UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The biological control of Centaurea diffusa lam. and C. maculosa lam. by Urophora affinis frauenfeld… Roze, Liga Dace 1981

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1981_A1 R69.pdf [ 13.15MB ]
Metadata
JSON: 831-1.0095186.json
JSON-LD: 831-1.0095186-ld.json
RDF/XML (Pretty): 831-1.0095186-rdf.xml
RDF/JSON: 831-1.0095186-rdf.json
Turtle: 831-1.0095186-turtle.txt
N-Triples: 831-1.0095186-rdf-ntriples.txt
Original Record: 831-1.0095186-source.json
Full Text
831-1.0095186-fulltext.txt
Citation
831-1.0095186.ris

Full Text

THE BIOLOGICAL CONTROL OF CENTAUREA DIFFUSA LAM. AND C. MACULOSA LAM. BY UROPHORA AFFINIS FRAUENFELD AND U. OUADRIFASCIATA MEIGEN (DIPTERArTEPHRITIDAE) by L i g a Dace Roze B.Sc. (Pint. Sc.) Uni v e r s i t y of Western Ontario, 1974 A Thesis Submitted i n P a r t i a l Fulfilment of the Requirments for the Degree of Doctor of Philosophy i n the Department of Plant Science We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May, 1981 © Liga Dace Roze, 1981 In present ing th is thes is in p a r t i a l fu l f i lment of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make it f ree ly ava i lab le for reference and study. I fur ther agree that permission for extensive copying of th is thes is for scho la r ly purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or publ icat ion of th is thes is for f inanc ia l gain sha l l not be allowed without my wri t ten permission. Department of The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date i i ABSTRACT The aim of t h i s study was to determine the e f f e c t s of the seed-reducing g a l l f l i e s , Urophora a f f i n i s Frauenfeld and U. quadrifasciata Meigen (Diptera: Tephritidae) on the rangeland weeds, Centaurea d i f f u s a Lam. and C. maculosa Lam. (Compositae; d i f f u s e and spotted knapweed, respectively) i n B r i t i s h Columbia, Canada. The gall-forming f l y larvae feed i n immature knapweed heads. Lar v a l d e n s i t i e s per head peaked i n 1977, f i v e and s i x years a f t e r the f l i e s were introduced against d i f f u s e and spotted knapweed, re s p e c t i v e l y . Average l a r v a l d e n s i t i e s above 0.5 and 1.0 per d i f f u s e and spotted knapweed head, r e s p e c t i v e l y , caused d i s t a l head abortion. Normally d i s t a l heads do not abort, but proximal heads may abort a"t anthesis: I t appeared that U. a f f i n i s caused d i s t a l head abortion by supe r p a r a s i t i z a t i o n of the heads. Proximal heads that would normally abort developed i n knapweed that had many superparasitized d i s t a l heads. Density-dependent mortality of f i r s t - i n s t a r larvae i n the p r e - g a l l stage probably r e s t r i c t s numbers of JJ. a f f i n i s . The two g a l l - f l y species attack the same part of the plant, and when U_. a f f i n i s f i r s t - i n s t a r l a r v a l d e n s i t i e s exceed 0.5 per head, U. qua d r i f a s c i a t a increase i s suppressed. U. a f f i n i s , which o v i p o s i t s i n younger heads than U. qua d r i f a s c i a t a , saturate o v i p o s i t i o n s i t e s before they reach the stage of maturity required by U. quad r i f a s c i a t a . These species are not e c o l o g i c a l homologues and U. a f f i n i s w i l l not displace U. qua d r i f a s c i a t a because U« quadrifasciata emerges one week e a r l i e r than U_. a f f i n i s , and when U. a f f i n i s d e n s i t i e s are high, U. quadrifasciata can survive i n large-headed plants during the f i r s t generation and c a p i t a l i z e on the heads produced by small-headed plants during the second generation when U. a f f i n i s d e n s i t i e s are low. i i i G a l l - f l y larvae not only reduce seed numbers i n attacked heads, but g a l l formation p u l l s nutrients from other parts of the plant, causing a further decrease i n the number of seeds i n unattacked heads and an increase i n the percentage of heads that remain undeveloped. V i a b i l i t y of d i f f u s e knapweed seed decreased by 30% accompanied by a loss of the waxy seed coat and dark-brown pigmentation. There was some loss i n v i a b i l i t y of spotted knapweed seeds, but the major e f f e c t on t h i s species was a 33% reduction i n average seed weight. The combined e f f e c t s of U_. a f f i n i s and U. quadrifasciata were an 80% reduction i n seed numbers on both d i f f u s e and spotted knapweed. Most of t h i s was caused by U. a f f i n i s . - Although the g a l l f l i e s reduce the number of seeds, they w i l l have a n e g l i -g i b l e e f f e c t i n slowing the rate of spread of d i f f u s e and spotted knapweed because of the extent of the i n f e s t a t i o n s and the ease of spread by seed. Experiments i n seeding showed that habitat and weather conditions determine knapweed germination and seedling s u r v i v a l i n new locations more than the density of seeds sown. Thinning t r i a l s on seedlings showed that density-dependent competition between seedlings regulated knapweed population densi-t i e s . I f a l l seeds were destroyed, eradication of d i f f u s e and spotted knap-weed i n f e s t a t i o n s would take over three years because both species can grow vegetatively f o r two or more years as rosettes before b o l t i n g . Spotted knapweed may b o l t and produce seed heads for as many as three years a f t e r the rosette stage. L i f e table studies in d i c a t e that the g a l l f l i e s attack knapweed too l a t e i n the l i f e cycle to be e f f e c t i v e . i v TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES v i LIST OF FIGURES v i i x STATISTICAL INFERENCES x i ACKNOWLEDGMENT x i i I INTRODUCTION A. BIOLOGICAL CONTROL OF WEEDS 1 B. KNAPWEED IN BRITISH COLUMBIA 2 C. BIOLOGICAL CONTROL OF DIFFUSE AND SPOTTED KNAPWEED 6 D. THIS STUDY 9 II KNAPWEED/GALL FLY SYSTEM A. LABORATORY AND GREENHOUSE STUDIES . i . D i f f u s e and spotted knapweed head production 10 ii.Urophora a f f i n i s and U. quadrifasciata a. Emergence and l a r v a l weights 19 b. Superparasitism of d i f f u s e knapweed by U. a f f i n i s 24 c. G a l l - f l y head-size preferences 29 B. FIELD STUDIES i.Synchrony of g a l l - f l y attack with knapweed phenology 37 i i . G a l l - f l y d e n s i t i e s on release s i t e s 49 i i i . T h e r o l e of heterogeneity i n d i f f u s e knapweed head s i z e on g a l l - f l y abundance ; 68 iv.Pa r a s i t i s m of d i f f u s e knapweed by f i r s t and second gener-ations of g a l l f l i e s i n 1977 74 v. E f f e c t of g a l l f l i e s on undeveloped heads 86 v i . E f f e c t of g a l l f l i e s on d i f f u s e and spotted knapweed seed weight, appearance and v i a b i l i t y 96 V Page vii.Survey of rangeland d i f f u s e and spotted knapweed seed y i e l d i n 1977 *. 102 v i i i . V a r i a t i o n i n seed weight and numbers i n a spotted knapweed population I l l i x . E f f e c t of g a l l f l i e s on knapweed seed number on the release s i t e s from 1975 to 1977 116 II I KNAPWEED DEMOGRAPHY A. EFFECTS OF SEED REDUCTION ON KNAPWEED POPULATIONS 135 B. SURVIVAL AND FERTILITY SCHEDULES 151 C. KNAPWEED SEED SOWING TRIALS 170 IV CONCLUDING DISCUSSION A. INTRODUCTION 187 B. REASONS FOR FAILURE OF THE GALL FLIES TO CONTROL KNAPWEED i.R e s i l i e n c e - o f knapweed 190 i i . Phenology of U. af f i n i s and U. quadrif a s c i a t a 191 i i i . L o n g e v i t y and aggressiveness of vegetative stages 192 i v . P l a s t i c reactions by knapweed to a decrease i n seed density. 194 v.Knapweed establishment from seed 195 C. FUTURE ASPECTS OF KNAPWEED CONTROL 195 D. RECOMMENDATIONS FOR BIOLOGICAL CONTROL OF WEEDS •. . 198 V LITERATURE CITED 202 v i LIST OF TABLES Table Page I B i o l o g i c a l control agents released i n B r i t i s h Columbia against d i f f u s e and spotted knapweed 7 II Comparison of U. a f f i n i s and U. qu a d r i f a s c i a t a l a r v a l weights 21 III E f f e c t of caging i n d i v i d u a l d i f f u s e knapweed heads with U. a f f i n i s 27 IV G a l l - f l y head-size preferences 32 V Mean d a i l y temperature (°C) and t o t a l p r e c i p i t a t i o n (mm) for the d i f f u s e and spotted knapweed release s i t e s 46 VI Average numbers of U. a f f i n i s per head on the d i f f u s e knapweed release s i t e from 1975 to 1978 54 VII Average numbers of U. a f f i n i s per head on the spotted knapweed release s i t e from 1975 to 1978 55 VIII Average numbers of U. qu a d r i f a s c i a t a per head on the d i f f u s e knapweed release s i t e from 1975 to 1978 59 IX Average numbers of U. q u a d r i f a s c i a t a per head on the spotted knapweed release s i t e from 1975 to 1978 59 X Comparison of average numbers qf U. a f f i n i s and U. quadri-f a s c i a t a per head and average head numbers per small- and la r g e -headed d i f f u s e knapweed plants on the release s i t e i n 1977 70 XI Comparison of average head and g a l l production during f i r s t and second f l y generations on the d i f f u s e knapweed release s i t e i n 1977 77 XII Estimate of f i r s t - and second-generation g a l l - f l y f ecundities on d i f f u s e knapweed i n 1977 79 XIII Comparison of g a l l - f l y attack of small- and large-headed d i f f u s e knapweed during f i r s t and second generations 81 XIV 1976 survey for undeveloped heads i n d i f f u s e knapweed populations 89 XV 1977 survey for undeveloped heads i n d i f f u s e knapweed populations 89 XVI 1976 survey for undeveloped heads i n spotted knapweed populations 90 XVII 1977 survey f o r undeveloped heads i n spotted knapweed populations 90 XVIII Climate and a l t i t u d e for surveys of d i f f u s e and spotted knapweed populations i n 1976 and 1977 94 Table XIX E f f e c t of g a l l f l i e s on d i f f u s e and spotted knapweed seed v i a b i l i t y and weight i n 1976 XX Comparison of germinated, non-viable and dormant seeds from d i f f u s e and spotted knapweed release s i t e s XXI Survey of d i f f u s e knapweed seed weight and y i e l d i n 1977 XXII Survey of spotted knapweed seed weight and y i e l d i n 1977 XXIII Expected and actual seed y i e l d s per p l o t on d i f f u s e and spotted knapweed release s i t e s , and the estimated reduction i n y i e l d from 1975 to 1977 XXIV Average numbers of d i f f u s e and spotted knapweed plants per p l o t fo r each thinning rate i n 1977 XXV Average number of d i f f u s e and spotted knapweed plants per p l o t for each s i z e class i n the spring of 1977, and t h e i r fates by the f a l l of 1977. Average head production f or each s i z e c l a s s i n the f a l l of 1977 XXVI Survivorship and f e r t i l i t y of d i f f u s e knapweed XXVII Survivorship and f e r t i l i t y of spotted knapweed XXVIII T o t a l d a i l y p r e c i p i t a t i o n (mm) for Kamloops and Westwold i n 1977 XXIX * Mean d a i l y temperature (°C) for Kamloops and Westwold i n 1977 XXX Comparison of Kamloops and Westwold seed sowing s i t e s XXXI Percentage dead d i f f u s e and spotted knapweed by the f a l l of 1977, of the number counted i n the spring of 1977, on the Kamloops, spring 1976 sowing t r i a l XXXII Comparison of d i f f u s e and spotted knapweed rosette establishment during the 1977 growing season for the Kamloops, f a l l 1976 sowing t r i a l XXXIII Average numbers of d i f f u s e and spotted knapweed plants produced by the f a l l of 1977 on the Westwold, f a l l 1976 sowing t r i a l , and the death rates during the 1977 growing season XXXIV 1976 and 1977 weather data f or Kamloops and Westwold LIST OF FIGURES Figure 1 Diffuse knapweed 2. Spotted knapweed 3 Diffuse and spotted knapweed seeds 4 U. a f f i n i s g a l l s i n a spotted knapweed head 5 Average numbers of d i f f u s e and spotted knapweed heads per plant appearing each four-day i n t e r v a l 6 Percent d i f f u s e knapweed heads i n developmental stages 7 Percent spotted knapweed heads i n developmental stages 8 Branching pattern and sequence of head appearance f or a t y p i c a l d i f f u s e knapweed plant 9 Knapweed head sizes a v a i l a b l e and accepted for o v i p o s i t i o n by g a l l f l i e s 10 Average numbers of heads i n young and old bud stages on d i f f u s e and spotted knapweed release s i t e s i n 1976 11 Average number of male and female g a l l f l i e s per branch on the d i f f u s e knapweed release s i t e i n 1976 12 Average numbers of U. a f f i n i s and U. quadrifasciata per p l o t on the d i f f u s e knapweed release s i t e i n 1977 13 Average numbers of heads i n young and old bud stages on the d i f f u s e knapweed release s i t e i n 1977 14 Average numbers of male and female g a l l f l i e s per s t a l k on the spotted knapweed release s i t e i n 1976 15 Mean r a t i o s of g a l l flies:heads on the d i f f u s e knapweed release , s i t e i n 1977 16 Relationship between the percentage dead f i r s t - i n s t a r larvae which died before g a l l formation and the average number of f i r s t - i n s t a r larvae per head f or f i r s t generations of U. a f f i n i s on d i f f u s e and spotted knapweed release s i t e s 17 Relationship between the average number of U. quadrifasciata and U. a f f i n i s f i r s t - i n s t a r larvae per head on spotted knapweed for f i r s t generations of f l i e s 18 Co r r e l a t i o n between the average number of U. qua d r i f a s c i a t a and U. a f f i n i s f i r s t - i n s t a r larvae per head on d i f f u s e knapweed for f i r s t generations of f l i e s i x Figure Page 19 D i s t a l head sup e r p a r a s i t i z a t i o n by U. a f f i n i s of d i f f u s e knapweed 71 20 D i s t a l head sup e r p a r a s i t i z a t i o n by U. a f f i n i s of spotted knapweed 71 21 Relationship between seed heads, undeveloped heads, superparasi-t i z e d heads, and average numbers of f i r s t - i n s t a r U. a f f i n i s and U. qua d r i f a s c i a t a per head a f t e r the end of the f i r s t adult f l y generation 88 22 Appearance of spotted knapweed seeds'from unattacked and attacked heads by-U. a f f i n i s and fusiform U. a f f i n i s g a l l s 98 23 Co r r e l a t i o n between the number of seeds per terminal seed head and s t a l k height f o r spotted knapweed 113 24 Regressions of the average number of seeds per seed head and time on the d i f f u s e and spotted knapweed release s i t e s 119 25 Relationship between the mean number of U. a f f i n i s f i r s t - i n s t a r larvae per head and the mean number of d i f f u s e knapweed seeds per seed head 120 26 Co r r e l a t i o n between the mean number of U. a f f i n i s f i r s t - i n s t a r larvae per head and the mean number of spotted knapweed seeds per seed head 120 27 E f f e c t of U. a f f i n i s on seed numbers per seed head on the d i f f u s e knapweed release s i t e i n 1975 123 28 E f f e c t of U. a f f i n i s on seed numbers per seed head on the d i f f u s e knapweed release s i t e i n 1976 123 29 E f f e c t of U. a f f i n i s on seed numbers per seed head on the d i f f u s e knapweed release s i t e i n 1977 124 30 E f f e c t of U. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1975 125 31 E f f e c t of U. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1976 126 32 E f f e c t of U. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1977 127 33 E f f e c t of U. quadrifasciata on numbers of seeds per seed head on the d i f f u s e knapweed release s i t e i n 1977 129 34 E f f e c t of U. quadrifasciata on d i f f u s e knapweed seed numbers per seed head i n Vernon, 1977 129 35 E f f e c t of U. quadrifasciata on d i f f u s e knapweed seed numbers per seed head i n Vernon, 1979 130 X Figure Page 36 Average d i f f u s e knapweed rosette diameter i n the spring and f a l l of 1977 for each thinning rate 144 37 The percentage of d i f f u s e knapweed rosettes i n each fate c l a s s r e l a t e d to spring s i z e classes 145 38 The percentage of spotted knapweed rosettes i n each fate class r e l a t e d to spring s i z e classes 145 39 The percentage of spotted knapweed rosettes which bolted success-f u l l y , unsuccessfully, or were arrested of the t o t a l number i n each s i z e class which attempted to b o l t . 146 40 Regressions of average s t a l k height and average numbers of seed heads per s t a l k on d i f f u s e knapweed rosette sizes 146 41 Average numbers of d i f f u s e and spotted knapweed small seedlings per p l o t i n 1977 159 42 Average numbers of d i f f u s e and spotted knapweed large seedlings per p l o t i n 1977 159 43 Cumulative numbers of d i f f u s e and spotted knapweed small rosettes per p l o t i n 1977 161 44 Diffuse and spotted knapweed regressions of the cumulative number of small rosettes produced per pl o t i n 1977 (Y) , where X = ' log^g(time) 162 45 Average numbers of d i f f u s e and spotted knapweed rosettes per pl o t i n the spring of 1977 for the Kamloops, 1976 sowing t r i a l 174 46 T o t a l number of d i f f u s e and spotted knapweed seeds produced per plot by the f a l l of 1977 for the Kamloops, spring 1976 sowing t r i a l 176 47 Average numbers of d i f f u s e and spotted knapweed large seedlings per plot i n the spring of 1977 for the Kamloops, f a l l 1976 sowing t r i a l 177 48 Average numbers of d i f f u s e and spotted knapweed large seedlings per plot i n the spring of 1977 for the Westwold, f a l l 1976 sowing t r i a l 179 49 Percentage d i f f u s e knapweed appeared of the number of seeds sown on the Westwold, f a l l 1976 sowing t r i a l 181 50 Percentage spotted knanweed appeared of the number of seeds sown on the Westwold, f a l l 1976 sowing t r i a l 181 x i STATISTICAL INFERENCES Throughout t h i s study numbers following - are standard errors of the means. V e r t i c a l l i n e s on graphs indi c a t e - one standard erro r . A l l s t a t i s t i c a l t e s t s were conducted at p^.05, unless otherwise stated. A l l two-way comparisons are two-tailed t - t e s t s , and a l l comparisons among three or more means were by analysis of variance as i n E l l i o t t , 1971 1. Where symbols are used, N = sample s i z e ; r = c o r r e l a t i o n c o e f f i c i e n t ; R q = reproductive e f f o r t ; x = mean and D.F. = degrees of freedom. 1 E l l i o t t , J.M. 1971. Some Methods for the S t a t i s t i c a l Analysis of samples  of Beiithic Invertebrates. Freshwater B i o l o g i c a l Association, S c i e n t i f i c P u b l i c a t i o n No. 25. x i i ACKNOWLEDGMENT My greatest thanks go to Dr. Bryan Frazer, my supervisor, for h i s help and d i r e c t i o n . I am indebted to Marina Devick and Laura Roze for t h e i r painstaking work. I am g r a t e f u l for the help and i n t e r e s t of the s t a f f at the Agrigulture Canada Research Station, i n Kamloops, and use of the f a c i l i t i e s at the Station. P a r t i c u l a r l y I thank Dr. D. E. Waldern, Dr. A. McLean, Dr. W. Majak, B i l l Hubbard and Imants Bergis. I also thank the members of the B r i t i s h Columbia M i n i s t r y of A g r i c u l t u r e , Roy S. Cranston and A l f H. Bawtree for t h e i r ready assistance, I thank Tom Wallace and other personnel at the B r i t i s h Columbia Forest Service i n Kamloops, and i n M e r r i t , and Mr. Ed Smith i n V i c t o r i a , for th e i r support. I am indebted to ranchers, William Bostock, Hugh Robertson, Larry Campbell and Larry Buff for granting protected sections of t h e i r rangeland for f i e l d studies. Special thanks to Dr. P. Harris for advice during the early part of t h i s research. I thank Dr. J . D. Shorthouse, Dr. H. ZwHlfer, Dr. A. K. Watson and Dr. D. Berube for t h e i r advice and information. I appreciate the assistance of the remaining members of my committee, Dr. A. J . Renney, Dr. V. C. Runeckles, Dr. J . Myers, and e s p e c i a l l y Dr. W. G. Wellington. This work was supported by the Bostock.Grant, the L. Klink Fellowship, the B r i t i s h Columbia Forest Service, and i n part by a National Research Council of Canada Grant to Dr. J . Myers. 1 I INTRODUCTION A. BIOLOGICAL CONTROL OF WEEDS B i o l o g i c a l c o n t r o l of pests i s not a new concept. The ancient Chinese used the ant, Oecophylla smaragdina, to con t r o l c a t e r p i l l a r s and beetles i n t h e i r c i t r u s groves (DeBach, 1974). The f i r s t known deliberate introduction of a b i o l o g i c a l c o n t r o l agent by man occurred i n medieval Arabia, when date growers c o l l e c t e d predatory ants i n the nearby mountains and transported them to oases. These were released against phytophagous ants, which attacked the date palm (van den Bosch and Messenger, 1973). B i o l o g i c a l c o n t r o l i s recognized to be part of "natural c o n t r o l " (DeBach, 1974; van den Bosch and Messenger, 1973; Huffaker et al'. , 1971). Natural control i s described by Huffaker e^ t a l . (1971) as " a l l factors of the envi-ronment d i r e c t l y or i n d i r e c t l y i n f l u e n c i n g n a t a l i t y and mortality or move-ments into and out of the population". However, one component of the envi-ronment of an organism, such as i t s natural enemies, may be able to regulate i t s numbers. This i s termed " n a t u r a l l y occurring b i o l o g i c a l c o n t r o l " . In contrast to n a t u r a l l y occurring b i o l o g i c a l c o n t r o l , where natural enemies are indigenous, " c l a s s i c a l b i o l o g i c a l c o n t r o l " i s the process whereby natural enemies of target pests are discovered and imported to con t r o l the pest (DeBach, 1974). The usual procedure i s to search f o r sui t a b l e agents i n the country of the o r i g i n of the pest. Agents are introduced without t h e i r natu-r a l enemies a f t e r screening tests have been conducted to ensure that they w i l l not be harmful to b e n e f i c i a l organisms. In the future, recognition of natu-r a l l y occurring b i o l o g i c a l c o n t r o l should equal, or even surpass, e f f o r t s i n c l a s s i c a l b i o l o g i c a l c o n t r o l (van den Bosch, 1971). 2 B i o l o g i c a l c o n t r o l of pests has been shown to be far cheaper than chemical co n t r o l (Harris, 1971; van den Bosch and Messenger, 1973). Although DeBach (1974) estimated that only 54% of a l l target pests were s u c c e s s f u l l y con-t r o l l e d with b i o l o g i c a l control agents, these successes have been specta-cular. A widely p u b l i c i z e d case of weed con t r o l involved two c a c t i , Opuntia  inermis and (). s t r i c t a , which occupied over 24.3 m i l l i o n ha i n A u s t r a l i a . The c o n t r o l of Hypericum perforatum (Klamath weed, or St. John's wort) by the beetles Chrysolina quadrigemina and C_. h y p e r i c i , was spectacular on 0.81 m i l l i o n ha i n C a l i f o r n i a (van den Bosch and Messenger, 1973) but the beetles have been les s successful i n B r i t i s h Columbia (Harris et_ a l . , 1969). The co n t r o l of the aquatic a l l i g a t o r weed (Alternanthera phylloxeroides) i n some parts of the United States by the f l e a beetle, Agasicles spp. (Andres and Goeden, 1971) shows that t h i s method i s not r e s t r i c t e d to t e r r e s t r i a l weeds. B. KNAPWEED IN BRITISH COLUMBIA B i o l o g i c a l c o n t r o l of the Composite weeds, Centaurea d i f f u s a Lam, and C^. maculosa Lam. ( d i f f u s e and spotted knapweed, respectively) was begun i n B r i t i s h Columbia because these weeds were threatening the 1.1 m i l l i o n ha of rangeland i n the province, and were spreading into parts of Alberta and Saskatchewan. The annual loss to the B r i t i s h Columbia C a t t l e Industry due to a reduction i n forage q u a l i t y by knapweed was estimated to be $350,000 per annum i n 1972. Further estimates suggested that allowing the weeds to spread unchecked to t h e i r geographic l i m i t s would ultimately increase spraying costs to $33 m i l l i o n . Picloram (Tordon 22-K) applied at a rate of 0.42 to 0.56 kg per ha i s the recommended treatment. Although the cost of the chemical was only $2.84 per ha i n 1976, the expense of providing spraying 3 equipment and manpower raised the cost to an average of $20.75 per ha. Where spot spraying was necessary to check small i n f e s t a t i o n s , the cost was as high as $398 per ha (Harris and Cranston, 1979). Diffuse and spotted knapweed are natives of Eurasia, and are thought to have entered the province as contaminants of a l f a l f a seed. Spotted knapweed was f i r s t reported i n V i c t o r i a i n 1893, and d i f f u s e knapweed was f i r s t - seen i n Washington i n 1907. These weeds w i l l r e a d i l y invade any s o i l i n the I n t e r i o r of B r i t i s h Columbia with a disturbed "A" horizon. (Watson1972) .. The • •• "Interior", "refers to an-uplifted plateau stretching from l a t i t u d e 54°N, j u s t north of Prince George, south to the i n t e r n a t i o n a l border between the Coastal Range and the Columbian Mountain System. On the average t h i s area i s 200 miles wide, and i s drained by the Ni c o l a , Okanagan, Thompson, C h i l c o t i n and Fraser watersheds. About 10% i s open grassland, i . e . , about 1,417,500 ha, and the re s t i s timber range, sub-alpine meadow and alpine tundra. (Brink, pers. comnw) I t i s the open grassland that knapweed r e a d i l y invades. In 1972 d i f f u s e knapweed was found as f a r north as the 51°N. Lat., and was spreading northward to i t s approximate geographical l i m i t , 53°N. Lat. Spotted knapweed, which grows more r e a d i l y i n cooler, moistor habitats, was predicted to spread even further north. (Watson, 1972). Spotted knapweed may be found at a l t i t u d e s as high as 1,200 m, but d i f f u s e knapweed generally grows only as high as 900 m (Watson and Renney, 1974). Di f f u s e knapweed i n f e s t s about 7.5 times the area occupied by spotted knapweed (Harris and Cranston, 1979). These weeds spread by seed d i s p e r s a l . Increase i n v e h i c l e t r a f f i c during the 1950's was l a r g e l y responsible for the spread of d i f f u s e and spotted 4 knapweed from only a few hectares to the estimated 25,953 ha infested by d i f f u s e and 3,410 ha infested by spotted knapweed i n 1972 (Brink, pers. comm.; Watson and Renney, 1974). A possible cause f o r the rapid deve-lopment of knapweed monocultures i s t h e i r secretion of a l l e l o p a t h i c com-pounds. Fletcher and Renney (1963) demonstrated the i n h i b i t i o n of germin-ati o n of barley and lett u c e by knapweed exudates i n the laboratory. Majak (pers. comm.) found that knapweed le a f l i t t e r had a greater i n h i b i t i n g e f f e c t on barley seed germination than did fresh knapweed le a f material, and that germination of a l f a l f a , barley and rye were i n h i b i t e d by leachates of s o i l i n which knapweed had been growing. Dif f u s e and spotted knapweed germinate either i n the spring or i n the f a l l , and produce a many-leaved rosette during the f i r s t season of growth. A "rosette" i s a c i r c u l a r c l u s t e r of leaves (Benson, 1959), and knapweed rosettes crown tap roots. In the Kamloops area rosettes b o l t i n May, pro-ducing si n g l e s t a l k s which bear composite heads, or "capitulae" (Marsden-Jones and T u r r i l l , 1954). Unopened heads w i l l be ref e r r e d to as heads i n the bud stage i n t h i s study. Some heads i n the bud stage abort at early stages of development - usually when the head i s less than 4 mm i n length. In B r i t a i n almost a l l heads i n the bud stage can abort on some knapweed plants (Marsden-Jones and T u r r i l l , 1954). Heads which have opened to expose the f l o r e t s w i l l be referred to as heads i n the flower stage. The spiny ^diffuse knapweed heads (Fig. 1) produce white or pink f l o r e t s , but some of the plants have purple f l o r e t s . Spotted knapweed heads.".usually have mauve f l o r e t s , and spotted knapweed derives 5 Figure 1. Diffuse knapweed Figure 2. Spotted knapweed 1 f t i l l f c. d i f f u s a » I I I ! t c. mac ul os a Figure 3. Diffuse and spotted knapweed seeds (magnified 2X) 6 i t s name from the black-tipped bracts, which subtend the receptacle, giving the heads a spotted appearance (Fig. 2). Opened-heads, with withered f l o r e t s and f e r t i l i z e d ovaries wiH""be> termed seed heads. Spotted knapweed produces about twice as many seeds (termed' "cypcelas" by Marsd'en-Jones and "T,urrill,:l954)' per -head (26.6) as does d i f f u s e knapweed (12.5). On the average d i f f u s e and spotted knapweed produce 832 and 372 seeds per plant, r e s p e c t i v e l y , on rangeland (Watson, 1972). Seeds of some Centaurea spp. have b r i s t l e s which spread on drying, thus fo r c i n g the seeds out of the heads (Harper, L o v e l l and Moore, 19^70). Spotted knapweed has many such b r i s t l e s , and seeds-of-this species are shed soon a f t e r maturation, i n August. Diffuse knapweed has very small b r i s t l e s , or they may be absent, and because the long, narrow heads often do not dehisce u n t i l the winter, seeds i n d i f f u s e knapweed heads may be retained u n t i l the following spring. Diffuse knapweed seeds are smaller (1.099 mg) than spotted knapweed seeds (1.778 mg; F i g . 