination of two series of aeciospores after 24 hours on different media at 15°C. Media June 21, 1967 July 1+, 1967 7 samples 3 samples Average % % Range Average % % Range Czapek-Dox broth + % agar 93-3 87.5-98.3 63.5 51.5-73.4 Czapek-Dox broth + 3 .5% agar 86.0 50.2-97.7 46.1+ 27.3-75.3 Potato dextrose agar 50.3 15.1-89.8 35-3 22.0-61.3 Dextrose + 1% agar 35.1 8.2-78.1+, 30.9 IO.O-56.I Nutrient agar 52.3 20.8-90.1+ 18.2 9.2-32.3 Water agar 0.5% 36.9 5.0-81.5 27.7 6.9-63.5 Water agar 2% 44.7 2.8-91.1+ 15.8 9.0-28.2 Yeast extract + 1% §ar 16.5 l+.5-!+1.6 11.1+ 1.0-21.0 Malt extract agar 13.6 0.3-1+3.8 1.7 1.6- 1.9 Malt extract broth + 1.5% agar 1.1+ 0.0- 5.9 1.4 0.7- 2.5 - 258 -and this was even more marked if only the three samples were compared. The percentage germination showed the same trend on both dates with re sults on the two Czapek-Dox media much superior, followed by the two on the dextrose media. Results on the yeast and malt extract media were poorer than on the water agar. The range of germination percentages showed the great variation in response of the various cankers, and this response was not always constant with regard to the media. In a subsequent test seven of the media (Table XXXV) were used to test three spore samples for the effect of the media on the number of germ tubes per spore and the average length of the longest germ tube after 2k hours at 15°C. Ten or more spores were observed in each sample at each treatment. The average number of germ tubes per spores varied between 1.5 and 2.0 on the different media, with the lowest number on the malt extract and potato dextrose agars and the highest on the dextrose and Czapek-Dox agars. The average length of the longest germ tube varied between 95 and 286u, with the longest occurring on the water agar closely followed by those on the dextrose and potato dextrose agars and the shortest ones on the nutrient agar. It is of interest, that when the individual samples were compared on all media, the sample having the least number of germ tubes had the highest average length, and the sample with the highest average number of germ tubes the shortest average length of the longest germ tube. Generally only one germ tube developed to any length and two germ tubes from one spore were rarely of equal length. When the lengths of all germ tubes were considered the same relative positions were retained between samples and on the various media, although the total growth on the water agar was increased (Table XXXV"). - 259 -Table XXXV. Average number of germ tubes and lengths of germ tubes per spore for three samples on seven different media after 24 hours at 15°C. Aver. no. Aver, length (u) Aver, length (u) Media of germ of longest germ of all germ tubes/spore tube/spore tubes/spore Czapek-Dox broth + 1% agar 2.0 215 228 Potato dextrose agar 1.7 244 264 Dextrose + 1% agar 2.0 272 283 Nutrient agar 1.9 95 119 Water agar 0.5% 1-9 • 286 348 Yeast extract + 1% agar 1.8 154 172 Malt extract agar 1.5 149 163 In further tests the effect of 0.5% water agar and Czapek-Dox agar were compared. Percentage germination was always better on Czapek-Dox at all levels of germination. There was no significant difference in the average number of germ tubes per spore from individual samples on the two media, but the difference in the length of the longest germ tube was considerable. In four tests, germ tube length on water agar averaged 3l8p, and on Czapek-Dox agar l67P, after 24 hours. b. Addition of sucrose to media Two tests were carried out with various concentrations of su crose introduced into agar media to give an additional carbohydrate source for germination. In one series a range of sucrose solutions from 1 to 40% was added to Bacto water agar (pH 7.0) and gave pH values from 7-2 to 7.4. Germination percentage varied little from that of the control between 1 and 30%, but