3). The highest germination % for seeds of both species i s on the s o i l surface. (Watson, 1972) Diffuse knapweed usually dies a f t e r maturation, but spotted knapweed produces numerous rosettes at the bases of mature s t a l k s , each of which can bolt i n following years. These rosettes, however, do not break o f f from the main root and grow as separate plants. The production of rosettes at bases of mature s t a l k s , and t h e i r subsequent b o l t i n g w i l l be termed perennation. A more d e t a i l e d account of knapweed biology can be found i n Watson and Renney (1974). C. BIOLOGICAL CONTROL OF DIFFUSE AND SPOTTED KNAPWEED Two seed-reducing g a l l f l i e s , Urophora a f f i n i s Frauenfeld and U. quadri- f a s c i a t a Meigen (Diptera:Tephritidae) were introduced into several l o c a t i o n s i n B r i t i s h Columbia i n an attempt to stop the spread of d i f f u s e and spotted knapweed, and to reduce e x i s t i n g i n f e s t a t i o n s (Table I ) . Both species r a p i d l y became established. The following i s summarized from ZwBlfer (1970) unless otherwise noted. Table I. B i o l o g i c a l c o n t r o l agents released i n B r i t i s h Columbia against d i f f u s e and spottzed .knapweed,, (Data..from.Harris, 1980a) knapweed species i n B.C. agent 1 country . ,. indigenous ... host o r i g i n l o c a t i o n date , > - estab-numbers . , of . ,of .. , lishment T released , N release release (status) diffuse knapweed U.a. France C. maculosa *••• P r i t c h a r d 1970 284 yes "A" 1971 209 yes diffuse knapweed R ussia C. s t e r i l i s P r i t c h a r d , ,„ n Bn 1972 797 yes diffuse knapweed * * Grand ± g u ^ Q Q Forks J diffuse knapweed U.q. Russia C. s t e r i l i s P r i t c h a r d „ B„ 1972 25 yes spotted knapweed U.a. France C. maculosa Chase 1971 97 yes spotted knapweed TT , 1970 297 yes Walachm 3 1971 280 yes ''"U.a. = U. af f i n i s ; U.q. = U_. quadrif a s c i a t a P r i t c h a r d "A" and "B" s i t e s are 265 m apart, on the Bostock ranch A** a c c i d e n t a l l y burned i n 1976; a c c i d e n t a l l y sprayed i n 1975 IJ. a f f i n i s and U. quadrifasciata females lay eggs amongst bracts of knapweed heads i n the bud stage. Copulation may occur within a few hours a f t e r emer-gence, but o v i p o s i t i o n does not occur u n t i l the second or t h i r d day a f t e r emergence. A l l heads i n the bud stage which are offered are probed by fe - r males i n the laboratory, however,1 only heads at p a r t i c u l a r stages of deve-8 lopment are accepted f o r o v i p o s i t i o n . U. a f f i n i s o v i p o s i t i n heads at younger stages of development than U. qua d r i f a s c i a t a. U. a f f i n i s f e -males lay about 120 eggs during t h e i r l i f e t i m e s under optimal conditions. If few heads are a v a i l a b l e f o r o v i p o s i t i o n , the eggs l a i d per head are not r e s t r i c t e d by the same, or other females. F i r s t - i n s t a r larvae chew into f l o r e t s or the immature ovaries. G a l l s are formed around the larvae. U_. a f f i n i s g a l l s are t y p i c a l l y hard and woody (Fig. 4), whereas the smaller JJ. q u a d rifasciata g a l l s are t h i n and papery. Both species feed on n u t r i -t i v e t i s s u e produced i n s i d e the g a l l s (Shorthouse, pers. comm.) and have three i n s t a r s . Figure 4. U. a f f i n i s g a l l s i n a spotted knapweed head Preliminary f i e l d work i n B r i t i s h Columbia indicated that most larvae pupate i n May a f t e r diapausing in s i d e the g a l l s during the winter. Adults emerge i n June and attack knapweed heads. The progeny of the f i r s t adult generation develop to the t h i r d l a r v a l i n s t a r by early August, when a proportion of the larvae pupate and emerge as the second adult generation. ZwBlfer (1970) found that i n Europe some f i r s t - g e n e r a t i o n IJ. a f f i n i s progeny also pupate i n mid-summer. Second adult generations emerge i n l a t e August. 9 D. THIS STUDY G a l l f l i e s and knapweed were studied mainly at Pritc h a r d and Chase, where the f l i e s had been released during the early 1970's. On the d i f f u s e knapweed release s i t e (Pritchard "B"; Table I) U_. a f f i n i s and U_. quadrif a s c i a t a were released together i n 1972, and on the spotted knapweed release s i t e (Chase) TJ. a f f i n i s was released alone i n 1971. Both s i t e s are i n the South Thompson River V a l l e y , approximately 20 km apart. The d i f f u s e knapweed release s i t e i s nearest Kamloops, at a lower elevation (366 m) than the spotted knapweed release s i t e (671 m). The d i f f u s e knapweed r e -lease s i t e i s hotter and d r i e r than the other s i t e . The d i f f u s e knapweed release s i t e i s a r e l a t i v e l y f l a t area of about 300 X 400 m, and was protected from l i v e s t o c k u n t i l the winter of 1977/1978, when i t was severely trampled by horses. In 1976 and 1977 the knapweed surrounding the s i t e was sprayed from the a i r , creating an i s l a n d of knawpeed i n the middle of pasture. G a l l f l i e s on the spotted knapweed release s i t e were released half-way up the crest of a ridge that was about 0.5 km long and 50 to 100 m high. The south-facing slope was a s o l i d i n f e s t a t i o n of spotted knapweed, but the north-facing slope was forested. The g a l l f l i e s spread over the e n t i r e infested ridge. In 1977 the area below the ridge was sprayed for knapweed cont r o l , and grazed, but there was no trampling by l i v e s t o c k within a 40 m radius around the release point of the f l i e s . On both release s i t e s the f l i e s increased exponentially i n population s i z e and i n 1977 population estimates reached 1 m i l l i o n (Harris, 1980a). By 1975 U. qu a d r i f a s c i a t a spread from the d i f f u s e knapweed release s i t e to the 10 spotted knapweed release s i t e , so that when t h i s study began i n 1977, both species were monitored on both release s i t e s . The aim of t h i s thesis i s to determine what e f f e c t s the g a l l f l i e s have had on d i f f u s e and spotted knap-weed r e l e a s e - s i t e populations from 1975 to 1978, and to predict the u s e f u l -ness of the f l i e s as seed-reducing b i o l o g i c a l c o n t r o l agents i n the I n t e r i o r of B r i t i s h Columbia against d i f f u s e and spotted knapweed. II KNAPWEED/GALL FLY SYSTEM A. LABORATORY AND GREENHOUSE STUDIES 1. Diffuse and spotted knapweed head production: 1.Introduction: Phytophagous insects may a f f e c t plant phenology. For example, Cameron (1935) found that damage to ragwort by T y r i a jacobaeae could cause shoot regener-at i o n and flower production l a t e i n the growing season. The present study was undertaken to define the growth pattern of knapweed when attacked by the head-feeding g a l l f l i e s . In t h i s study the term "head i n i t i a t i o n " r e f e r s to the appearance of new heads on knapweed s t a l k s , and "perennation" r e f e r s to rosette production at the bases of maturing s t a l k s . 2. Method: Diffuse and spotted knapweed rosettes were dug from the Kamloops area at the beginning of November, 1976, were transplanted into pots, 15.2 cm i n diameter and placed under a 16-hour photoperiod i n a greenhouse. When most of the rosettes had bolted (after about one month), three d i f f u s e and three spotted knapweed plants, which had produced t h e i r f i r s t heads on the same day, were a r b i t r a r i l y chos en. Thereafter the s i x plants were observed every 2 days - 3 hours for the appearance of hew heads, when heads previously pro-duced were c l a s s i f i e d into one of the following stages: 1) bud (unopened 11 heads, greater than 0.5 mm i n length); 2)flower head (opened heads, exposing f l o r e t s which had not w i l t e d ) ; 3)seed head (opened heads, but with wi l t e d f l o r e t s and f e r t i l i z e d o v a r i e s ) ; or 4)aborted head (heads which grow 3 to 5 mm i n length while i n the bud stage, then turn yellow and dry instead of completing development). Lengths and widths of .heads i n the bud stage were measured to the nearest 0.1 mm with c a l i p e r s . Heads at d i s t a l ends of primary branches were denoted by I, I I , I I I . . . ; of secondary branches by i , i i , i i i . . . ; of t e r t i a r y branches by a, b, c...; of quaternary branches by x', x", x'" ...; and of branches divided to the f i f t h degree by x^, x^, x^.... For example, a head appearing on the f i f t h primary branch, t h i r d secondary branch, fourth t e r t i a r y branch, second quaternary branch, and t h i r d branch divided to the f i f t h degree would be coded V i i i d^. Observations of d i f f u s e and spotted knapweed head appearance continued f o r 76 and 70 days, r e s p e c t i v e l y , when the plants started to per-ennate. Some of the spotted knapweed rosettes, which were produced at the bases of s t a l k s , started to b o l t . B o l t i n g from perennial rosettes does not occur u n t i l the following spring i n the f i e l d , and preliminary f i e l d work also indicated that few heads are i n i t i a t e d a f t e r the s t a r t of perennation. Therefore i t was assumed that d i f f u s e and spotted knapweed i n t h i s study had produced maximum numbers of heads 76 and 70 days, r e s p e c t i v e l y , a f t e r the beginning of head i n i t i a t i o n . 3.Results: The average number of new heads, which appeared at one sampling time, was greatest on day 28 for both d i f f u s e and spotted knapweed, with smaller peaks occurring on day 8 and 12 for d i f f u s e and spotted knapweed, r e s p e c t i v e l y 12 ( f i g . 5). Diffuse knapweed began to flower and set seed on day 32 and 36, respe c t i v e l y (Fig. 6), and spotted knapweed began to flower and set seed on day 28 and 34 (Fig. 7). Thetrate of head i n i t i a t i o n of d i f f u s e knapweed was greater before the f i r s t seed head appeared: from day 0 to day 36, heads appeared at a rate of 2.3% per day of the t o t a l number of heads produced, but from day 36 to day 76, only 0.5% of the t o t a l number of heads produced ap-peared per day. S i m i l a r i l y , the rate of spotted knapweed head i n i t i a t i o n from day 0 to day 34 was 2.6% per day of the t o t a l number of heads produced i n 70 days, but a f t e r day 34, the rate dropped to 0.3% per day. By day 34, 87% of the t o t a l number of spotted knapweed heads had appeared. Head abortion usually occurred when the f i r s t seed head appeared. For ex-sample, on d i f f u s e knapweed "A" plant (Fig. 8), the f i r s t seed head was ob-served on day 36, and seven out of the eight heads which aborted appeared on day 38. The remaining aborted head on t h i s plant, i n p o s i t i o n XII i i i , ap-peared on day 56. By the termination of the study, 10% and 20% of the t o t a l number of d i f f u s e and spotted knapweed heads produced, r e s p e c t i v e l y , had aborted. Aborted heads were usually i n proximal posi t i o n s on branches. The f i r s t head to appear was at the apex of the s t a l k on the f i r s t primary branch, thus preventing further s t a l k elongation. S i m i l a r i l y , lengths of primary, secondary, t e r t i a r y , etc. branches were l i m i t e d when the a p i c a l head appeared. Further head production on each branch always occurred from the d i s t a l to the proximal end. For example, on plant "A" the order of head appearance on primary branches I, I I , III...XI on the main axis was 2, 2, 4, 4, 6, 6, 8, 8, 8, 12, 20 and 34, re s p e c t i v e l y ; on secondary branches i , i i , i i i . . . i x on branch IV was 4, 6, 8, 8, 10, 18, 22, 22, and 24, r e s -p e c t i v e l y ; on t e r t i a r y branches a, b, c, and d on branch IV i i i was 8, 12, 30 i 0 10 20 30 40 50 60 70 80 DAYS AFTER THE BEGINNING OF HEAD PRODUCTION Figure 5. Average numbers of diffuse and spotted knapweed heads per plant appearing each four-day., i n t e r v a l , diffuse knapweed 0 0; spotted knapweed X- X 80 DAYS AFTER THE BEGINNING OF HEAD PRODUCTION Figure 6. Percent d i f f u s e knapweed heads i n developmental stages bud • •; flower H +; seed 0 -0: aborted * - *• ' i — * -percent heads appeared of t o t a l produced x x -P* ' DAYS A F T E R THE BEGINNING OF HEAD PRODUCTION Figure 7. Percent spotted knapweed heads i n developmental stages bud • — — — • ; flower H : + ; seed 0"--"0; aborted * — percent heads appeared of t o t a l produced x —x 9 Figure 8. Branching pattern and sequence of head appearance f or a t y p i c a l d i f f u s e knapweed plant. Numbers give the day of head appearance. C i r c l e d heads aborted. For branch code, see method. 17 12 and 22, r e s p e c t i v e l y ; on the quaternary branches c', c" and c 1" on branch IV i i i c was 12, 22 and 30, res p e c t i v e l y ; and on branch II i i b" the order of head appearance on the two branches divided to the f i f t h degree was 18 and 20, r e s p e c t i v e l y . The l a t t e r head was therefore coded I I i i b^. There was a shorter time, on the average, between consecutive head appear-ance on branches of a higher order ( i . e . , the larger primary and secondary branches) than on branches of lower order ( i . e . , the smaller t e r t i a r y and quaternary branches). Heads i n d i s t a l p o s i t i o n s reached the flowering stage i n s i g n i f i c a n t l y l e s s time (32.9 - 0.5 and 30.3 - 0.7 days f o r d i f f u s e and spotted knapweed, resp e c t i v e l y ) than did heads i n more proximal p o s i t i o n s (45.6 - 0.6 and 5 0 - 2 days for d i f f u s e and spotted knapweed, r e s p e c t i v e l y ) . There was no s i g n i f i c a n t d i f f e r e n c e i n development time from head i n i t i a t i o n to the end of the bud stage between d i f f u s e (39.6 - 0.8 days) and spotted (40 - 2 days) knapweed, when a l l the heads are averaged. Heads i n the bud stage can elongate up to 0.5 mm per day, but the average rate i s 0.25 mm per day. There was no s i g n i f i c a n t d i f f e r e n c e between d i f f u s e and spotted knap-weed i n the average time spent i n the flower stage (5.23 - 0.08 and 5.4 - 0.2 days, r e s p e c t i v e l y ) . On the average les s than 10% of the t o t a l number of heads for ei t h e r species were i n the flower stage at any one time. 4.Discussion: There are d e f i n i t e phases i n d i f f u s e and spotted knapweed head production. During the f i r s t phase terminal heads on primary branches appear, l i m i t i n g plant height and breadth. Af t e r a short pause the plants produce heads i n proximal p o s i t i o n s . By the time the t h i r d , or "flowering" phase i s reached, 75% of the t o t a l number of heads have appeared. In the t h i r d phase, d i s t a l -most heads flower i n the same order that they appeared. Only a few heads 18 flower at a time, and on small plants, heads flower one by one. This prob-ably ensures that i n the advent of poor growing conditions, such as a period of drought i n August, at least some heads w i l l have completed matur-ation. Few a d d i t i o n a l heads are i n i t i a t e d i n the fourth phase, the seed stage, and some heads i n the bud stage i n proximal p o s i t i o n s u s u a l l y abort. At t h i s point over 80% of the t o t a l heads have appeared. Development of heads i n proximal p o s i t i o n s i s very slow, and some heads may not elongate fo r weeks, u n t i l t h e i r turn comes for development. Regardless of the f a t e of heads, or the number of a d d i t i o n a l heads produced, the sequence of head i n i t i a t i o n of d i f f u s e and spotted knapweed i s so regular that head p o s i t i o n s can be coded for future reference. This eliminates the need for mechanical tagging of heads when heads must be marked for studies with g a l l f l i e s . Head development patterns for d i f f u s e and spotted knapweed are e s s e n t i a l l y the same, except that fewer heads are produced by spotted knapweed, with a shorter maturation time of the s t a l k . Branches of d i f f u s e knapweed are de-veloped i n the same manner as s t a l k s of spotted knapweed. Watson (1972) found that spotted knapweed rosettes could b o l t i n the green-house without vernalization. This study showed that not only spotted, but also d i f f u s e knapweed, could perennate without v e r n a l i z a t i o n . 5.Conclusions: 1. D i s t a l heads of d i f f u s e and spotted knapweed appear f i r s t and develop e a r l i e r and f a s t e r than proximal heads. 2. Normally only proximal heads abort. 3. Head abortion begins approximately when the f i r s t heads enter the seed stage. 19 4. Usually 75% of a l l heads appear before any heads enter the flower stage. 5. The percentage of heads i n the flower stage i s 10% or l e s s f o r d i f f u s e and spotted knapweed under greenhouse conditions. 6. Although spotted knapweed produces larger heads than d i f f u s e knapweed, the durations of the bud stages are s i m i l a r . 7. Heads coded according to branch positions can be found at l a t e r stages of plant development, regardless of the a d d i t i o n a l number of heads produced. i i . Urophora a f f i n i s and U. qua d r i f a s c i a t a: a)Emergence and l a r v a l weights: 1.Introduction: Varley (1947) found that U. jaceana r e s i s t e d d e s i c c a t i o n when i s o l a t e d from t h e i r g a l l s and completed development from the t h i r d i n s t a r to the adult stage i n gelatine capsules. Preliminary work showed that U_. a f f i n i s could do the same, and t h i s method of l a r v a l i s o l a t i o n was used to r e l a t e d l a r v a l weight with emergence time. This study also compares TJ. a f f i n i s and U. qua d r i f a s c i a t a l a r v a l weights and emergence times. 2.Method: Preliminary work showed that diapausing U. a f f i n i s larvae would not pupate u n t i l the spring unless they were c o l l e c t e d i n the f i e l d a f t e r December 1st. To ensure s u f f i c i e n t cold-shocking, spotted knapweed seed heads from the r e -lease s i t e were c o l l e c t e d i n January, 1976. These seed heads w i l l be c a l l e d spotted knapweed "1975" seed heads because they were produced during the 1975 growing season. The seed heads were stored i n a r e f r i g e r a t o r at 5°C. Batches of heads were removed, opened, and t h i r d - i n s t a r U. a f f i n i s larvae were dissected from t h e i r g a l l s , weighed to the nearest 0.001 mg and placed i n gelatine capsules. The larvae were incubated at 27°C under a 16-hour 20 photoperiod, and the capsules were checked d a i l y f o r emerged adults. The f i r s t l a r v a l batch was l e f t for 50 days, but l a t e r batches were l e f t for fewer days, due to a shortage of time. The weight of each larvae, and i t s sex, were correlated with development time to the adult stage. Seed heads were a r b i t r a r i l y c o l l e c t e d from the d i f f u s e and spotted knapweed release s i t e s i n January, 1977. These w i l l be c a l l e d "1976" seed heads. In a l l , 5520 d i f f u s e and 3350 spotted knapweed seed heads were incubated i n mesh cages at 29'C under a 16-hour photoperiod. Seed heads and sides of cages were misted d a i l y with water, and adult U_. a f f i n i s and U_. quadrifas-c i a t a were c o l l e c t e d d a i l y from the cages. Other seed heads, a r b i t r a r i l y c o l l e c t e d from the d i f f u s e and spotted knapweed release s i t e s i n October, 1976, were opened, and U. a f f i n i s and U_. q u a d r i f a s c i a t a larvae were dissected from t h e i r g a l l s , and weighed to the nearest 0.01 mg. Because few U_. quadri- f a s c i a t a larvae were found, d i f f u s e knapweed seed heads were randomly c o l -lected i n Vernon, where t h i s species had spread to, i n September, 1977. From these seed heads 60 U. quadrifasciata larvae were i s o l a t e d from t h e i r g a l l s , and weighed to the nearest 0.01 mg. 3.Results: The death rate of IJ. a f f i n i s i n gelatine capsules was 50%. The f l i e s died i n l a r v a l , pupal and adult (before emerging e n t i r e l y from the pupal case) stages. Some emerging adults had v e s t i g i a l wings. The following data was calculated using only larvae which developed into normal adults. Male U_. a f f i n i s larvae from spotted knapweed "1975" seed heads were s i g n i f i c a n t l y + + l i g h t e r (3.54 - 0.08 mg) than female larvae (4.7 - 0.1 mg). There was no s i g n i f i c a n t c o r r e l a t i o n between male U. a f f i n i s l a r v a l weight and days for emergence (r = 0.0851), nor between female U. a f f i n i s l a r v a l weight and days 21 for emergence (r = -0.0696). Males emerged s i g n i f i c a n t l y e a r l i e r than females (34.1 - 0.6 and 36.1 - 0.6 days, r e s p e c t i v e l y ) . The male:female sex r a t i o of normal U. a f f i n i s adults was 1.07 (N = 119). The average l a r v a l weight of U. a f f i n i s from "1975" spotted knapweed seed heads was s i g n i f i c a n t l y heavier than U_. af f i n i s from d i f f u s e and spotted knapweed release s i t e s i n 1976 (Table I I ) . Only f i v e IJ. q u a d r i f a s c i a t a larvae were found i n heads from the spotted knapweed release s i t e i n 1976, and t h e i r average weight was only 0.8 - 0.3 mg. The average l a r v a l weight of U. qu a d r i f a s c i a t a from d i f f u s e knapweed, c o l l e c t e d i n Vernon, was s i g -n i f i c a n t l y l e s s (1.74 - 0.08 mg) than that of U. a f f i n i s from the d i f f u s e and spotted knapweed release s i t e s i n 1976. There was a three- to four-f o l d v a r i a t i o n i n l a r v a l weight for both g a l l - f l y species. Table I I . Comparison of U. a f f i n i s and U. q u a d r i f a s c i a t a l a r v a l weights g a l l - f l y host knapweed species mean weight sample species and l o c a t i o n year (mg) s i z e U. a f f i n i s spotted knapweed, release s i t e 1975 4.15 + .09 a 119 U. a f f i n i s spotted knapweed, release s i t e 1976 2.40 + 0.5 b 472 U. a f f i n i s d i f f u s e knapweed, release s i t e 1976 3.1 + .2 c 67 U. quadri-f a s c i a t a spotted knapweed, release s i t e 1976 .8 + .3 5 U. quadri-f a s c i a t a d i f f u s e knapweed, Vernon 1977 1.74 + .08 d 60 "weights followed by a d i f f e r e n t l e t t e r vary s i g n i f i c a n t l y from one another 22 U_. quadrifasciata started emerging a f t e r 18 days of incubation at 29°C from d i f f u s e and spotted knapweed "1976" seed heads. No further emergence oc-curred a f t e r 45 days. IJ. a f f i n i s started emerging on days 21 and 23 from d i f f u s e and spotted knapweed, res p e c t i v e l y , but a few f l i e s were s t i l l appearing a f t e r 70 days of incubation, when the study was terminated. The male: female sex r a t i o s of TJ. quadrif a s c i a t a and U_. a f f i n i s from d i f f u s e knapweed were 1.32 and 1.33, re s p e c t i v e l y , and that of U. a f f i n i s from spotted knapweed was 1.32 (N = 88, 645 and 3361, r e s p e c t i v e l y ) . 4.Discussion: Although few IJ. q u a d r i f a s c i a t a were found i n the l a r v a l stage i n r e l e a s e -s i t e seed heads, there i s a strong i n d i c a t i o n that U_. q u a d r i f a s c i a t a weighs about one ha l f of IJ. a f f i n i s body weight. The l i g h t weight of U. quadri- f a s c i a t a from Vernon, where no U. a f f i n i s were found, supports t h i s conclu-sion. IJ. q u a d r i f a s c i a t a probably requires less energy to complete develop-ment than does U. a f f i n i s . This i s also indicated by the thinner n u t r i t i v e layer inside U. q u a d r i f a s c i a t a g a l l s , on which the larvae feed (Shorthouse, pers. comm.). Probably each U_. quadrif a s c i a t a g a l l reduces the p o t e n t i a l seed y i e l d of d i f f u s e and spotted knapweed by ha l f the amount of IJ. a f f i n i s . U_. quadrif a s c i a t a has an e a r l i e r and more concentrated period of emergence than IJ. a f f i n i s i n the laboratory. In the f i e l d IJ. quadrif a s c i a t a may emerge e a r l i e r than U. a f f i n i s to avoid i n t e r s p e c i f i c competition because both species attack knapweed heads i n the bud stage. E a r l i e r male than female emergence was also found by Storey (1976) for U. a f f i n i s and by Wadsworth (1914) f o r U. jaceana. This phenomenon may ensure that females are bred before they begin to o v i p o s i t when f l y d e n s i t i e s are sparse. 23 Like U. jaceana, U_. a f f i n i s can complete development from the t h i r d l a r v a l i n s t a r to the adult stage ins i d e gelatine capsules. The high death rate i n t h i s study i s a t t r i b u t e d to mechanical i n j u r y , or contamination by micro-organisms during d i s s e c t i o n of the g a l l s , rather than to desiccation. Pre-liminary work showed that p a r a s i t i z e d seed heads can be stored at room temp-erature from October u n t i l February i n a dry place, with no l a r v a l m o r t a l i t y . Because smaller larvae were more d i f f i c u l t to i s o l a t e than l a r g e r ones, average l a r v a l weight for U_. a f f i n i s from "1975" spotted knapweed seed heads i s probably overestimated. The male:female sex r a t i o was probably under-estimated because males have a smaller body s i z e than females. But t h i s study did show that JJ. a f f i n i s l a r v a l weight was not r e l a t e d to developmental time for each sex. Thus small females probably do not emerge e a r l i e r than large females i n the f i e l d . Fecundity i n insects i s r e l a t e d to female s i z e (southwood, 1971; Krebs, 1972). Therefore the p o t e n t i a l fecundity of U_. a f f i n i s i s probably constant throughout the emergence phase of the f i r s t gen-eration, given an abundance of o v i p o s i t i o n s i t e s and the same weather. Some reasons for heavier Tj. a f f i n i s larvae from d i f f u s e than from snotted knapweed i n 1976 are: the o r i g i n a l s t r a i n s released were d i f f e r e n t on the two release s i t e s , the indigenous host plants were d i f f e r e n t (Table I ) , only the stronger larvae survive i n d i f f u s e knapweed due to greater competition for food i n the smaller heads, or smaller larvae i n d i f f u s e knapweed seed heads are more e a s i l y injured because g a l l s tend to fuse together at lower den s i t i e s i n d i f f u s e than i n spotted knapweed. The male:female sex r a t i o s i n t h i s study were a l l greater than 1.0, but none were as high as 3.15, found by Storey (1976) for U. a f f i n i s i n spotted 24 knapweed. High male:female sex r a t i o s may, again, ensure mating before o v i p o s i t i o n . 5.Conclusions 1. U_. quadrif a s c i a t a has approximately half the body weight of U. a f f i n i s . 2. Male U. a f f i n i s weigh les s than females. 3. Male U_. a f f i n i s emerge e a r l i e r than females. 4. U. qu a d r i f a s c i a t a emerge e a r l i e r than U. a f f i n i s . 5. Developmental time f o r each sex of U. a f f i n i s i s not co r r e l a t e d with body weight. 6. Male:female sex r a t i o s of U. a f f i n i s and U. q u a d r i f a s c i a t a were about 1.3 i n 1976 on release s i t e s . b)SuperparasitJsm of d i f f u s e knapweed by U. a f f i n i s : 1.Introduction: Varley (1947) found that as U. jaceana egg d e n s i t i e s increased i n f i e l d Centaurea nemoralis heads, l a r v a l mortality i n the p r e - g a l l stage increased. Varley considerd that t h i s mortality was due to competition among young larvae for food and space. Both IJ. j aceana and IJ. a f f i n i s produce hard, woody g a l l s , larvae undergo three i n s t a r s before pupation, females o v i p o s i t i n knapweed heads which are 3 to 4 mm i n diameter, and females o v i p o s i t large numbers of eggs i n sin g l e heads when no other heads are a v a i l a b l e . IJ. a f f i n i s can lay as many as 59 eggs per spotted knapweed head (ZwBlfer, 1970) and U. jaceana was found by Varley (1947) to lay as many as 277 eggs per s i n g l e C_. nemoralis head. Roze and Frazer (19 79) predicted that above average U. a f f i n i s g a l l d e n s i t i e s of two and four per d i f f u s e and spotted knapweed head, r e s p e c t i v e l y , i n f i e l d populations, some heads would contain 25 a surplus of larvae and these heads would abort at early stages of develop-ment. This study determines the e f f e c t on d i f f u s e knapweed heads of exposure to IJ. a f f i n i s . 2.Method: When t h i s study was begun a l l the d i f f u s e knapweed growing i n the greenhouse had started to flower. To obtain plants i n the bud stage, bolted s t a l k s of twelve flowering d i f f u s e knapweed plants were removed at t h e i r bases, and the plants were allowed to perennate. A l l but two of the strongest peren-nating s t a l k s were removed, and these were allowed to reach the bud stage. The plants were a r b i t r a r i l y assigned a number, and three plants, f o r each of four treatments, were chosen at random. The treatments were low attack, medium attack, high attack and c o n t r o l . Both s t a l k s on the plants were treated the same, but were considered as separate sampling u n i t s . Therefore, there were s i x r e p l i c a t e s per treatment (three plants X two s t a l k s per p l a n t ) . The plants were attacked by mites, which were c o n t r o l l e d by p e r i o d i c a l l y washing the f o l i a g e with water. As soon as the heads on the f i r s t three primary branches, only, reached 3 to 4 mm i n length, the optimal s i z e for o v i p o s i t i o n by Tj. a f f i n i s (ZwHlfer, 1970), they were caged with a s i n g l e p a i r of f l i e s . H* a f f i n i s adults were obtained by c o l l e c t i n g p a r a s i t i z e d d i f f u s e knapweed seed heads on the d i f f u s e knapweed release s i t e i n January, 1977. The heads were incubated at 29°C under a 16-hour photoperiod, and were misted d a i l y with water. Adults were c o l l e c t e d d a i l y , and stored f o r three days before use because o v i p o s i t i o n occurs on the t h i r d or fourth day a f t e r emergence (ZwBlfer, 1970). 26 The tubular cages, closed at one end, were 5 cm long and 3 cm i n diameter, and were constructed of gauze. A piece of cotton was wrapped around each stem, one cm below the head, a pair of f l i e s introduced into a cage, and the open end of the cage was slipped over the head an tightened around the cotton with a drawstring. Previous work showed that plants were not damaged by caging, threrefore no heads on control plants were caged. The f l i e s were able to move f r e e l y around the heads, ins i d e the cages, and mating, as w e l l as head-probing were observed. P a i r s of f l i e s were l e f t f o r one, three and f i v e days for low-, medium-, and high-attack r a t e s , r e s p e c t i v e l y , on the singl e heads. Heads were coded, as previously (section II A i ) , when they appeared. Caging continued f o r three weeks, when the plants were l e f t to grow for another three weeks. The plants were harvested, and each head was found from the code, and c l a s s i f i e d into bud, flower/seed, or aborted stages. A l l seed heads were opened and searched f o r g a l l s . 3.Results: No g a l l s were found i n heads on control plants. Heads on c o n t r o l plants reached only about 3/4 of the s i z e of heads on plants which bolted for the f i r s t time. The percentage of aborted heads of the t o t a l number of caged heads d i f f e r e d s i g n i f i c a n t l y with attack rate (Table I I I ) . There was a s i g n i f i c a n t d i f f e r e n c e i n the percentage of aborted heads between c o n t r o l plants (18 - 5%) and high-attack-rate plants (49 - 5%), when a l l heads on the f i r s t three primary branches (caged + uncaged) are considered. Average g a l l d e n s i t i e s per seed head were low (Table I I I ) , and did not d i f f e r s i g -n i f i c a n t l y between attack rates. Only 12%, 5% and 26% of the caged heads, which matured to the seed stage, contained g a l l s f o r low-, medium-, and high-attack rates, r e s p e c t i v e l y . 27 Table I I I . E f f e c t of caging i n d i v i d u a l d i f f u s e knapweed heads with IJ. a f f i n i s _ _ „ . % aborted heads x number of g a l l s attactc rate - , . , _ .. of number caged per seed head low 2 1 - 6 % .2 - .2 + + medium 4 9 - 1 0 .05 - .05 high 5 2 - 3 .5 - .3 _ s i g n i f i c a n t d i f f e r e n c e between attack rates No d i s t a l heads on primary branches of con t r o l plants aborted, but 48% of the heads i n d i s t a l p ositions on high-attack-rate plants aborted. On attacked plants aborted heads were of two types: 95% resembled healthy heads i n the bud stage, but they were yellow and dried; and 5% appeared swollen, were dark brown or grey, and had hard outer s h e l l s . Some were f i l l e d with d e t r i t u s . The f i r s t type of head was not usually longer than 3.5 cm, but the second type was as long as 5 cm and one was found 6 cm i n length. 4.Discussion: Exposure of d i f f u s e knapweed heads to IJ. a f f i n i s can r e s u l t i n head abortion. It i s probable that high f l y d e n s i t i e s achieve the same e f f e c t on knapweed heads as prolonged exposure by one pa i r of f l i e s . Because eggs were not counted, there are two possible explanations. The f i r s t i s that super-numerary eggs or f i r s t - i n s t a r larvae caused head abortion. Normally, heads with U_. a f f i n i s act l i k e sinks, a t t r a c t i n g plant energy which would have been used for vegetative growth and sexual reproduction, but too many eggs or larvae could stimulate heads to a degree where they abort (Shorthouse, pers. comm.). These heads would be "superparasitized". Superparasitism . has been defined as "The phenomenon i n which more i n d i v i d u a l s of a given 28 species occur i n a host i n d i v i d u a l than can develop to maturity i n that host...Where t h i s occurs with s o l i t a r y endoparasites, internecine b a t t l e or p h y s i o l o g i c a l suppression of the supernumerary larvae or eggs r e s u l t s i n the s u r v i v a l of a dominant i n d i v i d u a l . In some cases, however, the host i t -s e l f succumbs prematurely before the supernumerary parasites are eliminated, and a l l p e r i s h . " (van den Bosch and Messenger, 1973) But knapweed heads are not i s o l a t e d hosts l i k e insects are. Only the superparasitized heads abort, and the plant survives. JJ. a f f i n i s , i n contrast, suffers high mortality i n superparasitized heads. The second explanation for d i s t a l head abortion a f t e r exposure to f l i e s i s damage by females t e s t i n g the heads for sui t a b i l i t y , ( S h o r t h o u s e , pers. comm.). The heads o f f e r e d i n t h i s study were smaller than average, probably because they were produced on perennating s t a l k s , and could have been e s p e c i a l l y susceptible to i n j u r y . I t would be to the advantage of the plant i f i t could detect probing before egg deposition and subsequently abort the probed head. I f heads are aborted early i n the growing season, seed y i e l d may not be affected (Tanskiy, 1969), and knapweed may have a better chance at producing seed l a t e r i n the growing season a f t e r peak f l y d e n s i t i e s . But Shorthouse (pers. comm.) found U_. a f f i n i s eggs i n aborted heads, and i t seems a.jpoor s u r v i v a l t a c t i c by t h i s species to destroy future o v i p o s i t i o n s i t e s simply by t e s t i n g t h e i r s u i t a b i l i t y - f o r egg deposition. Therefore i n f i e l d studies, any head i n a d i s t a l p o s i t i o n which aborts w i l l be assumed superparasitized, while heads i n proximal pos i t i o n s which abort w i l l be classed as undeveloped. Some heads can d e f i n i t e l y be i d e n t i f i e d as superparasitized on the basis of t h e i r swollen appearance, dark colour and hard outer s h e l l . The d e t r i t u s i n some of these heads indicates l a r v a l feeding and the i n a b i l i t y of f i r s t -i n s t a r s to induce g a l l formation, probably because of t h e i r surplus number. I n t r a s p e c i f i c competition i s further shown from preliminary f i e l d work, where some superparasitized heads contained s i n g l e , well-developed t h i r d -i n s t a r Uv a f f i n i s larvae, but without t h e i r fusiform g a l l s . JJ. a f f i n i s probably can cause head s u p e r p a r a s i t i z a t i o n of spotted knapweed, but because spotted knapweed heads contain larger and more numerous ovarioles than d i f f u s e knapweed, greater .numbers.of eggs must be l a i d per spotted knap-, weed head to cause head s u p e r p a r a s i t i z a t i o n . 5.Conclusions: • l.U_. a f f i n i s can cause d i f f u s e knapweed heads to abort at young stages of development, probably because more larvae hatch i n the*heads than there i s room and food f or g a l l development. 2. About 5% of the superparasitized heads can be p o s i t i v e l y i d e n t i f i e d on the basis of t h e i r swollen shape, -dark-colour and hard outer s h e l l . 3. Aborted heads i n d i s t a l p o s i t i o n s w i l l be classed as superparsitized by IJ. a f f i n i s and aborted heads i n proximal p o s i t i o n s on branches w i l l be classed as undeveloped. c ) G a l l - f l y head-size preferences: 1.Introduction: This study makes two comparisons of head sizes selected by g a l l f l i e s f o r o v i p o s i t i o n : IJ. a f f i n i s on d i f f u s e and spotted knapweed, and U. a f f i n i s and U. q u a d r i f a s c i a t a on d i f f u s e knapweed. 30 2.Method: Seed heads were c o l l e c t e d a r b i t r a r i l y from d i f f u s e and spotted knapweed r e -lease s i t e s i n January, 1976, and incubated at 29*C under a 16-hour photo-period, as i n section II i i a. Diffuse and spotted knapweed rosettes were-' -dug i n October, 1975, from the Kamloops area, and planted into pots.^••15.2 cm i n diameter. The plants were stored i n an unheated sh e l t e r i n Vancouver u n t i l December, when they were transferred to a greenhouse, and allowed to bolt under a 16-hour photoperiod. Aft e r emergence, f l i e s were stored f o r three days i n separate cages. Then the f l i e s were placed i n communal cages that contained young f l i e s , as w e l l as those which had been used i n the previous t r i a l s . U. a f f i n i s , emerging from d i f f u s e knapweed, were kept separate from those that emerged from spotted knapweed, and U_. qua d r i f a s c i a t a were separated from U. a f f i n i s . F l i e s were a r b i t r a r i l y chosen from communal cages for each new t r i a l . A l l caged f l i e s were provided with a yeast-honey s o l u t i o n from a cotton plug, and the sides of the cages were misted d a i l y with water. U. qua d r i f a s c i a t a emerged only from d i f f u s e knapweed i n s u f f i c i e n t quantity for t r i a l s . So few JJ. q u a d rifasciata were obtained that only four t r i a l s on d i f f u s e knapweed were made, compared with seven and thirty-one t r i a l s of IJ. a f f i n i s on d i f f u s e and spotted knapweed, r e s p e c t i v e l y . Experimental units were d i f f u s e knapweed branches and spotted knapweed s t a l k s . At least three heads i n the optimal s i z e range, >3 <8 mm i n length (ZwBlfer, 1970) for U. a f f i n i s were a v a i l a b l e per t r i a l . There were averages of 16 - 1 d i f f u s e and 2 3 - 3 spotted knapweed heads i n the optimal s i z e range per t r i a l with TJ. a f f i n i s . The same optimal s i z e range as for U. a f f i n i s was assumed for U. qu a d r i f a s c i a t a , and there were 5 7 - 5 d i f f u s e knapweed heads i n t h i s 31 range per t r i a l with U_. q uad r i f as c i a t a Maturity of the plants v a r i e d . In some t r i a l s only proximal heads were i n the optimal si z e range, whereas i n other t r i a l s d i s t a l heads were a v a i l a b l e . Heads not i n the optimal s i z e range were l e f t undisturbed on the s t a l k s . Lengths and widths of a l l the heads i n the bud stage were measured with c a l i p e r s to the nearest 0.1 mm (excluding bracts and spines), and head positions were coded as i n section II A i . Numbers of f l i e s per t r i a l depended somewhat on t h e i r a v a i l a b i l i t y , but on the average there were 7 - 1 and 8 - 1 pai r s of U. a f f i n i s per t r i a l for d i f f u s e and spotted knapweed, r e s p e c t i v e l y , and 7 - 1 p a i r s of U. quadri- f a s c i a t a for d i f f u s e knapweed. Cages were constructed of gauze, stretched over c y l i n d r i c a l wire frames, 30 cm i n length and 15 cm i n diameter. Cages were positioned on the plants without damaging branches or heads, and the f l i e s were caged for three daysy a f t e r which the f l i e s and cages were removed. The plants were l e f t to mature i n the greenhouse under natural l i g h t conditions f o r the month of A p r i l . Then the plants were harvested and the heads were c l a s s i f i e d into seed, un-developed or superparasitized stages. Heads were considered superparasitized i f they aborted a f t e r being exposed to the f l i e s , and were >3 mm i n length at the time of exposure. A l l other heads which aborted were assumed to be undeveloped. 3.Results: Head s u p e r p a r a s i t i z a t i o n occurred only by U. a f f i n i s on spotted knapweed. The test spotted knapweed plants had, i n some cases, unusually large heads,. and g a l l counts greater than 8 per head were not uncommon. The highest count was 17 U. a f f i n i s g a l l s per head. I t was assumed that superparasitized r. 32 spotted knapweed heads contained 9 U. a f f i n i s larvae i n t h i s study.because above?.this-density crowding seemed to decrease g a l l s i z e , and increase the mortality of the f l i e s . Some superparasitized heads were 5 to 6 mm long at the time of caging, but most were 3 to 4 mm long. D i s t r i b u t i o n s of the percentage a v a i l a b l e heads for o v i p o s i t i o n d i f f e r e d s i g n i f i c a n t l y with the percentages of p a r a s i t i z e d heads for a l l cases (Table 2 IV) according to X t e s t s . Table IV. " G a l l - f l y head-size preferences f l y species knapweed length/width head preference (mm) * degrees of freedom U. a f f i n i s d i f f u s e •^5 < 6, length ^3 < 4, width 81.16 110.15 4 2 spotted £4 < 5, length $3 <4, width 332.50 290.30 4 4 U. qua d r i f a s c i a t a d i f f u s e £5 < 6, length ^3 < 6, width 47.76 21.13 1 1 s i g n i f i c a n t d i f f e r e n c e between % a v a i l a b l e and % p a r a s i t i z e d heads for a l l cases Although most of the p a r a s i t i z e d heads were about 5 mm long and 3.5 mm wide, frequency d i s t r i b u t i o n s of the percentage of heads p a r a s i t i z e d v a r i e d . U. 2 a f f i n i s preferred wider spotted than d i f f u s e knapweed heads ( % • 2 0 1 . 2 4 ) , 2 and U. qua d r i f a s c i a t a oviposited i n s i g n i f i c a n t l y longer ( iC 30*32) and 2 wider ( ^ ( - J J = 13.51) d i f f u s e knapweed heads than did U. a f f i n i s . In t r i a l s with JJ. q u a d r i f a s c i a t a the plants had already reached the seed stage, and only proximal heads were i n the bud stage. These heads were s i g n i f i c a n t l y 2 2 shorter ( 32.99) and narrower ( X / O N = 24.85) than the d i f f u s e knapweed (.o) (z) heads used i n t r i a l s with U. a f f i n i s . %5(H 40 30-20-10 %30 A 2 0 10 l l l l i n , i , 1 I I II <2 £ 2 . ^ 3 *3<f *1<5 ?s<6 >6<7 >7<8 38<? >3<10 » | 0 HEAD LENGTH CLASS (MM) 1 <2 22<3 £3<4 ?4<5 ?S<6 >6<7 »7<8 sg<9 ^<jo HEAD WIDTH CLASS (MM) Figure 9. Knapweed head sizes a v a i l a b l e and accepted for o v i p o s i t i o n by g a l l f l i e s i and i i : U. a f f i n i s on d i f f u s e knapweed; i i i and i v : U. a f f i n i s on spotted knapweed; v and v i ; U. quadrif asciata on d i f f u s e knapweed. Heads a v a i l a b l e • • • ; heads accepted CZZ3 u> 34 (Figure 9 continued) %50 1 40 H 30 H 20 H 10 J i l II <2. * 2 0 >3«f » f < 5 >S<6 *<4<7 Z7<S »S<9 HEAD LE N G T H C L A S S (MM) <2 iZ<3 >f<5 HEAD WIDTH C L A S S (MM) The range of head-sizes p a r a s i t i z e d by the f l i e s v a r i e d greatly: U. a f f i n i s oviposited i n d i f f u s e knapweed heads <7 mm long (Fig. 9i) and < 5 mm wide (Fig. 9 i i ) , arid"in spotted knapweed heads which were < 8 mm long ( F i g . 9 i i i ) and <7 mm wide (Fig. 9iv) . IJ. quadrif a s c i a t a accepted d i f f u s e knapweed ".. heads which were >4 <7 mm long (Fig. 9v) and >2 <5 mm wide (Fig. 9 v i ) . 4.Discussion: ZwBlfer (1970) found that U_. a f f i n i s from spotted knapweed preferred spotted knapweed heads which were 3 to 7 mm i n length, and Storey (1976) found that U. a f f i n i s from spotted knapweed preferred spotted knapweed heads 4.1 - .5 mm i n width. Most of the present r e s u l t s were s i m i l a r . However, IJ. a f f i n i s i n -t h i s study preferred longer d i f f u s e knapweed heads than the 2.3 to 3 mm range that ZwBlfer (1970) recorded. The present r e s u l t s , however, are si m i ^ l a r to Berube's (1980), where U. a f f i n i s on d i f f u s e knapweed showed a preference f o r heads 3.5 to 5.5 mm i n length, but could o v i p o s i t i n heads 35 1.5 to 9.5 mm long. Berube (1980) l e f t the caged f l i e s with the plants for one week and measured the heads only a f t e r removing the cages. In t h i s study the heads were measured before the f l i e s were released, and the f l i e s were l e f t on the plants for three days. Knapweed heads i n the bud stage elongate an average of 0.25 mm per day (section II A i ) , therefore the actual length-preference of TJ. a f f i n i s may be larger or smaller than suggested by the above methods. This may explain, i n part, the wide acceptance ranges of heads sui t a b l e f o r o v i p o s i t i o n by g a l l f l i e s , but even when head elongation during the attack phase i s ignored, g a l l f l i e s accept a very wide rage of head sizes i n the laboratory. Head maturity, rather than external s i z e , i s probably a more important c r i t e r i o n f o r egg deposition. ZwBlfer (1970) sometimes ob-served JJ. a f f i n i s probing spotted knapweed heads for one hour without o v i -p o s i t i n g eggs. JJ. a f f i n i s o v i p o s i t o r s have small sensory pores, which may be used to detect the s i z e and developmental stage of the ovarioles (ZwHlfer, 1970). Knapweed used i n t h i s study had a great v a r i a t i o n i n head s i z e among plants, where head s i z e i s a measure of length and width j u s t before the head opens to flower. The d i f f u s e knapweed a v a i l a b l e to ZwBlfer may have had small heads, and t h i s may explain why ZwBlfer's (1970) JJ. a f f i n i s preferred such very short d i f f u s e knapweed heads compared with JJ. a f f i n i s i n t h i s study. S i m i l a r i l y , the heads offered i n t h i s study to JJ. quadrifasciata were smaller than the ones offered to IJ. a f f i n i s . Possibly proximal heads are smaller than d i s t a l heads within s i n g l e plants, and because JJ. quadrifas- c i a t a was exposed only to proximal heads, actual head preferences could be larger than i n t h i s study. This may explain why Berube (1980) found JJ. qua-" d r i f a s c i a t a to prefer such large d i f f u s e knapweed heads (7.5 to 9.5 mm long). Head length i s a better measure than head width for estimating the head si z e preferred by g a l l f l i e s because JJ. a f f i n i s on d i f f u s e and spotted knap-weed, U_. quadrif a s c i a t a on d i f f u s e knapweed, and U. j aceana on black knap-weed (Storey, 1976; and Varley, 1947) a l l preferred heads 3 to 4 mm wide. Probably a better method of estimating head s i z e preferences would be to base head s i z e on the f r a c t i o n which the head has elongated, assuming that maximum elongation occurs j u s t before anthesis. Thus U. a f f i n i s prefers heads 1/3 to 1/2 elongated, and IJ. q u a d r i f a s c i a t a prefers heads 1/2 to f u l l elongation. "...species having i d e n t i c a l e c o l o g i c a l niches ( i . e . e c o l o g i c a l homologs) cannot coexist for long i n the same habitat" (DeBach, 1974). U. a f f i n i s and U_. q u a d r i f a s c i a t a are not true e c o l o g i c a l homologs because the l a t t e r species prefers knapweed heads at more mature stages of development than does the former, even though there i s considerable overlap i n these p r e f e r -ences. Both species may coexist at low d e n s i t i e s . But because U_. a f f i n i s can u t i l i z e heads at younger stages of development, t h i s species may saturate-a l l o v i p o s i t i o n s i t e s before they have reached the stage preferred by U_. quadrif a s c i a t a . Thus JJ. a f f i n i s may suppress U_. quadrif a s c i a t a at high f l y densities.. -5.Conclusions: 1. U. a f f i n i s u s ually o v i p o s i t s i n smaller heads than does U. q u a d r i f a s c i a t a , but there i s considerable overlap i n head s i z e preferences by the two species. 2. A better measurement for g a l l - f l y preference of knapweed head sizes for o v i p o s i t i o n may be the f r a c t i o n which the head i n the bud stage has elong-ated, where f u l l elongation i s the point j u s t before the head opens to flower. 37 B. FIELD STUDIES 1. Synchrony of g a l l - f l y attack with knapweed phenology: 1.Introduction: For maximum seed destruction g a l l - f l y emergence should be synchronized with the appearance of knapweed heads at the r i g h t stage of development f or o v i p o s i t i o n . There should be enough f l i e s throughout the budding period of the weeds to heavily attack a l l heads. This i s a study of the abundance of JJ. a f f i n i s and JJ. qu a d r i f a s c i a t a i n r e l a t i o n to the number of knapweed heads at the r i g h t stage of o v i p o s i t i o n on the d i f f u s e and spotted knapweed release s i t e s . Diffuse knapweed phenologies i n 1976 and 1977 were compared, and d i f f u s e and spotted knapweed phenologies were compared i n 1976. 2. Method: Preliminary work showed that adult g a l l f l i e s could be counted, sexed and i d e n t i f i e d to species on. knapweed,- where they spend most of t h e i r time. In 1976, a rope, marked i n meters, was a r b i t r a r i l y stretched across the release s i t e on June 4th, one week before sampling took place. T h i r t y d i f f u s e knap-weed plants and t h i r t y spotted knapweed s t a l k s ( i . e . s i n g l e s t a l k s from perennial p l a n t s ) , c l o s e s t to each meter mark, were tagged at t h e i r bases with blue p l a s t i c ribbons. The following week, JJ. a f f i n i s and JJ. quadri- f a s c i a t a were counted and sexed before sunrise, when they were e s p e c i a l l y r e -luctant to f l y . A primary branch from each tagged d i f f u s e knapweed plant was taken by choosing a random number, and removing the corresponding branch. The e n t i r e spotted' knapweed s t a l k was removed. The rope was moved to a new l o c a t i o n , and the plants tagged f or the following week's sampling. Weekly sampling continued u n t i l no f l i e s were found at the end of the growing season. Heads i n the bud stage on the samples were c l a s s i f i e d as young, i f 38 they were $7 mm i n length of old i f they were >7 mm i n length. I f head s i z e appeared unusually large or small, heads were c l a s s i f i e d as young i f they were £2/3 elongated and old i f they were >2/3 elongated, where f u l l elongation was defined as head length j u s t before the head opened to flower. I t was assumed that young heads were p r i m a r i l y a v a i l a b l e to U. a f f i n i s while old heads were a v a i l a b l e to U. qu a d r i f a s c i a t a. In 1977 only the d i f f u s e knapweed release s i t e was monitored, and sampling was not begun u n t i l June 27th because 1976 r e s u l t s showed that the f l i e s did not emerge u n t i l the middle of June. Six 50 X 50 cm square pl o t s were marked out i n the spring of 1977 by throwing a square and del i n e a t i n g the pl o t with baling twine where the square landed. Before sunrise, once a week, adult f l i e s on knapweed and on other vegetation within the p l o t s were counted, and i d e n t i f i e d to species and sexed. Heads i n the bud stage were c l a s s i f i e d as young or old as above. 3.Results: In 1976 peak numbers of d i f f u s e knapweed heads i n the young bud stage ( F ig. lO i ) and f i r s t generations of U. a f f i n i s (Fig. H i ) and U. qu a d r i f a s c i a t a (Fig. l l i i ) occurred i n the middle of Ju l y . In 1977 highest counts of f i r s t generations of g a l l f l i e s (Fig. 12) and d i f f u s e knapweed heads i n the young bud stage (Fig. 13) occurred on the f i r s t sampling date (June 27th). Thus, unfortunately, the early period of g a l l - f l y emergence and knapweed head ap-pearance was missed. Male g a l l f l i e s on both release s i t e s appeared one week e a r l i e r than females i n 1976. In 1976, although peak numbers of heads i n the young bud stage appeared i n H-CW c CD i H > o fD c 1—' fD 3 fD 1-1 TO fo pu cn 09 cr ro fD c Cu cn 3 H- c CO rt g rt fD cr fu cn fD CN i-i ro H- 0) 3 c o I-1 H i 3-ON fD • fu c Cu V • cn 0 • H> M 3 Cu Cu H-cr Hi O c Hi CU d 3-' cn cn fD AVERAGE NUMBER OF SPOTTED KNAPWEED HEADS PER STALK AVERAGE NUMBER OF DIFFUSE KNAPWEED HEADS PER BRANCH 6e Figure 11. Average number of male and female g a l l f l i e s per branch the d i f f u s e knapweed release s i t e i n 1976 i . U. a f f i n i s ; i i . U. quadrifasciata. females 0 " 1 • 0; males X X 41 o _ j D_ QL LU 0. IS) LU < O LU 00 e LU 15 10 Figure 12. Average numbers of U. a f f i n i s and U. quadrifasciata per pl o t on the d i f f u s e knapweed release s i t e i n 1977. U. a f f i n i s O; IJ. quadrif a s c i a t a X—— LU LO =3 LL I— LU O l-H _ J Q 0. Ll_ IY. O LU D_ QC LU IS) 00 O LU a O LU LU Q. < 2 150 100 50 J Figure 13. JUNE 0 a 0 '1' JULY SEPTEMBER Average numbers of heads i n young and old bud stages on the d i f f u s e knapweed release s i t e i n 1977. young bud stage 0 — 0 ; old bud stage X- X Figure 14. Average numbers of male and female g a l l f l i e s per s t a l k on the spotted knapweed release s i t e i n 1976. P_- a f f i n i s ; i i . JJ. q u a d r i f a s c i a t a females 0 0; males X X early July on both d i f f u s e and spotted knapweed release s i t e s , heads i n the young bud stage were a v a i l a b l e f o r a shorter period on :the spotted ( F ig. l O i i ) than on the d i f f u s e ' ( F i g . i l Q i ) knapweed release s i t e ( 7i36.4). Very few new heads were i n i t i a t e d a f t e r August 1st on the spotted knapweed release s i t e , but heads were produced continuously i n low numbers on d i f f u s e knapweed u n t i l the middle of September. JJ. a f f i n i s was detected one week e a r l i e r on the spotted (Fig. 14i) than on the d i f f u s e ( Fig. H i ) knapweed release s i t e i n 1976. On both release s i t e s (Figs. 101 and l O i i ) i n 1976 no heads i n the old bud stage were seen u n t i l the middle of July, w e l l a f t e r the time when JJ. quadri- f a s c i a t a had emerged (Figs. l l i i and 1 4 i i ) , and numbers of old buds were to Q < LU X oo LU ZD Q < L L O o t-< LU (J < LU > < 43 . H Figure 15. Mean r a t i o s of g a l l flies:heads on the d i f f u s e knapweed release s i t e i n 1977. U_. affinis:heads i n the young bud stage O •—0 U- quadrifasciata:heads i n the old bud stage X X 44 never as p l e n t i f u l as young ones. Second g a l l - f l y generations on the d i f f u s e knapweed release s i t e peaked 3 to 4 weeks l a t e r i n 1976 (early September) than i n 1977 (early August), an i n t e r v a l s i m i l a r to that between f i r s t generations. In 1976 g a l l f l i e s were found on the d i f f u s e knapweed release s i t e as l a t e as the second week i n October, but i n 1977 no f l i e s were found a f t e r the middle of September. On spotted knapweed g a l l f l i e s did not appear i n samples u n t i l late August 1976, when very few new heads were being i n i t i a t e d . Fewer U. a f f i n i s were seen during the second-generation emergence period (37% and 8% of the t o t a l f l i e s seen i n 1976 on d i f f u s e and spotted knapweed release s i t e s , r e s -pectively) than during the f i r s t generation. The reverse was true for IJ. quadrif as d a t a : 69% and 58% of the t o t a l f l i e s seen i n 1976 on the d i f f u s e and spotted knapweed release s i t e s , r e s p e c t i v e l y , appeared as second-gener-a t i o n f l i e s . The r a t i o of g a l l f l i e s to d i f f u s e knapweed heads dropped to low values. between f i r s t - and second-generation peaks. On June 27th, 1977, there was one U. a f f i n i s to every seven heads i n the young bud stage, one-U. quadrifas-c i a t a to every t h i r t y - f i v e heads i n the old bud stage, and one g a l l f l y (JJ. a f f i n i s + JJ. quadrif asciata) to every s i x heads i n the young bud stage -(Fig. 15). On August 1st, between f i r s t and second g a l l - f l y generations, there was one U. a f f i n i s to every 618 heads i n the young bud stage, and. no IJ. q u a d r i f a s c i a t a were found i n samples, even though there were an average of 277 heads i n the old bud stage. 5.Discussion: 45 The mean d a i l y June temperature and t o t a l p r e c i p i t a t i o n i n 1976 was 16.1°C and 33.02 mm, whereas i n 1977 the temperature i n June was about 3 C°warmer (19.3°C) and about h a l f as much r a i n f e l l (16.5 mm; Table V). Because of the wetter, cooler spring of 1976, both knapweed heads and f i r s t - f l y gener-ations appeared l a t e r i n 1976 than i n 1977. The warmer spring of 1977 was t y p i c a l of the t h i r t y - y e a r average for the Kamloops area, therefore early June i s probably the normal time for head and adult *f-ly< appearance. Storey (1976) found that weather s i m i l a r i l y influenced the emergence of U_. a f f i n i s , which appeared l a t e r i n the cool spring of 1975 than i n the warmer one of 1974. In B r i t i s h Columbia the timing and duration of second.gall-fly gener-ations are also affected by weather conditions. The mean d a i l y temperature for August, 1976 was about 3 C° cooler (17.78°C) than i n 1977 (21.90°C), and there was an unusual amount of r a i n f a l l (122.17 mm) i n 1976 compared to 1977 (34.0 mm), and compared also the t h i r t y - y e a r average of 26.92 mm (Annonymous, 1976). Although the June temperature on the spotted knapweed release s i t e i n 1976 was cooler (13.9°C) than on the d i f f u s e knapweed release s i t e (16.11°C), g a l l f l i e s emerged e a r l i e r on the spotted than on the d i f f u s e knapweed r e -lease s i t e . The spotted knapweed release s i t e i s located on a steep, south-facing slope. I t therefore receives more i n s o l a t i o n than the f l a t bank, which slopes northward i n places, on which the d i f f u s e knapweed release s i t e i s found. Varley (1947) found that the average temperature in s i d e U. jaceana g a l l s was 5 C° warmer than a i r temperature. Usually the temperature i n plant t i s s u e s , such as buds or leaves, i s much warmer i n d i r e c t sunlight than a i r temperature (Wellington, 1950). G a l l - f l y larvae probably experience warmer temperatures on the spotted than on the d i f f u s e knapweed release s i t e due to the t e r r a i n , r e s u l t i n g i n e a r l i e r head and f l y appearance on the ' i . Table V. Mean d a i l y temperature (°C) and t o t a l p r e c i p i t a t i o n (mm) for the d i f f u s e 1 a n d spotted knapweed release s i t e s month -x d a i l y temp. (°C) on release s i t e s t o t a l precipitat*ion< (mm) on release s i t e d i f f u s e knapweed spotted knapweed d i f f u s e knapweed spotted knapweed 1975 1976 1977 1975 1976 1977 1975 " 1976 1977 1975 1976 1977 January -6.67 -2.78 -.41 -5.00 -2.78 -3.60 52.83 19.56 20.00 108.20 79.25 33.50 February -8.33 .56 2.50 -7.22 1.11 .40 38.86 10.16 5.40 89.92 41.15 52.10 March 1.67 2.22 4.70 .56 0.00 2.70 15.24 18.80 11.70 35.56 26.16 20.30 A p r i l 7.78 9.45 10.50 5.56 7.22 8.30 10.41 3.81 14.50 19.05 21.34 11.40 May 13.33 13.89 13.50 11.67 12.78* 11.60 18.80 21.08 22.60 29.72 55.88 44.70 June 16.67 16.11 19.30 15.00 13.89 16.90 29.21 33.02 16.50 34.80 51.05 37.60 July 22.78 19.45 19.90 20.56 17.22 17.70 13.97 32.26 24.60 31.75 32.77 44.20 August 18.33 17.78 21.90 16.11 16.11 20.20 23.88 122.17 34.00 42.16 139.45 23.40 September 15.56 15.56 13.50 13.89 13.89 12.00 8.13 8.89 26.50 7.62 11.43 46.10 October 7.78 8.33 8.30 7.22 6.67 7.00 9.65 20.07 8.30 61.26 23.88 20.80 November 2.78 2.78 .30 1.67 2.22 0.00 33.53' 14.22 50.10 66.29 14.22 84.50 December -1.67 0.56 -6.60 2.22 0.00 -5.60 32.77 38.10 63.30 83.82 42.16 79.50 Weather data for the d i f f u s e and spotted knapweed release s i t e s are from Kamloops Ai r p o r t and Sorrento East, respectively, found i n Climate of B.C., pub. by the B r i t i s h Columbia M i n i s t r y of A g r i -'culture. . * -data from Sicamous 47 -spotted than on the d i f f u s e knapweed release s i t e . .Earlier appearance of U_. a f f i n i s on spotted knapweed may by b e n e f i c i a l for t h i s fly. species be-cause spotted knapweed has a short, concentrated budding phase on t h i s s i t e . On d i f f u s e knapweed U. a f f i n i s does not gain much by emerging " e a r l i e r be-cause heads are p l e n t i f u l u n t i l August. Even i n the exceptionally cool, wet August of 1976, spotted knapweed ceased to i n i t i a t e new neads a f t e r the middle of August, when energy was diverted to the development of rosettes at bases of st a l k s of perennating plants. In most years few heads i n the bud stage can be found on the spotted knapweed release s i t e a f t e r August 1st. Thus second-fly generations here are a waste of energy on the part of the f l i e s unless they are able to disperse to locations where spotted knapweed matures l a t e . P r i o r to 1977 such i n -f e s t a t i o n s grew nearby, but have since been sprayed* In contrast d i f f u s e knapweed does not e n t i r e l y cease head production i n August because i t does not usually perennate. As long as there are energy reserves and the weather i s not exceedingly hot and dry, most d i f f u s e knapweed plants continue to i n i t i a t e new heads u n t i l the st a l k s are k i l l e d by cold temperatures i n October. (The roots are s t i l l a l i v e , and i f they are injured the plant perennates r e a d i l y . ) Second f l y generations are able to f i n d late-appearing heads on t h i s s i t e , and i n f a c t , on some plants, the r a t i o of flies:heads was larger than during the f i r s t generation. Thus the second generation i s e s s e n t i a l for seed reduction on d i f f u s e knapweed, but has l i t t l e importance i n terms of o v e r a l l e f f e c t on spotted knapweed. The large second generation produced by U. qua d r i f a s c i a t a may be a s u r v i v a l 48 t a c t i c by t h i s species here and i n countries of o r i g i n . Numbers of TJ. a f f i n i s are fewer f o r the second than for the f i r s t generation, and there are as many heads i n old as i n young stages of bud development i n August and September. Few heads reached the old bud stage during the f i r s t f l y generations because many heads were superparasitized by U. a f f i n i s or r e -mained undeveloped. The assumption that heads ^7 mm i n length were p r i m a r i l y a v a i l a b l e to U. a f f i n i s whereas heads >7 mm i n length were l e f t f o r TJ. quadrifasciata proved to be v a l i d . Berube (1980) found that i n the f i e l d U. a f f i n i s probed heads 7 mm or less i n length while i n no instance did TJ. quadri-\ f a s c i a t a attempt to probe heads l e s s than 6 mm long. On the d i f f u s e knapweed release s i t e the r a t i o of g a l l f l i e s to knapweed heads i s very low during a period of about seven days between f i r s t - and second-fly generations for two reasons: generations are non-overlapping, and JJ. a f f i n i s and U. qua d r i f a s c i a t a f i r s t - and second-generation emergence peaks are synchronized. Large numbers of heads escape attack between f l y generations, or may be attacked only very l i g h t l y , thus allowing for high seed y i e l d by heads appearing during t h i s i n t e r v a l . This i s one of the major weaknesses of the two Urophora species as b i o l o g i c a l c o n t r o l agents. 49 This study showed that male g a l l f l i e s emerge e a r l i e r than females, as found e a r l i e r i n the laboratory (section II A i i ) and also by Storey (1976). 5. Conclusions: •  •' •» • - •<- -1. G a l l f l i e s usually appear i n early June i n good synchrony with knapweed head appearance. Weather conditions can delay or hasten t h i s event. 2. F i r s t - and second-fly generations do not overlap, and U. a f f i n i s appear-ance i s synchronized with that of U. qu a d r i f a s c i a t a , allowing the heads that are produced between generations, i n early August, to escape heavy attack. 3.Second g a l l - f l y generations on the d i f f u s e knapweed release s i t e are e s s e n t i a l for e f f i c i e n t seed reduction, but second-fly generations on the spotted knapweed release s i t e probably have l i t t l e e f f e c t on seed y i e l d . 4.Spotted knapweed has a short, concentrated period of head production, but d i f f u s e knapweed i n i t i a t e s new heads u n t i l f r o s t k i l l s the stal k s i n the f a l l . 5.Males emerge e a r l i e r than females. i i . G a l l - f l y d e n s i t i e s on release s i t e s : 1.Introduction: Varley (1947) found that l a r v a l m o r t a l i t y of U. jaceana i n C_. nemoralis heads was "higher i n those flower heads containing many larvae than i n those containing only one or two. Thus early l a r v a l m o r t a l i t y i s density dependent, and overcrowding at t h i s stage i s p o t e n t i a l l y able to l i m i t the population density...Clearly some other factors prevent the population density r i s i n g to such a l e v e l that competition between larvae i s e f f e c t i v e as a c o n t r o l l i n g f a c t o r . " Varley subsequently showed that the c o n t r o l l i n g 50 factor was the chalcid p a r a s i t e , Eurytoma curta, and that Habrocytus trypetae could p o t e n t i a l l y contribute the regulation of TJ. jaceana. Enemies i n countries of o r i g i n of U_. a f f i n i s are the c h a l c i d parasites Eurytoma"tibialis, E_. robusta and Habrocytus spp. (ZwAlfer, pers. comm.). These parasites are of the same genera as found by Varley (1947) to be c o n t r o l l i n g TJ. jaceana. Furthermore, other enemies attacking U. a f f i n i s i n countries of o r i g i n , Torymus spp. and Metzneria spp. (ZwBlfer, pers. . . comm.) have also been reported by Varley (1947) attacking U. jaceana. TJ. a f f i n i s and TJ. j aceana may share heads with U. q u a d r i f a s c i a t a , and g a l l and l a r v a l development i s s i m i l a r "as?*dtiscussed previously (section II A i i b ) . The one major differ e n c e i s that the f i r s t U. a f f i n i s instar,; hatches from the egg, but the f i r s t U. jaceana i n s t a r develops ins i d e the egg and the second i n s t a r hatches from the egg (ZwBlfer, 1970; Varley, 1947). This study monitors U. a f f i n i s s u r v i v a l i n knapweed heads to de-termine i f Varley's (1947) p r e d i c t i o n of density-dependent l a r v a l competition for food and space as a p o t e n t i a l c o n t r o l l i n g factor of TJ. jaceana w i l l regulate U. a f f i n i s populations. TJ. a f f i n i s and U. q u a d r i f a s c i a t a have coexisted since 1972 and 1975 on the d i f f u s e and spotted knapweed release s i t e s , r e s p e c t i v e l y . The two species were released i n the same year on d i f f u s e knapweed, but TJ. a f f i n i s had been released on spotted knapweed four -years p r i o r to the detection of TJ. q u a d r i f a s c i a t a on t h i s s i t e . Varley ('1947) and Wadsworth (1914) indicated that U. jaceana numbers were not affected by TJ. q u a d r i f a s c i a t a . G a l l den-s i t i e s of U. q u a d r i f a s c i a t a were monitored to f i n d any regulating factors of U. q u a d r i f a s c i a t a populations. 51 In t h i s study the term " f i r s t - g e n e r a t i o n progeny" r e f e r s to the t o t a l number of f i r s t - i n s t a r larvae produced by the f i r s t adult generation, and the term -"second-generation progeny" r e f e r s to the t o t a l number of f i r s t - i n s t a r larvae produced by the second adult generation. F i r s t - and second-generation • •::' progeny can not be t o l d apart i n knapweed heads at the end of the growing season, and t h e i r d e n s i t i e s may d i f f e r greatly. Therefore the mortality for each f l y developmental stage was found only f or f i r s t - g e n e r a t i o n progeny by c o l l e c t i n g heads between f i r s t - and second- adult-generation emergence periods. 2.Method: On d i f f u s e and spotted knapweed release s i t e s U. a f f i n i s emergence i s syn-chronized with that of TJ. quadrif a s c i a t a (section II Bi) , therefore only one sample was taken between f i r s t - and second-generation peaks. .The time of t h i s sample varied from year to year depending on the date on which the l a s t , f i r s t - g e n e r a t i o n adults were seen, but most samples were taken i n the second week of August. This was one week a f t e r the l a s t few f i r s t - g e n e r a t i o n adults were seen to ensure that a l l f i r s t - g e n e r a t i o n progeny developed g a l l s . The samples were stored at room temperature for an a d d i t i o n a l week so a l l larvae, which were destined to pupate, had done so. The second sample was taken at the end of the growing season when second-generation adults had died o f f , usually by the end of September. Sampling was done by various methods. In 1975 four 50 X 50 cm square p l o t s on the release s i t e s were established i n May by throwing a square and de-l i n e a t i n g a p l o t where i t landed. Half of the heads were systematically harvested for the f i r s t sample, and the remainder comprised the second 52 sample. In 1976, 1977 and 1978, samples were e n t i r e primary branches of d i f f u s e knapweed and ent i r e stalks of spotted knapweed because both pro-duce approximately s i m i l a r numbers of heads. In 1976 heads were obtained from samples c o l l e c t e d f o r the previous study on g a l l - f l y synchrony with knapweed phenology, but i n 1977 and 1978 separate samples were c o l l e c t e d f o r t h i s study (N = 50 to 60 branches, and s t a l k s , f o r d i f f u s e and spotted • knapweed, r e s p e c t i v e l y ) . Heads were c l a s s i f i e d i n bud, seed (including those heads i n flower), superparasitized or undeveloped stages. Heads i n the seed stage were opened and g a l l s and seeds counted. G a l l s were i d e n t i f i e d as U_. a f f i n i s or U_. quadr i f a s c i a t a. G a l l contents were catagorized into one of the following classes: dead f i r s t - i n s t a r l a r v a a f t e r g a l l formation i f the g a l l was empty (found for TJ. j aceana by Varley, 1947, and assumed*'to occur here for U. af f i n i s ) ;• dead second- or t h i r d - i n s t a r larvae i f any dead l a r v a l remains were found (found also by Varley f o r IJ. j aceana); dead adult i f the adult had died before emergence from the g a l l or from the head; diapause ( t h i r d -i n s t a r larva) i f the l a r v a had eit h e r reached the t h i r d - i n s t a r stage or appeared healthy at a younger stage and would probably reach the t h i r d -i n s t a r ; pupa i f the larvae had pupated or was i n the process of doing so; or empty pupal case i f a second-generation adult had already emerged. The sum of pupae + empty pupal cases i n the f i r s t sample gave the number of f i r s t -generation progeny completing development to the adult stage i n one growing season. Superparasitized heads were distinguished from undeveloped heads as i n section II A i i b, and i t was assumed that only U_. a f f i n i s caused head sup e r p a r a s i t i z a t i o n . The number of dead U. a f f i n i s f i r s t - i n s t a r larvae i n the p r e - g a l l stage were assumed to be four and eight i n d i f f u s e and spotted 53 knapweed superparasitized heads, r e s p e c t i v e l y . These estimates are based on g a l l and seed counts at population peaks when there were sharp declines i n seed counts per head. Mean numbers of f i r s t - i n s t a r larvae per head were found f o r both TJ. a f f i n i s and TJ. qu a d r i f a s c i a t a on d i f f u s e and spotted knapweed as the estimate of f l y d e n s i t i e s i n heads. Included i n the term "head" were counts of seed and superparasitized heads. A value of n i l was given for JJ. qu a d r i f a s c i a t a when superparasitized heads were encountered. Angular transformations were made of the percentages of f l i e s i n each stage before analysis of variance. 3.Results: a ) f i r s t - g e n e r a t i o n TJ. a f f i n i s : The mean number of f i r s t - i n s t a r larvae per head-increased s i g n i f i c a n t l y from 1975 to 1977 on both d i f f u s e (Table VI) and spotted (Table VII) knapweed r e ' — - -lease s i t e s . From 1977 to 1978 there was a s i g n i f i c a n t decrease i n the mean number of f i r s t - i n s t a r larvae per head on the spotted knapweed s i t e , but there was no s i g n i f i c a n t d i f f e r e n c e between 1977 and 1978 on the d i f f u s e knapweed release s i t e . On the average, s i g n i f i c a n t l y fewer numbers of f i r s t -i n s t a r larvae were produced per head i n d i f f u s e compared to spotted knapweed 2 heads ( ^ (]_.).= 4.67). The percentage of f i r s t - i n s t a r larvae which died before g a l l formation i n superparasitized heads increased s i g n i f i c a n t l y from 1975 to 1977 on both d i f f u s e and spotted knapweed. S i m i l a r i l y to the number of f i r s t - i n s t a r larvae per head, the percentage of f i r s t - i n s t a r larvae which died before g a l l 54 Table VI. Average numbers of U. a f f i n i s per head on the d i f f u s e knapweed release s i t e from 1975 to 1978 SIG. 1 1975 1976 1977 1978 a b x no. of f i r s t - i n s t a r larvae per head .. 6 + .04 2.0 + .1 2.6 + .2 2.4 + .1 Lrst- dead f i f s f r r i n s t a r + + + Lrst-l arvae before ...... 0% 19 3% 22 4% 31 3% * o of fd head g a l l formation enerat: of fd head dead f i r s t - i n s t a r + + + enerat: • u D 0) 3 P-larvae a f t e r g a l l formation 0% 10 2% 6 2% 1.5 .5% * 60 dead second- and + + + 4-4-1 > m cd O rH t h i r d - i n s t a r 2 1% 2 1% 1 1% 1.5 T .5% 4—1 0) «_, larvae H •rl 14-1 stage instar t h i r d - i n s t a r larvae 87 + 2% 49 + 3% 47 + 4% 55 + 3% * % in pupae + empty pupal cases 11 + 2% 20 + 3% 24 + 3% 11 + 2% * * second o •H •U cd x no. of f i r s t - i n s t a r larvae per head .73 + .03 1.77 + .09 2.3 + .1 1.83 + .09 * * second 1 genet x no. of t h i r d - i n s t a r larvae per head .71 + .03 .82 + .06 .9 + .1 .46 + .04 * * X s i g n i f i c a n t d i f f e r e n c e among means from 1975 to 1978 (a) and between 1977 and 1978 (b) denoted by * formation decreased s i g n i f i c a n t l y from 1977 to 1978 on spotted, but not on d i f f u s e knapweed. The percentage of f i r s t - i n s t a r larvae which died a f t e r g a l l formation on d i f f u s e knapweed d i f f e r e d s i g n i f i c a n t l y with time, with the highest percentage occurring i n 1976 and a subsequent de c l i n e . u n t i l - 1978. On spotted knapweed the percentage of f i r s t - i n s t a r larvae which died a f t e r g a l l formation was not s i g n i f i c a n t l y d i f f e r e n t between years. There was no s i g -n i f i c a n t d i f f e r e n c e between years i n the percentage of second- and t h i r d -i n s t a r larvae which died, nor i n the percentage of dead adults, i n either d i f f u s e or spotted knapweed. 55 Table VII. Average numbers of TJ. a f f i n i s per head on the spotted knapweed release s i t e from 1975 to 1978 i SIG. 1975 .1976 1977 1978 a b X no. of f i r s t - i n s t a r larvae per head .9 ± .1 3.8 + .2 6.0 + .2 2.8 + .1 * * dead f i r s t - i n s t a r - 2% + I + larvae before 2 4 2% 23 ~r 3% 6 2% * * c i 4-1 cn g a l l formation o • H t-•r* dead f i r s t - i n s t a r + t + 4-1 cfl 14-cd larvae a f t e r 0% 3 1% 2.0 - r .7% 2.9 .9% u QJ r 4 4-1 QJ o g a l l formation t—< cu 60 •C M 0) dead second- and + i + cu cu ni 4-> > t h i r d - i n s t a r 0% 2 1% 1 - h 1% 2.2 .8% 4-1 larvae firs O i H CU M dead adults before emergence 0% 0% 0% .9 + .9% stag insta t h i r d - i n s t a r larvae 77 ±5% 76 + 3% 69 + 3% 74 + 2% c • H pupae + empty pupal cases 21 ± 5% 15 + 2% 5 + 1% 14 + 2% * A second C O H 4-1 tfl x no. of f i r s t - i n s t a r larvae per head 1.5 ± .1 3.9 + .2 4.9 + .2 2.6 + .1 * * second cu c QJ 60 x no. of f i r s t - i n s t a r larvae per head 1.33 ± .09 3.0 + .1 2.8 + .2 1.19 + .08 * * """significant d i f f e r e n c e among means from 1975 to 1978 (a) and between 1977 and 1978 (b) denoted by * The c o r r e l a t i o n between the density of f i r s t - i n s t a r larvae per head with the percentage of larvae which died .in the pre-^gall stage was not s i g n i f i c a n t in: eith e r d i f f u s e ,(r = 0.93, 'at 2 D.F.) or spotted knapweed (r = 0.97, at 2 -D.F.). .However, on spotted knapweed, as dens i t i e s of f i r s t - i n s t a r larvae increased from 1975 to 1977, the percentage of f i r s t - i n s t a r larvae which died before g a l l formation also increased, and i n 1978, when the f i r s t - i n s t a r l a r v a l density decreased, so did the percentage of larvae which died before g a l l formation (Fig. 16ii")'.i S i m i l a r i l y , as the density of f i r s t - i n s t a r '• %**0 -i 56 3 0 H 1978 X o 20 H 1976 X 1977 X LU O LL. LU 00 LU a: to CO CL i—i u_ O ft Q 10 4 %30 20H icH x i . -TT^ r— P .5 l.o 1975 X —l 1 1 f 1.5 2.0 2.5 3.0 iyz 1978 X 1976 X ~ r 3 T -1 2 .if 5 - 6 AVERAGE NUMBER OF FIRST-INSTAR U. AFFINIS .LARVAE PER HEAD Figure 16. Relationship between the percentage dead f i r s t - i n s t a r larvae which died before g a l l formation and the average number of f i r s t - i n s t a r larvae per head for f i r s t generations of U. a f f i n i s on d i f f u s e and spotted knapweed release s i t e s . i . d i f f u s e knapweed; i i . spotted knapweed 57 larvae i n d i f f u s e knapweed heads increased from 1975 to 1977, the mortality at the f i r s t - i n s t a r stage before g a l l formation increased ( Fig. 1 6 i ) . On d i f f u s e knapweed, between 1977 and 1978, there was no s i g n i f i c a n t d i f f e r e n c e between f i r s t - i n s t a r d e n s i t i e s per head, nor i n the percentage of f i r s t -i n s t a r larvae which died before g a l l formation. The percentage of t h i r d - i n s t a r larvae, or those larvae which tered d i a -pause i n mid-summer, decreased s i g n i f i c a n t l y from 87 - 2% i n 1975 to 49 -3% i n 1976 on d i f f u s e knapweed, and i n subsequent years did not d i f f e r s i g -n i f i c a n t l y with the percentage i n 1976. On spotted knapweed there was no s i g n i f i c a n t d i f f e r e n c e i n the percentage of t h i r d - i n s t a r larvae produced between years. The percentage of pupae plus empty pupal cases followed opposite trends on d i f f u s e compared to spotted knapweed. From 1975 to 1977 t h i s percentage increased s i g n i f i c a n t l y on diffuse knapweed and from 1977 to 1978 i t de-creased s i g n i f i c a n t l y . On spotted knapweed the percentage of pupae plus empty pupal cases decreased s i g n i f i c a n t l y from 1975 to 1977, and increased s i g n i f i c a n t l y from 1977 to 1978. Usually the percentage of f i r s t - g e n e r a t i o n progeny completing development to the adult stage did not exceed 20% on e i t h e r d i f f u s e or spotted knapweed. b ) f i r s t plus second U. a f f i n i s generations: The mean number of f i r s t - i n s t a r larvae per head increased s i g n i f i c a n t l y from 1975 to 1977 and decreased s i g n i f i c a n t l y from 1977 to 1978 on both d i f f u s e and spotted knapweed release s i t e s . There was no s i g n i f i c a n t d i f f e r e n c e i n the mean number of f i r s t - i n s t a r larvae produced per head between d i f f u s e and 2 spotted knapweed , ,= 2.14). However, i n most years there were about 58 twice as many f i r s t - i n s t a r larvae per spotted as per d i f f u s e knapweed head. The mean number of t h i r d - i n s t a r (diapause) larvae per head on d i f f u s e knap-weed increased s i g n i f i c a n t l y from 1975 to 1977 and decreased s i g n i f i c a n t l y from 1977 to 1978. No trend on spotted knapweed was observed, but there was a s i g n i f i c a n t d i f f e r e n c e i n numbers of t h i r d - i n s t a r larvae per head among years with the highest density i n 1977. c ) f i r s t - g e n e r a t i o n U. qu a d r i f a s c i a t a : The mean number of f i r s t - i n s t a r larvae per head decreased s i g n i f i c a n t l y from 1975 to 1978 on d i f f u s e knapweed (Table V I I I ) , but on the spotted knapweed release s i t e (Table IX), numbers remained low from 1975 to 1977, and then increased suddenly i n 1978 when U. a f f i n i s declined (Fig. 17). On the d i f f u s e knapweed release s i t e there was a s i g n i f i c a n t negative c o r r e l a t i o n between numbers of f i r s t - i n s t a r U_. qu a d r i f a s c i a t a and f i r s t - i n s t a r U. a f f i n i s larvae per head f o r f i r s t generations (Fig. 18). No empty U_. quadrifasciata g a l l s were found in- either d i f f u s e or spotted knapweed heads, and dead t h i r d - i n s t a r larvae and dead adults were found i n d i f f u s e knapweed only i n 1978 and 1976, r e s p e c t i v e l y . No dead U. qua d r i f a s c i a t a were found i n spotted knapweed heads. Therefore on both d i f f u s e and spotted knapweed the percen-tage of t h i r d - i n s t a r larvae followed opposite trends to the percentage of empty pupal cases. On -diffuse knapweed the percentage of pupating f l i e s r e -mained above 80% from 1975 to 1977, but f e l l to 45% i n 1978. On spotted knapweed 92% of the larvae pupated i n 1978 when U_. qua d r i f a s c i a t a was most dense, but i n previous years when U. qu a d r i f a s c i a t a was rare, the percentage of pupae varied from 67% to 100%. On the average the percentage of larvae which pupated was greater f o r U. quadrif a s c i a t a than f or TJ_. a f f i n i s on both s i t e s . 59 Table VIII. Average numbers of U. qua d r i f a s c i a t a per head on the d i f f u s e knapweed release s i t e from 1975 to 1978 i SIG. 1975 1976 1977 1978 a b a x no. of f i r s t - i n s t a r larvae per head .54 + .04 .24 + .05 .15 + .05 .10 + .03 * A teratic u a cu j-i .c cn 4-1 C TJ dead t h i r d - i n s t a r larvae 0% 0% 0% 9 1 CO 60 •H CO 14-1 1 CU O 4J X CD QJ U U dead adults before emergence 0% 4 + 2% 0% 0% first 60 -H CU nj 14-1 4J CO <4-4 0) o aj c > •H • U o cd N C H t h i r d - i n s t a r larvae pupae + empty pupal cases 12 88 + + 3% 3% 11 85 + + 6% 6% 19 81 + + 11% 11% 46 45 + + 14% 14% * A A A c o C 4-1 o cd x no. of f i r s t - i n s t a r larvae per head .29 + .02 .33 + .06 .41 + .07 .06 + .02 * A nCO CO x no. of t h i r d - i n s t a r + + + + cn c cu bC larvae per head .10 .01 .04 .01 .32 .07 .04 .01 A A s i g n i f i c a n t d i f f e r e n c e among means from 1975 to 1978 (a) and between 1977 and 1978 (b) denoted by * Table IX. Average numbers of U_. qu a d r i f a s c i a t a per head on the spotted knapweed release s i t e from 1975 to 1978 I SIG. 1975 1976 1977 1978 a b o •1-1 4-> X no. of f i r s t - i n s t a r larvae per head .01 +- .01 .15 + .06 .02 + .02 .49 + .06 A A genet stage t h i r d - i n s t a r larvae 0% 29 + 16% 33 + 33% 8 + 3% first % in pupae + empty pupal cases 100 - 100% 71 + 16% 67 + 33% 92 + 3% •u c a o •H 4-1 X no. of f i r s t - i n s t a r larvae per head 004 - .004 .05 + .02 .10 + .05 .62 + .07 A secc genei X no. of t h i r d - i n s t a r larvae per head 0 .01 + .01 .06 + .04 0 """significant d i f f e r e n c e among means from 1975 to 1978 (a) and between 1977 and 1978 (b) denoted by * CL < h-00 o CL LU CD CO CL LU CO CL •—1 LU O l Q ft X CL LU 0-LU < > CL • 6"i • 2 H 1978 X 1975 X, 60 1^76 X Figure 17. 1 2 3 4 5 6 AVERAGE NUMBER OF U. AFFINIS FIRST-INSTAR LARVAE PER HEAD Relationship between the average number of IJ. q u a d r i f a s c i a t a and U_. a f f i n i s f i r s t - i n s t a r larvae per head on spotted knapweed for f i r s t generations of f l i e s . CL co L L . O I CO Q CL CL < LU r-1 LU 00 U_ X < LU >-" o b < CO Q> < UJ U_ > >-< CL 3LU CL .6 1975 X - 1 — 1.0 T 1976 X 1977 1978 x * X 0.5 1.5 2.0 2.5 3.0 AVERAGE NUMBER OF U. AFFINIS FIRST-INSTAR LARVAE PER HEAD Figure 18. 'Correlation between the average number of U. qu a d r i f a s c i a t a and U. a f f i n i s f i r s t - i n s t a r larvae per head on d i f f u s e knapweed for f i r s t generations of f l i e s , (r = -.98; s i g n i f i c a n t at 2 D.F.) 61 d ) f i r s t plus second U. q u a d r i f a s c i a t a generations: There was a s i g n i f i c a n t d i f f e r e n c e i n the mean number of f i r s t - i n s t a r larvae which hatched per head by the end of the growing season on the d i f f u s e knapweed release s i t e : d e n s i t i e s increased from 1975 to 1977, but dropped s i g n i f i c a n t l y from 1977 to 1978. On the spotted knapweed release s i t e there was a s i g n i f i c a n t increase i n the mean number of f i r s t - i n s t a r larvae per head from 1975 to 1978. The mean number of t h i r d - i n s t a r larvae per head on the d i f f u s e knapweed s i t e varied s i g n i f i c a n t l y with time, but followed no trend. On spotted knapweed there was no s i g n i f i c a n t d i f f e r e n c e i n numbers of t h i r d - i n s t a r larvae per head from 1975 to 1978. In 1978, although there were an average of 0.49 - 0.06 g a l l s per head i n August on spotted knapweed, and 8% of them were t h i r d - i n s t a r larvae, no diapausing larvae were found i n September. 4.Discussion: Although there i s much controversy about theories of population regulation, most workers i n b i o l o g i c a l c o n t r o l f e e l that populations are regulated by density-dependent a c t i o n which "includes the actions of repressive environ-mental f a c t o r s , c o l l e c t i v e l y or s i n g l y , which so i n t e n s i f y as the population increases beyond a c h a r a c t e r i s t i c high l e v e l and so relax as density f a l l s that population increase beyond c h a r a c t e r i s t i c maxima i s prevented and de-crease to e x t i n c t i o n i s made les s l i k e l y . . . T h u s , a negative feed-back process between density and rate of increase i s involved" (Huffaker, et^. a l . , 1971). E a r l i e r , Varley (1947) showed that JJ. jaceana was regulated by t h i s process, and because U_. j aceana i s b i o l o g i c a l l y s i m i l a r to U_. a f f i n i s , much of t h i s discussion w i l l be based on Varley's (1947) paper. In t h i s study a more thorough i n v e s t i g a t i o n of TJ. a f f i n i s and U_. quadrif a s c i a t a m o rtality at each stage was not possible due to lack of time. Thus the following i s only a 62 p l a u s i b l e explanation of what factors may be regulating f l y populations; f a c t o r s , which may not be confirmed f o r several years u n t i l a " c h a r a c t e r i s t i c abundance", which most populations reach (Huffaker, e_t. a l . , 1971) i s attained. No s p e c i f i c predators of JJ. a f f i n i s were observed, but numerous spider webs with dead f l i e s i n them, were seen, mostly on d i f f u s e knapweed. Storey (1976) found that the spider, Dictyna major, preyed on U. a f f i n i s adults i n Montana, but he did not estimate the percentage of adults caught. I t i s l i k e l y that most spiders are non-specific predators, and i f g a l l f l i e s were to decline i n density, the spiders could catch the many other insects on the s i t e s (Watson, 1972) f o r food. Therefore spiders probably do not control f l y populations. Varley found that as many as 90% of JJ. j aceana g a l l s f e l l out during the winter, and that the larvae were probably eaten by mice.. Evidence of rodent a c t i v i t y on the release s i t e s was common: some plants were stripped of t h e i r seed heads and p i l e s of chaff and half-eaten seed heads were often found. Watson (1972) thought that mice fed on knapweed seeds, but the rodents prob-ably ate the larvae as w e l l . G a l l s could f a l l out when seed heads, e s p e c i a l l y spotted knapweed seed heads which open widely upon maturity, dehisce. Often JJ. a f f i n i s g a l l s i n spotted knapweed heads are not anchored to the receptacle. JJ. q u a d r i f a s c i a t a g a l l s i n spotted knapweed heads are usually held only by bracts, and few JJ. q u a d r i f a s c i a t a dipause larvae were found i n September i n spotted knapweed probably because they f e l l out. Preliminary work showed that many g a l l s can be found i n the s o i l i n the spring, and that a l l diapause larvae i n these g a l l s die before pupation. But probably the percentage . of g a l l s which f a l l out and are destroyed by predators i s independent of 63 f i r s t - i n s t a r l a r v a l d e n s i t i e s i n the p r e - g a l l stage, and f l y populations are not regulated by mortality at the t h i r d - i n s t a r stage. Egg mortality of U. j aceana was 9% and 15% i n 1935 and 1936, r e s p e c t i v e l y . Varley a t t r i b u t e d egg mortality to defective yolks, unmated females, or cross breeding with TJ. q u a d r i f a s c i a t a . He did not f i n d that egg m o r t a l i t y increased with increasing density of egg batches per head. Neither the cause of egg mortality, nor estimates of the death rate at the egg stage, were made for TJ. a f f i n i s or U_. q u a d r i f a s c i a t a . At high d e n s i t i e s unmated females would be rare, so t h i s factor probably does not contribute to population regulation. The only time cross-breeding between U_. a f f i n i s and U. q u a d r i f a s c i a t a was observed i n the f i e l d was when several hundred f l i e s were released together from a cage. Zwblfer (1974) found that g a l l f l i e s have elaborate behavioral mechanisms which prevent cross breeding. The e f f e c t of adult density on defective yolks i s not known. But, because Varley did not consider defective yolks to be a regulating factor for IJ. j aceana, TJ. a f f i n i s and TJ. q u a d r i f a s c i a t a probably are not regulated by t h i s f a c t o r . Only 2% to 3% of U. jaceana larvae died a f t e r g a l l formation. Varley a t t r i b u t e d death to "unknown causes" and f e l t that attack by b a c t e r i a and fungi of the larvae occurred a f t e r l a r v a l death. S i m i l a r i l y , few TJ. a f f i n i s larvae died a f t e r g a l l formation either i n the f i r s t - or during the second-and t h i r d - i n s t a r stages. The highest count was i n 1976, when 10% of the t o t a l number of f i r s t - i n s t a r U. a f f i n i s larvae died a f t e r g a l l formation i n d i f f u s e knapweed. Death i n the l a r v a l stage a f t e r g a l l formation was unrelated to f i r s t - i n s t a r d e n s i t i e s , e i t h e r for f i r s t - g e n e r a t i o n progeny, 64 or for f l y counts at the end of the growing season. Therefore from 1975 to 1978 death at second- and t h i r d - i n s t a r stages did not regulate JJ. a f f i n i s numbers. JJ. a f f i n i s populations are probably regulated by density-dependent competition for food and g a l l s i t e s before the f i r s t - i n s t a r larvae form g a l l s . Although co r r e l a t i o n s between the percentage of larvae which died before g a l l form-ati o n with f i r s t - i n s t a r d e n s i t i e s i n heads were not s i g n i f i c a n t , m o r t a l i t y at t h i s stage increased with l a r v a l density on both d i f f u s e and spotted knapweed from 1975 u n t i l 1977. From 1977 to 1978 both l a r v a l m o r t a l i t y and l a r v a l density i n heads decreased on the spotted knapweed s i t e and plateaued on the d i f f u s e knapweed s i t e . The reason for the plateau-effect on d i f f u s e knapweed may be the trampling of the knapweed during the winter, which was so severe that few plants were l e f t standing by the spring of 1978. However, Varley's p r e d i c t i o n of density-dependent competition between larvae at early stages of development as a c o n t r o l l i n g factor of U.j_a£eana pop-ul a t i o n s , i n the absence of natural enemies, appears to be true for JJ. a f f i n i s as w e l l . Probably i f JJ. a f f i n i s are able to i n i t i a t e gall-formation, most w i l l survive to the t h i r d i n s t a r . The f a c u l t a t i v e second generation of JJ. a f f i n i s confounds the i d e n t i f i c a t i o n of regulating factors of the population density. Varley did not report that JJ. j aceana produced two generations, but JJ. a f f i n i s may produce two generations both i n i t s countries of o r i g i n (ZwBlfer, 1970) and i n B r i t i s h Columbia. However, the proportion of f i r s t - g e n e r a t i o n progeny pupating i n August i s extremely v a r i a b l e . If the density of f i r s t - g e n e r a t i o n JJ. a f f i n i s adults i s high enough to superparasitize most of the heads, few larvae 65 survive to the the t h i r d i n s t a r and the s i z e of the second generation w i l l be small. Thus few a d d i t i o n a l diapause larvae w i l l be produced by the second generation, and the s i z e of f i r s t generation the following spring w i l l also be small. The rate of head s u p e r p a r a s i t i z a t i o n by the f i r s t IJ. a f f i n i s generation i s r e l a t e d to the number of heads a v a i l a b l e for o v i p o s i t i o n and the amount of food and space i n these heads. Average head s i z e may change from year to year, depending on weather conditions. Synchrony of f l y emergence with the appearance of heads i s also an important fa c t o r ; i f the f l i e s emerge s l i g h t l y e a r l i e r than heads appear, l a r v a l crowding due to high d e n s i t i e s of f l i e s i n r e l a t i o n to o v i p o s i t i o n s i t e s , can increase the expected rate of head s u p e r p a r a s i t i z a t i o n . These factors may explain why c o r r e l a t i o n s between superparasitized heads and l a r v a l d e n s i t i e s were not s i g n i f i c a n t . Densities of f i r s t - i n s t a r TJ. a f f i n i s larvae per head were about twice those i n spotted as i n d i f f u s e knapweed heads. On the average, there are 12.5 and 26.6 seeds per d i f f u s e and spotted knapweed head, r e s p e c t i v e l y , and each seed weighs 1.099 mg and 1.778 mg, r e s p e c t i v e l y (Watson, 1972). In 1977, at peak l a r v a l d e n s i t i e s , there were 2.3 - 0.1 and 4.9 - 0.2 f i r s t -i n s t a r TJ. a f f i n i s larvae per d i f f u s e and spotted knapweed head, r e s p e c t i v e l y , at the end of the second adult generation. Thus there were 5.97 mg and 9.65 mg t o t a l seed weight for d i f f u s e and spotted knapweed, r e s p e c t i v e l y , per one f i r s t - i n s t a r IJ. a f f i n i s l a r v a . In r e l a t i o n to a v a i l a b l e food, therefore, IJ. a f f i n i s was almost twice as dense on d i f f u s e as on spotted knapweed. This d i f f e r e n c e may have a r i s e n because greater d e n s i t i e s of f l i e s can coexist on a knapweed species, such as d i f f u s e knapweed, which compartmentalizes seeds 66 into smaller, more numerous units ( i . e . heads) than species which package seeds into fewer, larger u n i t s . In the laboratory the rate of d i f f u s e and spotted knapweed head maturation was the same. If the rates are s i m i l a r i n the f i e l d , then more eggs should be deposited i n spotted than i n d i f f u s e knapweed to achieve the same resource u t i l i z a t i o n . G a l l f l i e s appear to a t t a i n very high adult d e n s i t i e s i n the f i e l d i n r e l a t i o n to o v i p o s i t i o n s i t e s , but some behavioral mechanism must be l i m i t i n g d e n s i t i e s on spotted knapweed such that the f l i e s are not as e f f i c i e n t i n reducing spotted knapweed seed as they are i n reducing seed on the d i f f u s e knapweed s i t e . JJ. q u a d r i f a s c i a t a l a r v a l d e n s i t i e s also appear to depend on many f a c t o r s , so that changes i n t h i s insect's density are not e a s i l y explained. At d e n s i t i e s of more than 0.5 f i r s t - i n s t a r JJ. a f f i n i s larvae per head, JJ. quadrifasciata appears to be suppressed. At very high JJ. a f f i n i s d e n s i t i e s o v i p o s i t i o n s i t e s are saturated, or even superparasitized, because U. a f f i n i s can u t i l i z e knapweed heads at younger stages of development than JJ. quadri- f a s c i a t a . In t h i s s i t u a t i o n , the s u r v i v a l of JJ. quadrifasciata depends heavily on the success of the second adult generation; hot dry weather conditions i n August may reduce the number of heads i n the bud stage, thus threatening JJ. quadrifasciata with e x t i n c t i o n i n a population. This g a l l - f l y species, however, appears to be capable of increasing much more ra p i d l y than JJ. a f f i n i s from very low adult d e n s i t i e s . When conditions are r i g h t , few JJ. a f f i n i s and p l e n t i f u l o v i p o s i t i o n s i t e s , JJ. quadrif a s c i a t a 67 numbers can explode, only to crash when JJ. a f f i n i s catches up. TJ'.. quadri- f a s c i a t a may have many more s u r v i v a l t a c t i c s , adapted to s u r v i v a l i n the same population with more competitive species such as U. a f f i n i s and JJ. jaceana. The f i r s t U. q u a d r i f a s c i a t a l a r v a l i n s t a r causes only a s l i g h t swelling of the ovary w a l l , and i f the l a r v a dies, developing seeds or TJ. a f f i n i s g a l l s could crush immature TJ. q u a d r i f a s c i a t a g a l l s . Some immature TJ. quadrifas- c i a t a g a l l s could be mistaken for non-viable seeds. I t may be, however, that the m o r t a l i t y before and a f t e r g a l l formation i s indeed very low, as found i n t h i s study. Heads with'as many as f i v e JJ. q u a d r i f a s c i a t a g a l l s even i n d i f f u s e knapweed, usually do.jnot ,'appear swollen, and usually seeds are produced. This indicates that JJ. q u a d r i f a s c i a t a may have mechanisms which prevent i n t r a s p e c i f i c competition. Preliminary work showed JJ. quadri- f a s c i a t a to be much more a c t i v e than TJ. a f f i n i s , and d i s p e r s a l may be impor-tant i n the population regulation of t h i s species. 5.Conclusions: 1.Density-dependent l a r v a l m o r tality i n the p r e - g a l l stage probably r e s t r i c t s JJ. a f f i n i s numbers i n B r i t i s h Columbia. 2.On the average JJ. a f f i n i s d e n s i t i e s were greater i n spotted than i n d i f f u s e knapweed heads, but i n r e l a t i o n to net seed weight per head, JJ. a f f i n i s was denser i n d i f f u s e than i n spotted knapweed heads. 3.JJ. q u a d r i f a s c i a t a appears to be suppressed by U. a f f i n i s when JJ. a f f i n i s d e n s i t i e s exceed 0.5 f i r s t - i n s t a r larvae per head, but the regulation of JJ. q u a d r i f a s c i a t a appears to depend on many other f a c t o r s . 68 i i i . The r o l e of heterogeneity i n d i f f u s e knapweed head s i z e on g a l l - f l y  abundance: Habitats of natural animal populations can be very heterogeneous with r e s -pect to microweather, food, shelter and natural enemies (den Boer, 1971; Birch, 1971; Pajunen, 1971 and Dempster, 1971). G a l l - f l y larvae cannot leave the heads i n which they have hatched and the p r o b a b i l i t y of l a r v a l s u r v i v a l may depend not only on numbers hatched per head but also on the amount of o v a r i o l e tissue per head, and the s i z e of the ovaries. On spotted knapweed the average number of seeds per seed head increased s i g n i f i c a n t l y (r = .68 at 53 D.F.) with head diameter when unattacked terminal heads on the spotted knapweed release s i t e i n 1975 were measured. On the d i f f u s e knapweed release s i t e some plants are "small-headed", producing only short, narrow heads, while other plants are "large-headed", producing long, wide heads, where head s i z e i s a measure of the head i n the bud stage j u s t before the head opens to flower. Probably head s i z e for d i f f u s e knapweed i s a good i n d i c a t i o n of net ovariole weight, and/or numbers of seeds per head as i t i s for spotted knapweed. This study assumes that there i s l e s s food and space for g a l l f l i e s i n small- than i n large-headed d i f f u s e knapweed, and compares g a l l - f l y attack on the release s i t e s between these two plant types. 2.Method: In July, 1977, f i v e small- and f i v e large-headed d i f f u s e knapweed plants were chosen and tagged. Heads on the small-headed plants reached no more than 8 mm long and heads on large-headed plants reached an average length of 12 mm at anthesis. A l l the plants were of s i m i l a r s i z e and shape, had the same numbers of heads, were at s i m i l a r stages of development, and a l l had white flowers. In September, the plants were harvested and the heads were - ... - 69 systematically removed from the branches, and c l a s s i f i e d according to branch p o s i t i o n , and stage of development (seed, superparasitized or undeveloped) as previously. Superparasitized heads were assumed to contain four dead f i r s t - i n s t a r TJ. a f f i n i s larvae i n the p r e - g a l l stage, as i n the previous section. Seed heads were opened and U_. a f f i n i s and U. qua d r i f a s c i a t a g a l l s and seeds were counted. G a l l - f l y developmental stages were c l a s s i f i e d as t h i r d - i n s t a r larvae, emerged second generation adults, dead f i r s t - i n s t a r larvae a f t e r g a l l formation, dead second- or t h i r d - i n s t a r larvae, or dead adults i n section II B i i . In a n a l y s i s , the average number of f l i e s per head for U_. a f f i n i s and TJ. q u a d r i f a s c i a t a were calculated f o r each stage, where "head" equalled the sum of seed and superparasitized heads for both f l y species. 3.Results: Greater numbers of heads were superparasitized on small- than on large-headed plants, and although the differe n c e was not s i g n i f i c a n t , about twice as many heads were undeveloped on small- (62 - 13) than on lar g e - (36 - 10) headed plants (Table X). Altogether 70% of the t o t a l heads on small-headed plants, but only 44% of the t o t a l heads on large-headed plants, were either superparasitized or undeveloped. The appearance of small-headed plants was s t r i k i n g l y a l t e r e d because most of the heads i n d i s t a l p o s i t i o n s were super-p a r a s i t i z e d or undeveloped, but those i n proximal posi t i o n s developed to the seed stage (Fig. 19). Large-headed plants resembled knapweed growing on rangeland where there were no g a l l f l i e s , with d i s t a l heads developing to the seed stage. These heads, however, were swollen with many g a l l s , and there appeared to be an increase i n the number of proximal heads which r e -mained undeveloped compared to con t r o l rangeland knapweed. 70 Table X. Comparison of average numbers of TJ. a f f i n i s and U. quadrifasciata per head and average head numbers per small- and large-headed d i f f u s e knapweed plants on the release s i t e i n 1977. small-headed large-headed SIG.'1 t o t a l heads 141 - 22 95 - 12 seed heads 4 0 - 6 5 2 - 6 undeveloped heads 62 - 13 36 - 10 superparasitized heads 3 8 - 6 7 - 3 % superparasitized heads of t o t a l heads 27.1 - .7 7 - 2 % undeveloped heads of t o t a l heads 4 3 - 4 3 7 - 7 seeds per seed head 4.1 - .6 3 - 1 f i r s t - i n s t a r larvae 2.9 - .6 2.8 - .4 t h i r d - i n s t a r larvae .34 - .08 1.4 - .2 * empty pupal cases .02 - .01 .50 - .08 dead f i r s t - i n s t a r larvae before „ , + , c + „ j- ^ . Z . 4 - . 6 . 5 - . 2 g a l l formation dead f i r s t - i n s t a r larvae a f t e r n , + _„ n , + T-, R . .04 - .02 .06 - .03 g a l l formation dead second- and t h i r d - i n s t a r larvae .04 ^ .02 .09 - .06 dead adults .002 - .002 .3 - .1 u CO 3 f i r s t - i n s t a r larvae .4 - .1 .21 - .08 t h i r d - i n s t a r larvae .3 - .1 .05 - .03 empty pupal cases .07 - .05 .13 - .05 dead second- and t h i r d - i n s t a r larvae .01 - .01 0 dead pupae 0 .004 - .004 dead adults .004 - .004 .01 - .01 """significant d i f f e r e n c e between small- and large-headed plants denoted by * Most small-headed plants produced new heads u n t i l l a t e i n the growing season, but large-headed plants usually ceased head i n i t i a t i o n by the middle of ." August. Although the diff e r e n c e between means was not s i g n i f i c a n t , i t appears that small-headed plants i n i t i a t e greater numbers of heads (141 - 22) than large-headed plants (95 - 12). There was no s i g n i f i c a n t d i f f e r e n c e i n the 72 average number of seed heads produced by small- (40 - 6) and l a r g e - (52 - 6) headed plants, but s i g n i f i c a n t l y greater numbers of seed heads for a l l f i v e r e p l i c a t e s had no g a l l s i n small-headed plants (55) compared with l a r g e -2 headed plants (22; % 1 0 . 1 8 ) . There was no s i g n i f i c a n t d i f f e r e n c e i n the mean number of seeds per seed head between small- and large-headed plants. The mean number of f i r s t - i n s t a r U_. a f f i n i s larvae per head did not d i f f e r s i g n i f i c a n t l y between small- and large-headed plants. However, there were s i g n i f i c a n t l y fewer dead IJ. a f f i n i s f i r s t - i n s t a r larvae before g a l l form-. ..• ation, and s i g n i f i c a n t l y more IJ. a f f i n i s t h i r d - i n s t a r larvae, empty pupal cases and dead adults per head i n large- than i n small-headed plants. The-dif-ference between means,1 was not s i g n i f i c a n t , but there i s a strong i n d i c a t i o n that the average numbers of U. q u a d r i f a s c i a t a larvae per head are greater i n small- (0.4 - 0.1) than i n l a r g e - (0.21 - 0.08) headed plants. This i s supported by f i n d i n g s i g n i f i c a n t l y more t h i r d - i n s t a r U_. q u a d r i f a s c i a t a larvae per head i n small- than i n large-headed plants, and no s i g n i f i c a n t d i f f e r e n c e between la r g e - and small-headed plants for the remaining U_. q u a d r i f a s c i a t a stages. 4.Discussion: From estimates of f i r s t - i n s t a r d e n s i t i e s per head, i t appears that U. a f f i n i s o v i p o s i t s s i m i l a r numbers of eggs i n small- and large-headed d i f f u s e knapweed. But, because there i s l e s s food and space i n "small" heads than i n " l a r g e " heads', given both heads are at the same maturiy, the small heads are more l i k e l y to be superparasitized, but large heads usually develop to the seed stage, producing not only U. a f f i n i s g a l l s but also mature seeds. There i s great egg wastage i n small-headed d i f f u s e knapweed plants at high popu-73 l a t i o n d e n s i t i e s , such as i n 1977, when 83% of U. a f f i n i s were estimated to die before g a l l formation, compared to only 23% i n large-headed plants. On the spotted knapweed release site,' i n 1977, when U_. a f f i n i s was at peak numbers, 15% of TJ. a f f i n i s larvae were estimated to die i n the p r e - g a l l stage (section II B i i ) . Spotted knapweed plants which produced small heads were also more susceptible to su p e r p a r a s i t i z a t i o n by U_. a f f i n i s (Fig. 20) than pOsants producing larger heads. Heterogeneity i n d i f f u s e knapweed head s i z e i s b e n e f i c i a l to both the g a l l f l i e s and the knapweed. Some U. a f f i n i s w i l l always survive i n large-headed plants even when average mortality i n the p r e - g a l l stage i s high. This prevents & sudden e x t i n c t i o n of the population. Heterogeneity i n head s i z e may also enable U_. quadrifasciata to p e r s i s t with TJ. a f f i n i s because high d e n s i t i e s of TJ. q u a d r i f a s c i a t a were found i n small-headed plants, probably the progeny of the second generation. This aspect i s further discussed i n the following section. Knapweed benefits from heterogeneity i n head s i z e because some seeds are produced by large-headed plants even at high f l y de n s i t i e s early i n the growing season. Late i n the growing season small-headed plants compensate fo r head su p e r p a r a s i t i z a t i o n during the f i r s t generation by producing s i m i l a r numbers of seeds per plant (40 seed heads per plant X 4.1 seeds per head = 164 seeds per plant) to large-headed plants (52 seed heads per plant X 3 seeds per head = 156 seeds per p l a n t ) . Thus heterogeneity i n d i f f u s e knap-weed head s i z e may be a mechanism f or s u r v i v a l of t h i s weed i n countries of o r i g i n , where i t i s attacked by many head-feeding in s e c t s . Because hetero-geneity i n head s i z e occurs i n knapweed not attacked by f l i e s , and among 74 plants grown i n the greenhouse from f i e l d - c o l l e c t e d rosettes, the basis f o r t h i s v a r i a t i o n i s probably genetic. Dry weather conditions may decrease the average head s i z e i n f i e l d populations, but heterogeneity i n head s i z e would probably s t i l l e x i s t . 5.Conclusions: 1. JJ. a f f i n i s does not appear to show an o v i p o s i t i o n a l preference for small-or large-headed d i f f u s e knapweed. 2. JJ. q u a d r i f a s c i a t a i s more successful i n small- than i n large-headed d i f f u s e knapweed. 3.Similar numbers of seeds per plant were produced by small- and la r g e -headed plants, even though high percentages of heads on small-headed plants were superparasitized or undeveloped. 4.Small-headed d i f f u s e knapweed i s more susceptible to superparasitism than i s large-headed d i f f u s e knapweed. i v . Parasitism of d i f f u s e knapweed by f i r s t and second generations of g a l l f l i e s i n 1977: 1.Introduction: E a r l i e r (section II B i ) i t was shown that on the d i f f u s e knapweed release s i t e i n 1977, when g a l l f l i e s were at peak d e n s i t i e s , the r a t i o of adult f l i e s to heads acceptable f o r o v i p o s i t i o n was s i m i l a r for both f i r s t - and second-generation f l i e s . This study determines i f attack rates are s i m i l a r during f i r s t and second generations. In the previous section, i t appeared that great numbers of heads on small-headed d i f f u s e knapweed plants were superparasitized by f i r s t - g e n e r a t i o n JJ. a f f i n i s , but that large-headed 75 plants may produce seed during the f i r s t generation. This study further compares small- and large-headed d i f f u s e knapweed reaction to g a l l - f l y attack on the release s i t e i n 1977. 2.Method: The six 50 X 50 cm square p l o t s on the d i f f u s e knapweed release s i t e , which had been used to monitor adult f l y d e n s i t i e s (section II B i i ) were used for t h i s study. On August 1st, 1977, between f i r s t and second adult gen-erations, the number of heads per plant were counted, and c l a s s i f i e d as seed, superparasitized, undeveloped, and bud stages as i n section II B i i . A l l the heads i n the bud stage were assumed to be a v a i l a b l e for attack by the second generation, but heads i n the remaining stages were assumed to have been attacked only by the f i r s t generation of adults. The plants were harvested i n September, and each head was coded as i n section II A i , accor-ding to i t s p o s i t i o n on branches, and c l a s s i f i e d as being i n seed, undeve-loped or superparasitized stages. Heads i n the seed stage were opened, and numbers of seeds and JJ. a f f i n i s and JJ. quadrif a s c i a t a g a l l s were counted. G a l l contents were c l a s s i f i e d as i n section II B i i . I t was assumed, as i n section II B i i that only U. a f f i n i s caused head sup e r p a r a s i t i z a t i o n , and that superparasitizediVheads contained four dead f i r s t - i n s t a r larvae i n the p r e - g a l l formation stage. Each plant skeleton was sketched, and the heads were indicated on the branches by using d i f f e r e n t symbols for each head stage. Knowing the number of heads i n each stage on August 1st, and assuming that heads, which ap-peared during the f i r s t adult generaiton occupied d i s t a l positions on bran-ches, i t was possible to i d e n t i f y "heads which developed during each gener-a t i o n on the sketches. Then the average number of seeds and f l i e s i n heads 76 attacked by f i r s t adult generations were compared to heads attacked by second adult generations. Sizes of the f i r s t U. a f f i n i s and U. q u a d r i f a s c i a t a adult generations per plot were found from data c o l l e c t e d for e a r l i e r sections (Tables VI and V I I I ) . Average numbers of t h i r d - i n s t a r larvae per seed head were m u l t i p l i e d by the average number of seed heads per plo t on the s i t e i n the f a l l of 1976 to give absolute f i r s t - g e n e r a t i o n population estimates. The sizes of the se-cond generations were estimated from the average numbers of empty pupal cases per p l o t , produced by the f i r s t adult generation. Sex r a t i o s f o r each generation were found from counts of males and females i n section II B i i . An estimate of fecundity was obtained by d i v i d i n g the average number of f i r s t - i n s t a r larvae per plo t by the average number of females per plo t for each generation. This assumes no death or movement of adults. Six small-and s i x large-headed plants from the p l o t s were a r b i t r a r i l y chosen for head and g a l l comparisons between f i r s t and second generations. 3.Results: There was no s i g n i f i c a n t d i f f e r e n c e i n the average number of t o t a l heads per plot between f i r s t and second generations (Table XI). But, s i g n i f i c a n t l y more seed heads were produced during the second than the f i r s t generation, with fewer superparasitized and undeveloped heads being produced i n the second than i n the f i r s t generation. There was no s i g n i f i c a n t d i f f e r e n c e between generations i n the average number of seeds per seed head, but be-cause more seed heads were being produced during the second generation, i t appears that more seeds per plo t were produced during the second than the f i r s t generation, even though t h i s d i f f e r e n c e was not s i g n i f i c a n t . F i r s t -77 Table XI. Comparison of average head and g a l l production during f i r s t and second f l y generations on the d i f f u s e knapweed release s i t e i n 1977 f i r s t second generation generation SIG. t o t a l heads per p l o t % seed heads % superparasitized heads % undeveloped heads seeds per seed head seeds per pl o t 322 28 19 53 3 232 t 66 +- 5 ±5 ±3 +- 78 215 - 61 6 9 - 3 4 + 2. -27 - 3 3 + 1 620 - 279 A t o t a l f i r s t - i n s t a r larvae per pl o t 422 + 136 . 135 + 52 * f i r s t - i n s t a r larvae per head 2.6 + .4 .9 + .2 A CO •H % t h i r d - i n s t a r larvae 24 + 5% 78 + 2% A fin % empty pupal cases 12 + 4% 6 + 2% m cd % dead p r e - g a l l , f i r s t - i n s t a r larvae 60 + 9% 5 + 2% P i % dead f i r s t - i n s t a r larvae a f t e r g a l l formation 2.0 + .8% 7 + 1% % dead second- and t h i r d - i n s t a r larvae .7 + .3% 3 + 1% A % dead adults before emergence 1.3 + .4% 1.0 + .4% iata t o t a l f i r s t - i n s t a r larvae per pl o t 24 + 8 93 + 41 o CO CCl f i r s t - i n s t a r larvae per head .15 + .04 .4 + .1 rif, % t h i r d - i n s t a r larvae 47 + 13% 97 + 1% A uad % empty pupal cases 47 + 14% 2 + 1% cr % dead t h i r d - i n s t a r larvae .5 + .5% .4 + .4% o l % dead adults 5.5 + 3% .6 + .6% s i g n i f i c a n t d i f f e r e n c e between f i r s t and second generations denoted by * i n s t a r TJ.' a f f i n i s ; d e n s i t i e s per p l o t and per head were s i g n i f i c a n t l y greater during the f i r s t than the second generation1 Because "the rate of head s u p e r p a r a s i t i z a t i o n was high during the f i r s t generation, 60 - 9% of TJ. a f f i n i s f i r s t - i n s t a r larvae died before g a l l formation'. Therefore s i g -n i f i c a n t l y more t h i r d - i n s t a r U_. a f f i n i s larvae were found i n heads produced during the second generation, when the o v e r a l l death rate was low. Some empty pupal cases were found i n heads designated as having been produced 78 during the second generation (6 - 2%), i n d i c a t i n g that there was some error by t h i s method i n estimating first-vahdrsecond-generation f l y and head d e n s i t i e s . U. a f f i n i s l a r v a l death a f t e r g a l l formation appeared to be higher during the second than the f i r s t generation, but the d i f f e r e n c e was s i g n i f i c a n t only for second- and t h i r d - i n s t a r larvae. A few dead adults i n g a l l s were found for both generations. Because U. quadrifasciata d e n s i t i e s per head were low during the f i r s t gen-eration (.15 - .04 per head), comparisons between f i r s t and second gener-ations are d i f f i c u l t to make. There was no s i g n i f i c a n t difference between the average number of U. q u a d r i f a s c i a t a f i r s t - i n s t a r larvae p e r - p l o t , and per head, but t h i s species appeared to be more common i n heads produced during the second than the f i r s t genration. About 50% U. quadrifasciata larvae pupated during the f i r s t generation, and death rates i n the remaining stages were low. During the second generation death rates were also low, and most U. q u a d r i f a s c i a t a (97 - 1%) reached the t h i r d - i n s t a r stage. A few empty pupal cases (2 - 1%) were found i n heads designated as being produced during the second generation, i n d i c a t i n g , again, some error by this-method. Average numbers of empty pupal cases per p l o t were found for both f l y species by the addition of seed heads containing empty punal cases for both f i r s t and second generations. These estimates gave the average number of f l i e s r. emerging per p l o t during the second generation (Table XII). Male:female sex r a t i o s were 1.0, or greater, for both generations. Fecundity, defined here as the number of f i r s t - i n s t a r larvae produced per female, did not d i f f e r s i g n i f i c a n t l y between f i r s t (9.63) and second (7.96) U. a f f i n i s generations 79 Table XII. Estimate of f i r s t - and second-generation g a l l - f l y f e c u n dities on d i f f u s e knapweed i n 1977 A x no. of a d u l t s 1 no. of females x no. o f f i r s t -k AA G emerged per sex emerged p e r i n s t a r l a r v a e F p l o t r a t i o p l o t p e r p l o t a f f i n i s 1 2 114 ± 38 3 9 - 7 1.6 1.3 43.84 16.96 422 ^ 136 135.- 52 9.63 7.96 U. q u a d r i -f a s c i a t a 1 2 6 ± 3 1 2 - 4 1.0 2.0 3 4 24 J 8 93 - 41 8.0 23.25 A f i r s t generation (1) and second generation (2) AA F = fecundity (number of f i r s t - i n s t a r larvae produced per female) """from counts of diapause larvae i n the f a l l of 1976 tfrom counts of adult f l i e s i n section II B i i from table XI 2 TC^J^- 0.16, but was s i g n i f i c a n t l y greater during the second U. quadrifa sciata 2 generation (23.25) than the f i r s t (8.0; 7.44). 2 + S i g n i f i c a n t l y more ( X^y= 355.17) f i r s t - i n s t a r U_. a f f i n i s larvae (422 - 136) than f i r s t - i n s t a r TJ. q u a d r i f a s c i a t a larvae (24 - 8) were produced by f i r s t -generation adults. But, i f the number of progeny which survived to the t h i r d - i n s t a r and older stages (157 and 23.9 for U. a f f i n i s and U. quadri- f a s c i a t a , r e s p e c t i v e l y ) are divided by the number of females per p l o t during the f i r s t generation (43.84 and 3 for U_. a f f i n i s and TJ. q u a d r i f a s c i a t a , r e s -2 p e c t i v e l y ) , there i s no s i g n i f i c a n t d i f f e r e n c e ( X?1.65) i n the average number of progeny produced by f i r s t - g e n e r a t i o n adults (3 ;59 and.7.i96 for U. a f f i n i s and TJ. q u a d r i f a s c i a t a , r e s p e c t i v e l y ) . Again, during the second gen-2 e r a t i o n , s i g n i f i c a n t l y more ( X^^= 7.74) U. a f f i n i s (135) than IJ. q u a d r i f a s c i a t a (93) f i r s t - i n s t a r larvae were produced. But, the r a t i o of second-generation progeny which survived to or past the t h i r d i n s t a r : the number of second-generation females which emerged per p l o t was s i g n i f i c a n t l y 80 2 greater ( ^ ^ ) ~ ^«'98) for IJ. quadrif a s c i a t a (23.16) than for U. a f f i n i s (6.77). Small-headed plants produced almost equal numbers of heads during both gen-erations, but large-headed plants produced 75 - 12% of t h e i r heads during the f i r s t generation (Table XIII). Only 4 - 2% of the heads reached the seed stage during the f i r s t generation for small-headed plants, whereas about h a l f of the heads reached the seed stage f o r large-headed plants. During the second generation most of the heads reached the seed stage on small-headed plants, and about 61 - 10% did so for large-headed plants. As a r e s u l t greater numbers of seeds per plant, and per seed head, were pro-duced during the second than during the f i r s t generation by small-headed plants, but more seeds per large-headed plants were produced during the f i r s t generation. Average numbers of seeds per seed head, and U. a f f i n i s and U. qu a d r i f a s c i a t a per head were s i m i l a r for the two generations for large-headed plants. 95 -2% of the t o t a l number of U_. a f f i n i s f i r s t - i n s t a r larvae i n small-headed plants died i n superparasitized heads during the f i r s t generation compared to only 31 - 6% i n large-headed plants. As a r e s u l t few U. a f f i n i s t h i r d -instari larvae were produced i n small-headed plants during the f i r s t gener-at i o n , but 79 - 4% were produced during the second generation. Large-headed A plants produced s i m i l a r numbers of t h i r d - i n s t a r larvae during the f i r s t (43 -7%) and second (64 - 15%) generations. Large-headed plants also produced the second generation: 16 - 4% U. a f f i n i s empty pupal cases were found i n l a r g e -headed plants compared to only 2 - 1% i n small-headed plants. A f t e r g a l l formation, m o r t a l i t y rates were greater during the second than the f i r s t generation i n both small- and large-headed plants. In small-headed plants 81 Table XIII. Comparison of g a l l - f l y attack of .'small- and large-headed d i f f u s e knapweed during f i r s t and second generations i head f i r s t second size generation generation t o t a l heads per plant s L 75 44 + I+I 12 7 73 14 % seed heads * S L 4 47 + + 2 5 73 61 % undeveloped heads S L 57 46 + + 4 6 26 35 u % superparasitized heads * S L 39 7 + + 5 2 .9 4 seeds per plant •k S L 4 45 + + 2 12 190 9 seeds per seed head S L .8 1.4 + .4 .7 3.7 1.3 f i r s t - i n s t a r larvae per head S L 2.8 2.5 + + .6 .4 .5 2.1 CO •H a % t h i r d - i n s t a r larvae * S L 2 43 + + 1% 7% 79 64 ± 4% - 15% affi % empty pupal cases S L 2 16 + + 1% 4% 2 11 J 1% - 8% % dead f i r s t - i n s t a r larvae i n the p f e - r g a l l .'stage: . ' * S L 95 31 + + 2% 6% 6 8 l « - 8% % dead i n remaining stages * S L 1.0 10 + + .5% 2% 13 17 J 3% - 8% cfl •u f i r s t - i n s t a r larvae per head S L .3 .2 + + .2 .1 .9 .03 !•* - .03 ascia % t h i r d - i n s t a r larvae A S L 58 49 + + 42% 21% 98 100 J 7% - 100% .adrif % empty pupal cases S L 42 27 + + 42% 15% 2 ± 1% 0% c % dead i n remaining stages S L 24 24% 0% 0% """significant d i f f e r e n c e according to a 2 X 2 X- test of a s s o c i a t i o n denoted by * S = small-headed plants; L = large-headed plants 82 most JJ. a f f i n i s died i n the f i r s t - i n s t a r stage a f t e r g a l l formation, but i n large-headed plants most died i n the adult stage before emergence. Although the d i f f e r e n c e was not s i g n i f i c a n t , i t appears that more JJ. quadri- f a s c i a t a were produced during the second generation (.9 - .2) than during the f i r s t generation (.3 - .2) i n small-headed plants. JJ. quadrifasciata i n large-headed plants were rare. Due to low numbers of JJ. q u a d r i f a s c i a t a found i n both small- and large-headed plants, estimates of percentages of JJ. q u a d r ifasciata i n each stage had large standard e r r o r s . From these data i t appears that some second-generation adults emerge from both small- and large-headed plants, and that these f l i e s mostly attack small-headed plants. Their progeny appear to have very low death rates. 4.Discussion: If g a l l - f l y m o r t a l i t y at the egg stage i s not density-dependent so that estimates of f i r s t - i n s t a r l a r v a l d e n s i t i e s can be taken as a measure of attack-rate, JJ. a f f i n i s i s l e s s e f f i c i e n t during the second than the f i r s t generation. JJ. qua d r i f a s c i a t a, i n contrast, i s more successful during the second than the f i r s t generation. Heterogeneity i n head s i z e among d i f f u s e knapweed plants on the release s i t e created great v a r i a b i l i t y i n f l y and seed counts when comparisons were made between f i r s t and second generations. On the average, when f l y d e n s i t i e s are high, such as i n 1977 on d i f f u s e knapweed, seeds are produced l a t e i n the growing season because most heads during the f i r s t generation are attacked so heavily that they are superparasitized, and high proportions of v unattacked heads remain undeveloped. Seed y i e l d s during the second generation may be high because JJ. a f f i n i s does not appear to be as e f f i c i e n t i n attacking heads 83 as during the f i r s t generation. And, even though d e n s i t i e s of U_. quadri- f a s c i a t a are much higher during the second than the f i r s t generation, t h i s species appears to have about one-half the e f f e c t on seed y i e l d as does U. a f f i n i s because U. qua d r i f a s c i a t a produces thinner g a l l s and has a smaller body s i z e . U_. qua d r i f a s c i a t a survive the great competition from U. a f f i n i s during the f i r s t generation by leaving a few progeny i n large-headed plants, where there i s food and space enough for both species and often seeds too, and by o v i p o s i t i n g i n the few heads on small-headed plants that escape super-parasitism by U. a f f i n i s . During the second generation TJ. q u a d r i f a s c i a t a takes advantage of less competition by U. a f f i n i s and the resurgence i n head i n i t i a t i o n by small-headed plants, and the r e s u l t i s that U. quadri- f a s c i a t a appears to have a higher fecundity during the second than the f i r s t generation. Other "factors thanrhead.size may be influ e n c i n g g a l l - f l y d e n s i t i e s i n knap-weed heads. Plant s i z e , spine and bract length and colour, and i n t e r n a l a t t r i b u t e s of the heads which the f l i e s determine by probing, may account for some of the v a r i a b i l i t y i n the data. Heterogeneity i n these factors and head s i z e probably a f f e c t s g a l l - f l y attack rate on spotted knapweed s i m i -l a r i l y to d i f f u s e knapweed. I t appears that l i t t l e energy i s expended by d i f f u s e knapweed i n developing heads to the 3 to 4 mm long bud stage, which i s most susceptible to super-parasitism by TJ. a f f i n i s . Because g a l l s and seeds do not develop i n these heads, energy,by small-headed plants at high f l y a t t a c k - r a t e s , i s a c t u a l l y v 84 conserved, and used l a t e r i n the growing season when JJ. a f f i n i s attack-rate i s lower. A s i m i l a r e f f e c t was found by Tanskiy (1969) who removed terminal cotton buds, and determined that head production during the l a t t e r part of the growing season was stimulated with no loss in' cotton y i e l d , but with a delay i n plant maturation. Some species of plants, whose flowers are eaten by in s e c t s , can compensate for t h i s loss by the production of a d d i t i o n a l flowers (Janzen, 1971). Spotted knapweed also seems to have the c a p a b i l i t y of reacting to head s u p e r p a r a s i t i z a t i o n during the f i r s t generation by increasing the rate of head i n i t i a t i o n in-August, even when i t normally commences vegetative rosette growth. This indicates that spotted knapweed al l o c a t e s a net amount of energy to be spent on sexual reproduction, and continues head development u n t i l t h i s energy i s spent. For both d i f f u s e and spotted knapweed, i t i s d i f f i c u l t to show the resurgence i n head production i n August, when terminal heads have been superparasitized, because the t o t a l number of heads i n i t i a t e d under normal growing conditions i s highly v a r i a b l e . A resurgence i n head production may be shown by a greater than expected number of heads produced, e s p e c i a l l y when a,:high proportion of these heads are produced during the second fly-".generation. Hot, dry weather conditions i n August may decrease head production, causing a crash i n f l y numbers, because the density of diapausing larvae may be severely r e s t r i c t e d by fewer and/or smaller heads. ZwOlfer (1970) found that JJ. a f f i n i s produced about 120 eggs per female, and Varley (1947) found that JJ. j aceana could lay between 52 and 70 eggs during t h e i r l i f e t i m e . However, egg mortality for JJ. jaceana was not higher than 15%. If JJ. a f f i n i s and JJ. quadrif a s c i a t a egg m o r t a l i t i e s i n the f i e l d are s i m i l a r , and each female lays 50 or more eggs, then fecundity i n t h i s study 85 i s underestimated. But, adult f l i e s may have died, or dispersed before o v i -p o s i t i o n , and high adult d e n s i t i e s f o r some Tephritids have also been found to decrease fecundity (Watt, 1960). Numbers of TJ. a f f i n i s eggs l a i d i n superparasitized heads could also have been greater than four, as e s t i -mated. Empty pupal cases may have been found i n heads designated as having been produced during the second generation because there may be a small overlap between f i r s t and second generations. Samples could have been taken one week l a t e r , but then some of these'heads'may contain second-generation . progeny, with the r e s u l t that second-generation s i z e would be underestimated. 5.Conclusions: 1. U. a f f i n i s had a 75% death rate i n the p r e - g a l l stage i n 1977 during the f i r s t generation, and on small-headed plants t h i s percentage was 95% or higher. 2. Enough U. a f f i n i s and U. qua d r i f a s c i a t a survive i n large-headed plants and small-headed plants during the f i r s t generation to produce the second generation even at peak f l y d e n s i t i e s . 3.Small-headed plants 1 undergo a resurgence i n head production during the second generation i f most of the heads i n i t i a t e d during the f i r s t gener-ati o n are superparasitized. 4. TJ. q u a d r i f a s c i a t a appears to have a higher fecundity during the second generation when U. a f f i n i s numbers are low and there are p l e n t i f u l heads on small-headed plants. 5. Even though the g a l l f l i e s destroy many, potential seeds during the f i r s t generation, d i f f u s e knapweed can produce high numbers of seeds l a t e i n the growing season. 86 v. E f f e c t of g a l l f l i e s on undeveloped heads: 1.Introduction: Undeveloped heads i n weedy species such as Hieracium floribundum (Thomas and Dale, 1975) are common, and i n Europe (Marsden-Jones and T u r r i l l , 1954) and i n B r i t i s h Columbia, undeveloped heads on knapweed' are common l a t e i n the growing season.- 4J.- »aff-inis and TJ. qua d r i f a s c i a t a act l i k e sinks, draining energy into heads with g a l l s (Shorthouse, pers. comm.). This study determines what e f f e c t g a l l f l i e s have on the percentage of undeveloped heads produced by d i f f u s e and spotted knapweed. 2.Method: The average number of -heads- -in^'seed, superparasitized and undeveloped stages produced by the end of the f i r s t and second adult generations from 1975 to 1978 were obtained from«samples "collected on the d i f f u s e and spotted knapweed release s i t e f o r a previous':study on g a l l - f l y abundance (section II B i i ) . In 1976 and 1977 surveys were made of rangeland populations f o r the percen-tage heads, where any aborted head was counted as "undeveloped". D i f f e r e n t sampling methods were used i n 1976 and 1977, and except f or release s i t e s and the d i f f u s e knapweed population i n Vernon, d i f f e r e n t populations were sampled each year. In' 1976 t h i r t y branches of d i f f u s e and t h i r t y s t a l k s of spotted knapweed were a r b i t r a r i l y taken. In 1977 a 50 X 50 cm square was randomly thrown on the populations, and a l l knapweed within f i v e such squares was r e -moved. Heads f or 1976 and 1977 samples were c l a s s i f i e d as above. Forty seed heads from each sample were chosen, at random, and opened to determine whether g a l l f l i e s had dispersed to the populations. The seeds were used f o r a l a t e r study. In ana l y s i s , the percentages of undeveloped heads for range-land populations were compared with r e l e a s e - s i t e populations using Duncan's 87 New M u l t i p l e Range Test (LeClerge, et a l . , 1962). On release s i t e s the e f f e c t on undeveloped heads by JJ. q u a d r i f a s c i a t a was assumed to be one-half that of JJ. a f f i n i s , because JJ. quadrif a s c i a t a larvae are smaller than JJ_. a f f i n i s larvae (section II B i i a ) and because JJ. qu a d r i f a s c i a t a require l e s s energy for l a r v a l and g a l l development than do JJ. a f f i n i s (Shorthouse, pers. comm.). 3.Results: There was a s i g n i f i c a n t negative c o r r e l a t i o n (r = -.976) between the average number of f i r s t - i n s t a r JJ. a f f i n i s + ^ JJ. quadrif a s c i a t a larvae per head (henceforth termed "the average number of g a l l f l i e s per head") and the percentage of heads i n the seed stage from 1975 to 1978 on the d i f f u s e knap-weed release s i t e (Fig. 2 1 i ) . There were s i g n i f i c a n t p o s i t i v e c o r r e l a t i o n s between the percentage of undeveloped heads (r = .975) and the percentage of superparasitized heads (r = .958) with the average number of g a l l f l i e s per head on d i f f u s e knapweed, but these c o r r e l a t i o n s on spotted knapweed (Fig. 21 i i ) were not s i g n i f i c a n t (r = .934 and .905, r e s p e c t i v e l y ) . There was, however, a s i g n i f i c a n t negative c o r r e l a t i o n between the percentage of heads i n the seed stage and the average number of g a l l f l i e s per head on the spotted knapweed release s i t e (r = -.954). Greater percentages of heads were 2 superparasitized on d i f f u s e than on spotted knapweed ( % 1 3 - 6 7 ) , but 2 there was no s i g n i f i c a n t d i f f e r e n c e f o r undeveloped heads ( ^ ( 2 ) = z-55). The percentage of undeveloped heads among rangeland knapweed populations varied greatly. Probably the g a l l f l i e s , which were found at low den s i t i e s on some s i t e s , had not yet affected the percentage of undeveloped heads to any great extent. Thus, with few exceptions, rangeland populations of %90 80-70-60-50-40-30-20-10-1975 . JO- - • - • - 0 • - — i r ~ 1976 1977 ( i ) - . - • - 0 1978 '3.0 2.5 2.0 1.5 1.0 .5 ft < u co Q < < LL LU HH X CL Q CL < LU ZD 0_ C LU . < ZD > CL LA + 00 < 1—H 1— 00 H-1 -ZL u_ HH 1 < V-00 • C£ d z IX %904 80-70-60-50-40-30-20. 10 X.' _ • -o-JO — r X "0 1975 1976 1977 ( i i ) 1978 -8.0 "7.0 ft < HH o 00 < LL HH CL CD < Q ft X C£ LU CL LU < > P IS A " + ft oo z : •6.0 gj •5.0 •4.0 3.0 2.0 1.0 IX i i— oo CL Figure 21. Relationship between seed heads, undeveloped heads, superparasitized heads, and average numbers of f i r s t - i n s t a r IJ. a f f i n i s and U. quadrif asciata. per head a f t e r the end of the f i r s t adult f l y generation. i . d i f f u s e knapweed release s i t e ; i i . spotted knapweed release s i t e % seed heads + + ; % undeveloped heads X X; % superparasitized heads C— •~0; average numbers of TJ. a f f i n i s + % U. quadrifasciata per head at the end of the f i r s t adult f l y gen-erat i o n N = 77, 402, 300 and 546 for 1975, 1976, 1977 and 1978, r e s p e c t i v e l y on d i f f u s e knapweed, and N =163, 176, 441 and 336 for 1975, 1976, 1977 and 1978, r e s p e c t i v e l y , on spotted knapweed for the percentage of heads i n each stage. . Average f l y densities are from Tables VI, VII, VIII and IX. co co 89 Table XIV. 1976 survey f o r undeveloped heads i n d i f f u s e knapweed populations l o c a t i o n % undeveloped heads Boston Bar Lytton Keremeos Penticton Vernon 1 Falkland Sweetsbridge'' release, s i t e k average Means of a l l s i t e s are s i g n i f i c a n t l y d i f f e r e n t (p <.01) with the mean of the release s i t e , according to Duncan's New M u l t i p l e Range Test (LeClerge et a l . , 1962) U. qu a d r i f a s c i a t a found at l e s s than .5 g a l l s per seed head This s i t e was probably sprayed with Tordon. "average f o r s i t e s which had not been sprayed, and where there were no f l i e s (N = 6) 27 ± 2% 17 - 2 4 - 1 1 0 - 2 1 2 - 2 6 - 2 20 J 4 5 0 - 5 1 4 - 3 Table XV. 1977 survey for undeveloped heads i n d i f f u s e knapweed populations f Vernon 1 2 - 2 f P r i t c h a r d (grazed) 1 4 - 2 f P r i t c h a r d (ungrazed) 11 v 1 locat i o n 1 - % undevei Boston Bar 11 Lytton 21 12 14  Yellow Lake 20 Falkland 9 Lumby (grazed) 11 Lumby (ungrazed) 12 Summerland 23 White Lake Observatory 25 Oli v e r 15 Osoyoos 34 Penticton 16 r e l e a s e . s i t e average'' 43 19 + + + + + + + + + + + + + means of a l l s i t e s are s i g n i f i c a n t l y d i f f e r e n t (p < .01) with the mean of the release s i t e , according to Duncan's New Mu l t i p l e Range Test (LeClerge, et a l . , 1962) """locations preceded by " f " had g a l l f l i e s i n samples ''average for s i t e s without f l i e s and grazing (N = 9) 90 Table XVI. 1976 survey for undeveloped heads i n spotted knapweed populations l o c a t i o n % undeveloped heads 1 Barriere 35 J 2% Sicamous 27 - 3 ** Barnhartvale 30 7 2 ** HWY 97 (Campbell ranch) 31 ^  2 ** Monte Lake 24 r 2 ** Westwold-1 34 j 2 Westwold-2 31 - 3 ** Lake Superior (Ontario) 33 - 2 * r e l e a s e . s i t e 3 5 - 4 average"' 3 0 - 1 X s i g n i f i c a n t d i f f e r e n c e with release s i t e i f followed by * (p < .05) or ** (p < .01) according to Duncan's New M u l t i p l e Range Test (LeClerge et a l . , 1962) -'average f o r a l l s i t e s , excluding Lake Superior arid-th^ release s i t e (N = 7) Table XVII. 1977 survey for undeveloped heads i n spotted knapweed populations — l o c a t i o n % undeveloped heads Westwold-3 4 ^ 1 Westwold-4 61 7 4 Walachin 18 I 4 Keremeos 2 1 - 3 r e l e a s e . s i t e 5 2 - 2 average 24 - 12 _ means of a l l s i t e s are s i g n i f i c a n t l y different (p < .01) with the mean of the release s i t e , according to Duncan's New Mu l t i p l e Range Test (LeClerge et^ a l . , 1962) A v e r a g e for a l l s i t e s excluding the release s i t e (N = 4) d i f f u s e (Tables XIV and XV) and spotted (Tables XVI and XVII) knapweed had lower percentages of undeveloped heads than did release s i t e populations. The average percentage of undeveloped heads f or rangeland populations with-out f l i e s , grazing or spraying were 14 - 3% and 30 - 1% for d i f f u s e and spotted knapweed, re s p e c t i v e l y i n 1976, and 19 - 3% and 24 - 12% for d i f f u s e and spotted knapweed, r e s p e c t i v e l y , i n 1977. The grazed s i t e s at Pr i t c h a r d and Lumby (Table XV) had s i m i l a r percentages of undeveloped heads to un-91 grazed s i t e s at these l o c a t i o n s , i n d i c a t i n g that grazing may not have a great e f f e c t on the percentage of heads that remain*undeveloped. I t i s also i n t e r s t i n g to note that the spotted knapweed population at Lake Superior, Ontario (Table XVI) had a s i m i l a r percentage of undeveloped heads as the average for rangeland spotted knapweed populations. • 4.Discussion: Proximal heads are probably developed only when there are s u f f i c i e n t energy reserves and moisture l a t e i n the growing season to f i l l ovaries. "Seeds are i n i t i a t e d before they are supplied with reserves, and a p e r f e c t l y orga-nized plant would presumably i n i t i a t e only the number of seeds which i t could adequately service. In p r a c t i c e , because environmental conditions vary within and between seasons, the number of seeds produced may exceed or f a l l short of that number" (Harper, L o v e l l and Moore, 1970). Presumably, knapweed must have enough reserves to f i l l a l l the ovarioles i n the heads, or the head remains undeveloped. G a l l f l i e s may be robbing proximal heads of energy reserves because t h i s study indicates that g a l l - f l y d e n s i t i e s are c orrelated with the percentage of undeveloped heads on d i f f u s e and spotted knapweed._ This c o r r e l a t i o n was not s i g n i f i c a n t on spotted knapweed, but the c o r r e l a t i o n c o e f f i c i e n t was high (r = .934). Because undeveloped and superparasitized heads were distinguished mostly on the basis of t h e i r r e s p e c t i v e l y proximal and d i s t a l p o s i t i o n s on branches, s l i g h t errors i n judgement could have resulted i n the above non-significant t e s t . I t i s very probable, however,that the g a l l f l i e s are increasing both the percentage of undeveloped and superparasitized heads on spotted knapweed, as they are on d i f f u s e knapweed. Knapweed i n i t i a t e s 75% or more of the t o t a l number of heads produced before the f i r s t head flowers (section II A i ) , and the plants 92 cannot "forsee" that parasitism of terminal heads w i l l greatly increase energy requirments to these heads. I t also appears «t«hat even i f d i s t a l heads are superparasitized, as many were i n 1977, there i s a depletion of plant energy above the usual amount required f o r proximal head development because the percentage of undeveloped heads was high i n 1977. Head su p e r p a r a s i t i z a t i o n i s normally.related to the average number of female JJ. a f f i n i s per numbers of heads at the r i g h t stage of o v i p o s i t i o n . But, an increase i n the expected s u p e r p a r a s i t i z a t i o n rate may'occur i f the f l i e s emerge e a r l i e r than knapweed heads appear; i f weather conditions are cool and wet, promoting rapid head development, :but slowing f l y a c t i v i t y ; or i f average head s i z e i s smaller than usual due to dry weather conditions. In t h i s study i t was assumed that head s i z e was constant from 1975 to 1978, that f l y emergence and head appearance was well-synchronized each year, and that f l y populations were increasing while the average number of heads i n i t i a t e d each year was the same. But a l l of these, and other v a r i a b l e s may change from year to year, a f f e c t i n g the percentages of heads i n each stage. In some years the density of IJ. q u a d r i f a s c i a t a i s so low, that t h i s species-adds very l i t t l e to the s i g n i f i c a n c e of the . c o r r e l a t i o n - c o e f f i c i e n t s between g a l l - f l y d e n s i t i e s and the percentage of heads i n each stage. Halving the density of JJ. q u a d r i f a s c i a t a i n t h i s study and assuming that t h i s species had half the e f f e c t that JJ. a f f i n i s had i n seed reduction i s supported by Harris' (1980b) find i n g s , where JJ. q u a d r i f a s c i a t a were shown to have ha l f the c a l o r i e s as U. a f f i n i s larvae. 93 The average percentage of undeveloped heads for d i f f u s e and spotted knapweed rangeland populations varied between 1976 and 1977 probably because d i f f e r e n t populations were sampled i n the two years, and because weather conditions were d i f f e r e n t i n 1976 and 1977, as discussed i n section II B i . On d i f f u s e knapweed the average percentage of undeveloped heads, when means of the two years are averaged, i s about 17%. The sample siz e i n 1977 for spotted knap-weed was too small to give a r e l i a b l e estimate of the percentage of undeve-loped heads. I t i s probably close to the 30% found i n 1976. In 1977 the g a l l f l i e s increased the expected percentage of undeveloped heads from the average of 19 - 3% found on rangeland populations to 43 - 2% on the release s i t e . On spotted knapweed, i f the expected average of undeveloped heads i s taken to be 30%, as found i n the rangeland survey i n 1976, g a l l f l i e s i n -creased the percentage of undeveloped heads by 22%. Thus the f l i e s reduce seed numbers not only i n heads which they attack, but they reduce the number of heads which develop to the seed stage, and prevent seed production a l t o -gether i n these heads. Thirty-year averages of the mean annual temperature and t o t a l annual p r e c i -p i t a t i o n do not appear to be r e l a t e d to the percentages of undeveloped heads for rangeland knapweed populations. For example, the percentage of undeve-loped heads i n Osoyoos (34 - 3%) was greater than i n Boston Bar (11 - 2%) for d i f f u s e knapweed even though Boston Bar i s cooler and moister (Table XVIII). O l i v e r and Osoyoos are both hot and dry, but the percentage of undeveloped heads on d i f f u s e knapweed i n Osoyoos was approximately twice that of O l i v e r . Many factors probably a f f e c t the percentage of undeveloped heads:,: such as weather, s o i l type, other vegetation and i n t r i n s i c and genetic factors i n the population. Tordon appears to increase the percentage of undeveloped heads 94 Table XVIII. Climate 0* and a l t i t u d e f o r surveys of d i f f u s e and spotted knapweed populations i n 1976 and 1977 lo c a t i o n * mean annual temperature (°C) * .total annual p r e c i p i t a t i o n (mm) A l t i t u d e (m) Barnharvale Kamloops Kamloops 375 Barriere Bostock ranch 6.7 445 375 - • • • Kamloops Kamloops 366 Boston Bar He l l ' s Gate H e l l ' s Gate 228 Campbgll ranch Westwold Westwold 580 Chase Sicamous Sorrento 671 Falkland 7.2 424 600 H e l l ' s Gate 9.5 1176 114 Kamloops 8.3 261 345 Keremeos 9.5 249 549 Lumby Vernon r Vernon 550 Lytton H e l l ' s Gate Kamloops 357 Monte Lake Westwold Westwold 670 O l i v e r 8.9 299 305 Osoyoos 10.0 342 342 Penticton 8.9 296 610 Pri t c h a r d Kamloops Kamloops 433 Sicamous 7.8 604 426 Sorrento Sicamous 515 390 Summerland Vernon Vernon 853 Sweetsbridge Vernon Vernon 580 Vernon 7.2 393 555 Walachin Kamloops Kamloops 305 Westwold 6.1 316 640 White Lake Observatory Penticton Penticton 762 Yellow Lake Penticton Penticton 853 * for locations where no data i s a v a i l a b l e , see l o c a t i o n c i t e d , which i s the nearest s i m i l a r weather s t a t i o n ** A l t i t u d e given i s where the actual sample was taken and not where the weather s t a t i o n was located d i f f u s e knapweed release s i t e spotted knapweed release s i t e ^Climate data from Climate of B.C. , published by the B r i t i s h Columbia Mi n i s t r y of Ag r i c u l t u r e (standard 1941 - 1970 averages) 95 even when the plant i s not k i l l e d . The g a l l f l i e s are only one of the numerous factors a f f e c t i n g the percentage of undeveloped heads. On the average rangeland spotted knapweed populations- had higher percentages of undeveloped heads than d i f f u s e knapweed populations. Spotted knapweed heads are la r g e r , and produce more numerous, and larger seeds than d i f f u s e knapweed. Thus each spotted knapweed head requires a larger commitment on the part of the plant to develop' the head to maturity than does d i f f u s e knapweed. If energy reserves are low, and the weather i s dry, d i f f u s e knapweed may be able to develop some of the heads, while spotted knapweed may have to abort a l l heads already i n i t i a t e d . 5.Conclusions: 1. A's well as increasing the percentage of superparasitized heads, g a l l f l i e s can increase the percentage of undeveloped heads i n d i f f u s e and spotted knapweed populations. 2. The percentage of undeveloped heads varies greatly among rangeland knapweed populations, but on the average i t i s about 17% for d i f f u s e and 30% for spotted knapweed. 3. Climate does not appear to regulate the percentage of undeveloped heads on rangeland knapweed populations. 96 v i . E f f e c t of g a l l f l i e s on d i f f u s e and spotted knapweed seed weight, appearance and v i a b i l i t y : 1.Introduction: Phytophagous insects may a l t e r seed s i z e , composition or dormancy (Cavers, 1971). Even unattacked seeds i n f r u i t s which have been p a r t i a l l y eaten by predators may have a lower v i a b i l i t y (Janzen, 1971). Diffuse and spotted knapweed heads may contain both g a l l s and mature seeds, but seed v i a b i l i t y may be reduced. This study describes some e f f e c t s of g a l l - f l y attack on d i f f u s e and spotted knapweed seed a t t r i b u t e s . 2.Method: In August, 1976, undehisced seed heads were a r b i t r a r i l y c o l l e c t e d on the d i f f u s e and spotted knapweed release s i t e s , and on co n t r o l s i t e s where there were no g a l l f l i e s (Vernon and Westwold for d i f f u s e and spotted knapweed cont r o l s i t e s , r e s p e c t i v e l y ) . The heads were stored at room temperature for four months before they were weighed. Heads were chosen at random, and the mature seeds i n each head were counted, weighed to the nearest 0.01 mg, and transferred to moistened Whatman #2 germination paper i n P e t r i dishes that were then placed i n a closed cupboard at room temperature (23°C) , found to be the optimal temperature for germination (Watson, 1972). The seeds were checked every three days for r a d i c l e growth. After twelve days the seeds were c l a s s i f i e d as: germinated ( r a d i c l e greater than 2 mm i n length), non-viable ( r a d i c l e l e s s than 2 mm or i f the seed had become diseased), or dormant ( i f the seed did not germinate). The percentage i n each cl a s s of the t o t a l number tested was used i n the analyses. 3.Results: 97 A s i g n i f i c a n t l y greater percentage of seeds from the d i f f u s e knapweed release s i t e (46%) was nonviable compared with seeds from the con t r o l s i t e (16%). S i m i l a r i l y a greater percentage of seeds from the spotted knapweed release s i t e (12%) was nonviable compared with the control s i t e (4%) . Seeds from both d i f f u s e and spotted knapweed control s i t e s were s i g n i f i c a n t l y heavier than seeds from release s i t e s , but the reduction i n weight f or d i f f u s e knapweed was only 7% compared to 33% for spotted knapweed (Table XIX). Table XIX. Effect, of g a l l f l i e s on.diffuse and spotted knapweed seed s' v i a b i l i t y and weight i n 1976 * knapweed . % non-viable mean seed l o c a t i o n , i . , >. species seeds weight (mg) d i f f u s e release c o n t r o l s i t e s i t e 46% 16 0.93 1.01 I+I + 0. 0. 02 04 spotted release c o n t r o l s i t e s i t e 12 4 1.52 2.27 + I+I 0. 0. 03 06 s i g n i f i c a n t d i f f e r e n c e between release and con t r o l s i t e s Nonviable seeds were usually grey, or p a r t l y white i n colour, and had d u l l , porous seed coats, whereas germinated and dormant seeds were usually brown or black, and had hard, waxy seed coats. Nonviable seeds from both d i f f u s e and spotted knapweed release s i t e s were s i g n i f i c a n t l y l i g h t e r than germinated or dormant seeds (Table XX). Germinated and dormant seeds did not d i f f e r i n weight f or ei t h e r d i f f u s e or spotted knapweed. 2 There were s i g n i f i c a n t l y ( % ^ = 21.49) more nonviable seeds from the d i f f u s e knapweed release s i t e (46%) than from the spotted knapweed release s i t e (11%) i 2 but there was no s i g n i f i c a n t d i f f e r e n c e ( % ^ = 1.69) for germinated seeds (51% and 65% for d i f f u s e and spotted knapweed, r e s p e c t i v e l y ) . A s i g n i -98 Table XX. Comparison of germinated, non-viable and dormant seeds from d i f f u s e and spotted knapweed release s i t e s class mean seed weight (mg) % i n c l a s s d i f f u s e 1 spotted 1 knapweed knapweed d i f f u s e spotted j knapweed knapweed germinated non-viable dormant 1.06 - .02 k 1.49 - .03 k .77 - .03 j .80 - .06 j .94 - .07 k 1.54 - .04 k 51 65 46 11 * 3 24 * s i g n i f i c a n t d i f f e r e n c e between classes i f followed by d i f f e r e n t l e t t e r s s i g n i f i c a n t d i f f e r e n c e between d i f f u s e and spotted knapweed denoted by * f i c a n t l y greater percentage (X ^0.05 of seeds fromfhe spotted knapweed release s i t e (24%) were dormant compared to the d i f f u s e knapweed release s i t e (3%). Most d i f f u s e knapweed seeds from the release s i t e which germinated or were dormant, were well-rounded, or f i l l e d . Some spotted knapweed seeds which germinated appeared to be much smaller and s h r i v e l l e d , compared with seeds from the c o n t r o l s i t e (Fig. 22). The s h r i v e l l e d seeds were often sickle-shaped and were usually found i n heads containing g a l l s . C. maculosa 3 4 * 1 4 144 a l t I f f ! 9 V c I I I I • I • I a) U. a ff inis galls bjseeds from un — parasifi zed heads c)seeds from para-sifi zed heads Figure 22. Appearance of spotted knap-weed seeds from unattacked and attacked heads by IT. a f f i n i s , and fusiform TJ. a f f i n i s g a l l s (magnification 1.75X) 99 4.Discussion: The greatest e f f e c t of the g a l l f l i e s on d i f f u s e knapweed seed i s one of a decrease i n the percentage of v i a b l e seed, although there was some de-crease i n average seed weight. Spotted knapweed seed weight was more d r a s t i c a l l y affected than v i a b i l i t y . This dif f e r e n c e could occur as a r e s u l t of the smaller d i f f u s e than spotted knapweed ovaries; any reduction i n nutrients to d i f f u s e knapweed heads may cause the seeds to abort, .but spotted knapweed seeds would be able to withstand a considerable reduction i n nutrients before the seeds lose t h e i r v i a b i l i t y . The s h r i v e l l e d appearance of spotted knapweed seeds i n heads with JJ. a f f i n i s g a l l s may be, i n part due to the pressure by the woody g a l l s on the immature seeds, squeezing some into s i c k l e shapes. The g a l l f l y , JJ. j aceana, which i s s i m i l a r to JJ. a f f i n i s (as described e a r l i e r i n section II B i i ) appears to have a s i m i l a r e f f e c t on C.. nemoralis (Wadsworth, 1914) as the g a l l f l i e s i n t h i s study have on d i f f u s e and spotted knapweed (when Wadsworth made h i s study, C_. nemoralis was c a l l e d C_. nigra and JJ. j aceana was c a l l e d JJ. s o l s t i t i a l i s ; Varley, 1947). Wadsworth (1914) found that "probably the nourishment that normally goes to the developing seeds became diverted to the growing larvae, and also to the formation of the thick-walled, woody g a l l " . He described JJ. j aceana g a l l s as having a layer of s o f t , n u t r i t i v e t i s s u e i n s i d e the woody outer e x t e r i o r , and he found that the n u t r i t i v e layer was "connected at the base of the chamber with the vascular system of the plant". Only 28.5% of the seeds from ga l l e d C_. nemoralis heads germinated, whereas 89% of the seeds from ungalled heads germinated. The o v e r a l l e f f e c t of JJ. j aceana on C_. nemoralis was to r e -duce v i a b l e seed y i e l d by 50% of the t o t a l number of v i a b l e seeds produced. This e f f e c t i s s i m i l a r to that of JJ. a f f i n i s and JJ. quadrif a s c i a t a on d i f f u s e knapweed, where 46% of the seeds were not v i a b l e . 100 Storey (1976), working with IJ. a f f i n i s i n Montana, did not f i n d a s i g n i -f i c a n t decrease i n spotted knapweed seed v i a b i l i t y , and the germination percentage was 96.7% from heads with 4 or more g a l l s . In the present study i f "dormant" and "germinated" seeds are considered to be v i a b l e , the percen-tage of seeds which could p o t e n t i a l l y germinate i s 89% on the spotted knap-' weed release s i t e . This r e s u l t i s only s l i g h t l y lower than found by Storey. Storage of weed seeds often increases the germination percentage, and Watson (1972) found that spotted knapweed seeds, stored for 20 months at room temp-erature, had a germination percentage of 93%, whereas those seeds stored for only 3 and 25 days had germination percentages of 20% and 80%, r e s p e c t i v e l y . Seeds i n t h i s study were stored for four months at room temperature before they were moistened, and the germination percentage obtained for spotted knapweed on the release s i t e i n 1976 was not unlike percentages found by Watson (1972) for seeds stored f o r 25 days. Therefore the differ e n c e i n germination percentages between seeds from the release and control s i t e s may be mainly due to s i t e d i f f e r e n c e s . Even i f g a l l f l i e s do lower the germin-ati o n rate of spotted knapweed seed, the net e f f e c t seems small (see also Storey,- 1976) . Si t e differences probably did not e n t i r e l y account f or the great decrease i n d i f f u s e knapweed seed v i a b i l i t y on the release s i t e because the a t t r i b u t e s of the seed coat appeared to be changed. Seeds from the release s i t e were l i g h t brown or grey i n color, and had d u l l and porous coats, whereas seeds from the con t r o l s i t e were dark brown, and had hard, waxy seed coats. Maun and Cavers (1971a) found s i m i l a r r e s u l t s by d e f o l i a t i n g Rumex crispus. 101 Seeds from nondefoliated plants were "plumper, bright reddish-brown i n c o l o r , and very shiny. Seeds from d e f o l i a t e d plants were d u l l , smaller, and l i g h t brown i n co l o r " . They found also that "the average thickness of the i n -vesting structures of seeds was reduced from 62.2 u i n seeds from c o n t r o l plants to 52.6 u i n seeds from d e f o l i a t e d plants" and that the seeds from the d e f o l i a t e d plants took up water more r a p i d l y and germinated f a s t e r than seeds from co n t r o l plants. Thus one e f f e c t of the g a l l f l i e s on d i f f u s e knapweed may be to decrease the seed bank i n the s o i l , because seeds from t h i s i . s i t e appeared to have thinner seed coats, and would therefore germinate fa s t e r than seeds from co n t r o l s i t e s . Small seeds tend to germinate more r e a d i l y than large seeds (Harper, 1963),-therefore the g a l l f l i e s may also be decreasing the seed bank on the spotted knapweed release s i t e because spotted knapweed seed weight was<reduced by 33%. Some of t h i s d i f f e r e n c e i n weight between the co n t r o l and release s i t e may be due to s i t e d i f f e r e n c e s , and Watson (1972) found an average of . 1.78 mg for spotted knapweed seeds, which i s lower than the average of 2.27 mg found i n t h i s study for the co n t r o l s i t e . Average weight for spotted knapweed r e l e a s e - s i t e seeds was lower than Watson's (1972) weight, i n d i c a t i n g that the g a l l f l i e s are reducing average seed weight. Probably seeds i n heads with g a l l s are more severely a f f e c t e d , while seeds i n unattacked heads may s u f f e r only s l i g h t weight l o s s . However, average seed weight would be greatly reduced at peak f l y "population, idensities, such as i n 1977, when very few heads were found without g a l l s . Seedlings produced by small seeds s u f f e r a greater m o r t a l i t y than seedlings produced by larger seeds, when seeds of both sizes are found i n the same 102 population. However, i f o v e r a l l seed s i z e i s small, then the same population growth can be achieved from the smaller sizes as for populations started from large seeds. (Harper, L o v e l l and Moore, 1970). On both release s i t e s seed s i z e varied f o u r f o l d for the class of seeds which germinated i n 12 days. Thus there was a mixture of small and large seeds, and a long-term e f f e c t of the g a l l f l i e s could be an increase i n seedling m o r t a l i t y by increasing the proportion of small seeds produced on the release s i t e s . The c o l l e c t i v e e f f e c t s of the g a l l f l i e s on seed a t t r i b u t e s w i l l not r e s u l t i n a decrease i n knapweed population density unless the surviving seedlings cannot replace a l l those i n d i v i d u a l s which died. This question demands a lengthy study of tte decay-rate of the seeds i n the s o i l , and knapweed population dynamics. 5.Conclusions: l . G a l l f l i e s appear to reduce both v i a b i l i t y and weight of d i f f u s e and spotted knapweed seeds, but d i f f u s e knapweed seed has a comparatively greater de-• crease i n v i a b i l i t y , whereas spotted knapweed seed s u f f e r s a greater weight l o s s . v i i . Survey of rangeland d i f f u s e and spotted knapweed seed y i e l d i n 1977: 1.Introduction: Young g a l l - f l y larvae feed d i r e c t l y on knapweed ovaries (Shorthouse, pers. comm.) before c e l l s p r o l i f e r a t e and form n u t r i t i v e t i s s u e . E a r l i e r i t was shown that head s i z e within d i f f u s e knapweed populations a f f e c t s g a l l - f l y s u r v i v a l at high population d e n s i t i e s . Therefore c h a r a c t e r i s t i c f l y den-s i t i e s per head for rangeland knapweed populations may depend on the average 103 numbers of seeds produced per seed head. And, although most authors agree that seed s i z e i s one of the le a s t p l a s t i c characters of a plant (Harper, 1963; Palmblad, 1968), there could be v a r i a t i o n i n knapweed seed s i z e among populations. This -study surveys d i f f u s e and spotted knapweed seed weight and y i e l d : on rangeland. 2. Method: Seed heads from fourteen d i f f u s e and four spotted knapweed populations, and from the d i f f u s e and spotted knapweed release s i t e s , were obtained from knapweed c o l l e c t e d i n 1977 for the survey i n head abortion (section II B v ) . A l l the seed heads from each s i t e were counted, and the average number of seed heads per plo t (50 X 50 cm) were found. Forty undehisced seed heads were chosen, at random, and opened. The number of v i a b l e seeds i n each head was counted, where v i a b l e seeds were distinguished from non-viable seeds and u n f e r t i l i z e d ovaries as noted i n the previous section (II B v i ) . I f any s t e r i l e seed heads were found, they were classed as seedless. I f any TJ. a f f i n i s or TJ. q u a d r i f a s c i a t a g a l l s were found, t h e i r density per seed head was recorded. Seed heads with g a l l s were included i n the estimation of the mean number of seeds per seed head. A l l v i a b l e seeds from the f o r t y heads were pooled and spread over a g r i d . Sixty seeds, closest to predetermined points on the g r i d , were weighed to the nearest 0-01 mg. 3. Results: D i f f u s e knapweed seeds from the release s i t e ranked second-heaviest out of the d i f f u s e knapweed populations surveyed, with the Yellow Lake s i t e having s i g n i f i c a n t l y heavier seeds (Table XXI). The Yellow Lake s i t e had been heavily grazed during the middle and l a t t e r parts of the growing season. 104 Seeds from a l l other d i f f u s e knapweed s i t e s were s i g n i f i c a n t l y l i g h t e r than those from the d i f f u s e knapweed release s i t e . Seeds from the Lumby-2 s i t e , which was moderately grazed, were s i g n i f i c a n t l y heavier than those from the ungrazed Lumby-1 s i t e , but the di f f e r e n c e between grazed and ungrazed Pritchard-2 and Pritchard-1 populations, r e s p e c t i v e l y , was not s i g n i f i c a n t . The Pritchard-2 s i t e had been grazed only i n the early spring, before r o-settes had started to b o l t , while the Lumby-2 s i t e had been moderately grazed throughout the growing season. For a l l but two d i f f u s e knapweed populations, the mean number of seeds per head was s i g n i f i c a n t l y greater than on the release s i t e . But at Osoyoos, where there were no g a l l f l i e s , and at Summerland, where U. qua d r i f a s c i a t a was found at a low density (0.3 - 0.1 g a l l s per seed head), there were only 3.3 seeds per head. This average was almost the exact density found on the d i f f u s e knapweed release s i t e (3.2 seeds per head). I f net seed weight i s defined as the weight of the t o t a l number of seeds per seed head, and i s found by multip l y i n g the mean seed weight by the mean number of seeds per head, then f or the d i f f u s e knapweed populations surveyed, there was a s i x -f o l d gradation i n net seed weight. At Osoyoos the. seeds were l i g h t (0.90 mg) and few per head (3.3), so net seed weight was only 2.97 mg per head. But, at Yellow Lake, seeds were both heavy (1.37 mg) and many per head (12.9), so net seed weight was 17.67 mg per head. S i m i l a r i l y , there was a large var-i a t i o n i n net seed, weight f or spotted knapweed. The p l o t s at Lumby-1 and Boston Bar had s i g n i f i c a n t l y greater numbers of seeds per plo t than the d i f f u s e knapweed release s i t e , but there were s i g -n i f i c a n t l y fewer seed heads per plo t on four s i t e s (Oliver, Osoyoos, Lumby-2, Table XXI. Survey of d i f f u s e knapweed seed weight and y i e l d i n 1977 sxte seed weight" (mg) number per seed head seeds" U. a f f i n i s g a l l s U. qua d r i f a s c i a t a g a l l s number of seed-' heads per plo t (50 X '50 'cm),. number of seeds-per pl o t (50 X 50 cm) Vernon Summerland M Penticton Oliver' " Osoyoos . Falkland Lytton Pritchard-1 Pritchard-2 H Lumby-1 Lumby-2 M White Lake Observatory H Yellow Lake H Boston Bar release s i t e 1.02.. 1.07 1.02 1.12 .90 1.10 .97 .99 -'.99 ,04 ,10 + + + + + + + + + + + 37 .93 1.21 .02 ** ,03 ** ,03 ** ,03 ** ,03 ** ,03 ** ,03 ** ,03 ** .03 ** ,02 ** - .02 ** + 05 - .03 ** + + + ,04 ** ,02 ** ,04 5.5 5.0 9.2 9.1 8. 12.9 8.0 3.2 + + + + + + + + + + ±1 + + + 1 ** .5 .8 ** .5 ** .7 .7 ** .8 ** .5 ** , 9 * * .7 ** . 9 * * •k* , 9 * * ,7 ** ,6 0 0 0 0 0 0 $ .15 - .07 0 0 0 0 0 1.2 - .2 .2 .1 o o .6 - .1 •9 .20 - .08 .08 - .06 0 0 1.0 - .2 .08 308 J 149 J 243 J 130 J 105 -279 T 258 r 237 T 198 r 728 -67 -9± 262 t 380 -235 -61 82 51 80 74 94 27 15 48 70 58 47 135 ** 20 ** 2768 -492 | 2209 r 622 J 345 -2289 -1417 T 1184 T 1820 r 6628 -724 -66 ± 2345 t 3038 -751 -714 ** 244 852 ** 130 51 390 ** 383 291 433 ** 1228 ** 108 13 732 ** 652 ** 163 s i t e was grazed moderately (M) or heavily (H) ^ s i g n i f i c a n t difference with release s i t e i f followed by * (p <.05) or ** (p <.01) according to Duncan's New Mul t i p l e Range Test (LeClerge et a l . , 1962) Table XXII. Survey of spotted knapweed seed weight and y i e l d i n 1977 site' seed weight'' (mg) number per seed head number of seed>J heads per plo t (50 X 50 cm) number of seeds^ per pl o t (50 X 50 cm) seeds'' U. a f f i n i s U. qua d r i f a s c i a t a g a l l s g a l l s Keremeos Walachin Westwold-1 Westwold-2 release s i t e 1.82 J .03 ** 2.00 - .05 ** 1.62 T .03 1.52 - .04 ** 1.68 - .04 19 t. 2 ** 6 + 1 1 1 - 2 ** 1 3 - 2 ** 7.0 - .9 .02 - .02 0 0 4.7 - .5 <? 1.0 - .3 4-.4 - .2 54 J 14 26 j 8 ** 28 - 7 : *•*'• 108 T 18 ** 72 - 15 1034 ^ 273 ** 155 T 49 * 312 - 75 1407 - 230 ** 507 - 102 s i g n i f i c a n t difference with release s i t e i f followed by * (p <.05) or ** (p <.01) according to Duncan's New Mul t i p l e Range Test (LeClerge et^ a l . , 1962) o ON 107 and White Lake Observatory). Seven out of the fourteen d i f f u s e knapweed populations surveyed produced more seeds per plo t than the r e l e a s e - s i t e population, but the seed y i e l d on the remaining s i t e s did not d i f f e r s i g -n i f i c a n t l y with the release s i t e . JJ. q u a d r ifasciata was found i n Vernon, Summerland and Falkland, the clos e s t of these s i t e s being 60 km from the d i f f u s e knapweed release s i t e . The f l i e s on the Pritchard-1 and Lumby s i t e s came from e a r l i e r releases made for other studies. Seeds on the Westwold-2 s i t e , where JJ. q u a d r i f a s c i a t a g a l l s were found at a density of 0.1 - 0.1 per seed head, were s i g n i f i c a n t l y l i g h t e r than seeds from the spotted knapweed release s i t e (Table XXII). The mean seed weight on the Westwold-1 s i t e did not d i f f e r s i g n i f i c a n t l y from the release s i t e , and seed weight of the remaining two populations i n the survey, at Keremeos and Walachin, had s i g n i f i c a n t l y heavier seeds than the populations on the release s i t e . The number of seeds per seed head at Walachin did not vary s i g n i f i c a n t l y from the number of seeds per head on the release s i t e , but on a l l other s i t e s , there were more seeds per seed head than on the release s i t e . S i m i l a r i l y , the number of seed heads per p l o t , and the number of seeds per plo t exceeded averages found f or the spotted knapweed release s i t e on only one and two s i t e s , r e s p e c t i v e l y . Because o n l y ' w e l l - f i l l e d seeds were weighed i t was assumed that the average seed weight was not affected by the occasional g a l l found i n the seed head samples. The average weight of d i f f u s e knapweed rangeland seed was 1.05 -.03 mg, and there were an average of 7.4 - .7 seeds per seed head, 240 - 46 108 seed heads per pl o t and 1853 - 447 seeds per p l o t . For spotted knapweed the average seed weight was 1.7 - .1 mg, and there were an average of 12 - 3 seeds per seed head, 54 - 19 seed heads per pl o t and 727 - 297 seeds per p l o t . 4.Discussion: There was considerable v a r i a t i o n i n the average number of seeds per seed head and i n average seed weight between knapweed populations. Thus average f l y d e n s i t i e s per head, which are aff e c t e d by net seed weight per head, w i l l probably vary from one population to another when the g a l l f l i e s have dispersed throughout the I n t e r i o r . Knapweed populations where net seed weight per head i s low w i l l support lower average numbers of g a l l s per head than knapweed with many, heavy seeds per. head, because smaller heads, are more e a s i l y superparasitized than larger ones . Other phenomena, such as the percentage of undeveloped and superparasitized heads, and the r a t i o of TJ. a f f i n i s to TJ. quadrifasciata w i l l probably be affected by the changes i n net seed weight per head. It i s i n t e r e s t i n g to note that, although the g a l l f l i e s were at peak den-s i t i e s on release s i t e s i n 1977, other rangeland populations of knapweed with none or only a few f l i e s produced fewer or the same number of seeds per seed head, and per p l o t . This indicates that r e l e a s e - s i t e knapweed populations are very f e r t i l e and are probably producing seed f a r i n excess of that needed for the maintenance of the e x i s t i n g populations. G a l l f l i e s may a f f e c t knapweed populations only a harsh s i t e s , where surplus of seed i s smaller. 109 V a r i a t i o n i n seed weight and y i e l d are affected by many f a c t o r s , pooled under "the environment" by Goodall (1970). Harper (1960) found that s o i l f e r t i l i t y could a f f e c t seed y i e l d : i n les s f e r t i l e s o i l Papaver rhoeas produced an average of four seeds per capsule and one capsule per plant, but i n more f e r t i l e s o i l 2,000 seeds were produced per capsule and there were 400 capsules per plant. S i m i l a r i l y , d i f f u s e and spotted knapweed, which grew i n i r r i g a t e d pasture produced more seeds per seed head, and more seed heads per plant than knapweed growing on rangeland (Watson, 1972). Seed weights may be affected by seasonal weather changes (Harper, L o v e l l and Moore, 1970), and by d e f o l i a t i o n (Maun and Cavers, 1971a). Where plants grow i n mixed populations, stress due to i n t e r s p e c i f i c competition at the time when the seeds are " f i l l e d " can a f f e c t seed weight and number (Harper, L o v e l l and Moore, 1970). The time of harvest can a f f e c t average seed weight. For example, i f Rumex  crispus i s c o l l e c t e d early, average seed weight w i l l be higher than the r e a l averages for the population because seeds, which mature f i r s t are heavier than late-maturing seeds (Maun and Cavers, 1971b). Diffuse knapweed seed i n t h i s study appears to be affected by grazing. At Yellow Lake grazing occurred during mid-summer when- d i f f u s e knapweed was flowering. A reduction i n the number of heads probably resulted i n an i n -creased supply of nutrients to seeds'which escaped destruction and were s t i l l being f i l l e d , because averagerseed weight appeared to be very heavy. On the Pritchard-2 s i t e grazing occurred only i n the spring, so that only the ro- .. settes were damaged, and plants, that bolted, produced and f i l l e d seeds l i k e the plants growing on the ungrazed Pritchard-1 s i t e . , 110 Because so many factors may influence seed weights and y i e l d s , d i f f e r e n t estimates f o r the same species can be found i n the l i t e r a t u r e . For example, Stevens (1932) reported seed weights for Cirsium arvense from d i f f e r e n t sources ranging from 0.900 to 1.575 mg. In t h i s study, average seed weights of d i f f u s e (1.05 mg) and spotted (1.74) knapweed were s i m i l a r to Watson's (1972) estimates of 1.10 mg and 1.78 mg f o r d i f f u s e and spotted knapweed. This indicates that the spotted knapweed con t r o l s i t e i n the previous section (II B v i ) produced exceptionally heavy seeds (2.27 - .06 mg) and that unless the spotted^knapweed'release-site seeds ' were also heavier than average, g a l l - f l i e s may be reducing seed''weigHfi.'less- than estimated. Fewer seeds per d i f f u s e knapweed seed head (7.36) were found i n t h i s study than by Watson (1972; 12.5). Watson (1972) also found a larger value for the average number of spotted knapweed seeds per seed head (26.6) than i n th i s study (12.25), but the sample s i z e f o r spotted knapweed here,was , too low (only four s i t e s ) to confirm that Watson's estimate i s d e f i n i t e l y higher. U. a f f i n i s was not found at any new s i t e s where i t had not been previously released, but U. quadrifasciata was found at several new l o c a t i o n s . This indicates that IJ. quadrifasciata disperses much fa s t e r than U_. a f f i n i s , and may be able to a t t a i n high population sizes before U. a f f i n i s catches up. 5.Conclusions: 1. Net seed weight per seed head v a r i e s considerably among rangeland knapweed populations because there were s i g n i f i c a n t differences i n average seed weight and average numbers of seeds per seed head among these populations. 2. "Average seed d e n s i t i e s per head and per unit area f o r some rangeland popu-l a t i o n s were- l e s s than or equal to those on the d i f f u s e and spotted knap---- . I l l weed populations on the release s i t e s i n 1977, at peak g a l l - f l y d e n s i t i e s . 3.JJ. quadrif a s c i a t a disperses f a s t e r than U_. a f f i n i s . v i i i . V a r i a t i o n i n seed weight and numbers i n a spotted knapweed population: 1.Introduction In the previous section i t was shown that seed weight and average numbers of seeds per seed head varied among rangeland populations of d i f f u s e and spotted knapweed. Here comparisons are made of average seed weights and numbers of seeds per seed head among i n d i v i d u a l plants i n a sing l e spotted knapweed population, and between i n d i v i d u a l seed heads on sing l e spotted knapweed s t a l k s . 2.Method: . This study was made on a spotted knapweed rangeland population i n Westwold, i n 1976. Single s t a l k s , a l l of which came from separate plants, were tagged while the heads were s t i l l i n the bud stage, and s t a l k heights were measured to the nearest cm. A mature seed head was defined as one which had turned yellow and dried, i n d i c a t i n g that the vascular ti s s u e had been blocked, and the seeds were re c e i v i n g no a d d i t i o n a l n u t r i e n t s . When the heads were har-vested they were coded according to t h e i r p o s i t i o n on branches as i n section II A i , and stored i n separate envelopes. Heads on s i x t y stalks were har-vested, but afterwards i t was found that U. qu a d r i f a s c i a t a had attacked some heads on f i f t e e n of the s t a l k s . These sta l k s were discarded, leaving only forty-five r e p l i c a t e s . The seed heads were opened and v i a b l e seeds were counted. Viable seeds were defined as previously (section II B v i ) , as those seeds with hard, shiny, brown or black coats. Viable seeds from each seed head were weighed together and divided by the number of seeds per head 112 for an estimate of average seed weight per head. 3.Results: There was a s i g n i f i c a n t d i f f e r e n c e i n average seed weight and average num-bers of seeds per head. Usually the seed head on the f i r s t primary branch, which also flowered f i r s t , had the highest seed count and the heaviest seeds f o r the s t a l k . The r e s u l t s varied from 9 to 41 seeds per terminal head (a 2^-fold range) and from 0.93 to 2.57 mg for seed weight (a 3k-fold range). The average seed weight and numbers of seeds per seed head for heads on primary branches usually decreased from the top to the bottom of the s t a l k , as did the order of flowering. For one t y p i c a l s t a l k , the average numbers of seeds per head were 20, 20, 18, and 17 for I, I I , III and IV head p o s i -t i o n s , and average seed weights were 1.87, 1.76, 1.61 and 1.38 mg, respec-t i v e l y . However there were exceptions. On another s t a l k the number of seeds per seed head decreased with increasing order of flowering (24, 19, and 11 for heads i n I, II and III p o s i t i o n s ) , but seed weight was v a r i a b l e (1.52, 1.47 and 1.77 mg for heads i n I, II and I I I p o s i t i o n s , r e s p e c t i v e l y ) . If heads i n terminal positions on the f i r s t four primary branches are con-sidered to be be i n d i s t a l p o s i t i o n s , whereas any head on a secondary branch i s i n a proximal p o s i t i o n , then d i s t a l heads appeared to have greater numbers of seeds per seed head (26.3 - 6) than those heads i n proximal positions (19 - 1). Although t h i s d i f f e r e n c e was not s i g n i f i c a n t , i t might have been i f sample s i z e f o r proximal heads which reached maturity (N = 14) had been la r g e r . The reason for the small proximal-head sample s i z e was that many heads i n t h i s class aborted. And, although there were exceptions, average seed weight was higher i n d i s t a l than i n proximal heads. The end r e s u l t 4oH 1 30' X X X , X X x X 10 H x x x xx X X 2 0 H x x x X x x x x, I— 70 10 —r— 20 —i— 30 40 50 STALK HEIGHT (CM) 60 80 Figure 23. Correlation between the number of seeds per terminal seed head and sta l k height f spotted knapweed. (r = .39; si g n i f i c a n t at 43 D.F.) 114 was that net seed weight was greater i n d i s t a l than proximal heads, and t h i s was often v i s u a l l y apparent by the smaller seed-head siz e for proximal than for d i s t a l heads. There was a s i g n i f i c a n t p o s i t i v e c o r r e l a t i o n between the mean number of seeds per seed head with s t a l k height when only f i r s t - f l o w e r i n g seed heads i n p o s i t i o n I were considered. But, the c o r r e l a t i o n c o e f f i c i e n t was low (0.39) i n d i c a t i n g only a weak r e l a t i o n s h i p (Figure 23). There was no s i g -n i f i c a n t c o r r e l a t i o n between s t a l k height and average seed weight i n terminal heads (r = 0.26; N = 45). However, a s i g n i f i c a n t c o r r e l a t i o n c o e f f i c i e n t (r = .51, N = 45) was obtained between s t a l k height and net seed weight per head (numbers of seeds per head X average weight per seed). 4.Discussion: The food resource for g a l l f l i e s , i n the form of ov a r i o l e t i s s u e , v a r i e s among i n d i v i d u a l s t a l k s i n a spotted knapweed population, as well as with the order of flowering and head p o s i t i o n on i n d i v i d u a l s t a l k s . Because a l l experimental st a l k s were from d i f f e r e n t plants, t h i s v a r i a t i o n probably _ -occurs among plants. The number of g a l l s and mature seeds which can develop i n spotted knapweed heads probably depends to some extent on the net seed weight, as discussed previously. Thus great care must be taken when stalks are caged with f l i e s f o r the purpose of determining the e f f e c t of f l y attack on seed reduction. Storey (1976) had to c a l c u l a t e separate regression equa-tions for each of three, separately-caged stalks with JJ. a f f i n i s i n order to determine the e f f e c t of increasing g a l l numbers on spotted knapweed seed y i e l d . He found that when there were no g a l l s , there were re s p e c t i v e l y 21.36, 26.10 and 28.07 seeds per seed head for the three r e p l i c a t e s . He 115 also found that there was a great v a r i a b i l i t y i n the t o t a l number of seed heads produced by each r e p l i c a t e . The average number of seeds per head and net seed weight for terminal head positions probably increases with plant height because the t a l l e r s t a l k s appear to be produced by l a r g e r , more vigorous rosettes. Net seed weight decreases with the order of flowering because energy flows into heads which appear f i r s t to ensure s e e d - f i l l i n g for at l e a s t some heads i n case adverse weather conditions set i n . Heads i n proximal positions usually have lower seed weights probably because l i t t l e energy i s l e f t f o r sexual reproduction by the time t h e i r turn comes for development. Often proximal heads remain undeveloped. Larger heads produced by knapweed e a r l i e r i n the growing sea-son give f i r s t - g e n e r a t i o n g a l l f l i e s an advantage over second-generation f l i e s . Ovariole number, as w e l l as ovariole s i z e , i s probably c o n t r o l l e d by genetic factors as well as by environmental conditions, hence the exceptions to the above patterns of seed weight and number. Harper, L o v e l l and Moore (1970) found that plants with a determinate type of growth pattern have a greater v a r i a t i o n i n seed s i z e than those plants with indeterminate growth forms. Also species which flower more synchronously have a greater v a r i a t i o n i n seed weight than species which flower continuously. Both d i f f u s e and spotted knapweed are determinate, but v a r i a t i o n i n seed weight among heads i n a d i f f u s e knapweed plant would probably be le s s than i n a spotted knapweed plant because spotted knapweed has a short, concentrated period of flowering compared to d i f f u s e knapweed. Harper, L o v e l l and Moore (1970) c i t e d numerous examples of v a r i a t i o n i n seed s i z e within plants. They found that i n some cases the heavier seeds are shed l a t e r , germinate l a t e r and 116 produce larger cotyledons than the l i g h t e r seeds. B i o l o g i c a l control agents can a l t e r many seed properties (Cavers, 1971), but i n order to determine those affected, normal d i s t r i b u t i o n patterns of these a t t r i b u t e s f o r i n d i -v i d u a l plant types must be found, or else sample sizes:must be very large. 5.Conclusions: l.Seed weight and number of seeds per seed head vary among i n d i v i d u a l spotted knapweed plants i n a population, and with head p o s i t i o n and order of flowering on i n d i v i d u a l s t a l k s . 2. T a l l e r spotted knapweed sta l k s usually produce more seeds per head and heavier seeds i n terminal heads than do shorter s t a l k s . 3. D i s t a l heads appear to produce heavier and more numerous seeds per seed head than do proximal heads. i x . E f f e c t of g a l l f l i e s on knapweed seed number on the release s i t e s from 1975 to 1977: 1.Introduction: This i s a study of the e f f e c t s of U_. a f f i n i s and TJ. q u a d r i f a s c i a t a on the seed y i e l d s of d i f f u s e and spotted knapweed populations on the release s i t e s from 1975 to 1977. 2.Method: In 1975 and 1977 a 50 X 50 cm square was thrown at random on the d i f f u s e and spotted knapweed release s i t e s and a l l the heads within the square were har-vested at the end of the growing season. In 1975 there were four r e p l i c a t e s and i n 1977 there were five.-' Numbers of heads i n seed, superparasitized and undeveloped stages were counted. Samples of at l e a s t f i f t y undehisced seed 117 heads, chosen at random from the c o l l e c t e d heads, were opened, and the v i a b l e seeds (defined as previously) were counted, as were JJ. a f f i n i s and JJ. quadri- f a s c i a t a . S t e r i l e heads, i . e . , those heads which flowered, but produced no seed and were unattacked, were counted as "seed heads" with n i l seeds. In 1976 numbers of seed heads were estimated by multipl y i n g the mean number of stalks per 50 X 50 cm plo t by the mean number of seed heads per s t a l k . Counts of seeds and g a l l s per seed head were obtained from plant material c o l l e c t e d for the study of g a l l - f l y d e n s i t i e s on release s i t e s (section I I B i i ) . Two assumptions were made i n estimating the percentage of seed loss which could be a t t r i b u t e d to g a l l f l i e s : the average number of t o t a l heads per plo t was not affected by g a l l - f l y attack, and the expected percentage of undeveloped heads was the average f or knapweed growing on rangeland i n 1977, where there were no f l i e s , found to be 14 - 3% and 30 - 1% for d i f f u s e and spotted knap-weed, re s p e c t i v e l y (section I I B i v ) . The expected number of seed heads per plot was found by subtracting the expected number of undeveloped heads from the t o t a l number of seed heads + superparasitized heads + undeveloped heads. The expected number of seeds per seed head was found by a r b i t r a r i l y c o l l e c t i n g undehisced seed heads on nearby co n t r o l s i t e s i n 1977 and 1975, only, f o r d i f f u s e (Pritchard) and spotted (Chase) knapweed, r e s p e c t i v e l y . Preliminary work had indicated that average numbers of seeds per seed head were s i m i l a r i n unattacked heads. G a l l f l i e s on con t r o l s i t e s were at low de n s i t i e s (0.07 - .02 and 0.21 - .03 JJ. a f f i n i s g a l l s per seed head on d i f f u s e and spotted knapweed, r e s p e c t i v e l y , and 0.14 - .03 and 0.05- .01 JJ_. quadrif a s c i a t a g a l l s per seed head on d i f f u s e and spotted knapweed, r e s p e c t i v e l y ) . Any seed heads containing g a l l s were not used i n the-estimates of expected numbers of seeds per seed head, and these estimates were used i n analysis from 1975 to 1977. It was assumed that the few heads that were gal l e d did not a f f e c t seed counts i n unattacked heads. The actual and ex-118 pected seed y i e l d s per pl o t were calculated by multip l y i n g the actual and expected number of seed heads per p l o t by the actual and expected number of seeds per seed head, r e s p e c t i v e l y . The "percentage reduction i n seed y i e l d " i s defined here as the (expected number of seeds per p l o t — the actual number of seeds p l o t ) / (the expected number of seeds per plo t ) X 100%. U. qua d r i f a s c i a t a was detected i n Vernon i n 1976, i n a population without JJ. a f f i n i s . Branch samples were a r b i t r a r i l y c o l l e c t e d from the same d i f f u s e knapweed population i n 1977 and 1979, and the seed heads were opened, and g a l l s and v i a b l e seeds were counted. 3.Results: On the average-10V7-. 0.3 and 25.3 T 0; 4-seeds-per seed Ahead we're found on the d i f f u s e and spotted" knapweed con t r o l s i t e s , r e s p e c t i v e l y . Assuming that i n 1971, before the f l i e s were released on d i f f u s e knapweed, and when f l y numbers were s t i l l low at Chase, the numbers of seeds per head on the release s i t e s were the same as on the con t r o l s i t e s , there was a s i g n i f i c a n t decrease i n the mean number of seeds per 'seed head from 1971 to 1977 on both release s i t e s (Fig. 24). The slope of the regression l i n e was s i g n i -f i c a n t l y l e s s f o r d i f f u s e (-1.28) than for spotted (-2.80) knapweed, i n d i -cating that the rate of decrease i n seed numbers per seed head was greater on spotted than on d i f f u s e knapweed. Previously (section II B i i ) i t was shown that d e n s i t i e s of U. a f f i n i s f i r s t - i n s t a r larvae increased s i g n i f i -cantly from 1975 to 1977 on both d i f f u s e and spotted knapweed release s i t e s . From 1975 to 1977 there was a s i g n i f i c a n t negative c o r r e l a t i o n between the mean number of U. a f f i n i s f i r s t - i n s t a r larvae per head and the mean number of seeds per seed head for spotted knapweed ' (Fig. 26), but t h i s r e l a t i o n s h i p 00 c ro AVERAGE NUMBER OF SEEDS PER SEED HEAD 611 H-TO 3 K (D K3 ON CD Ci n fD 3 o (D CL H CL fu i-i CO rt Hi fD 3* Hi 1—1 fD H- CU fD 3 rt H g H- H-fD CO O CO p 3 fD 3 Hi fD H- cr1 Cu 3 1-1 fD 3 CD rt 3* S rt s! fD o- 1 fD CO fD H* fD CL l-t 3 3 • CD O rt rt s s Hi to 3* H H fD CD II X) M 3 1 O Co fD • rt CO VO rt < 3 VO fD CO CL fD 3 ^ • 3 i V d 3 CD 3 fD cr H* fu H fD TO 13 3 s: 3* H- fD fD o Hi fD fo Hi H- CL CL O CO 3 rt Co rt H-1 a —^ o -i ho o -« I VO •3 CL . PL XI H- fD Hi t-j Hi c & CD fD fD CO _ CL vo IV N> 3 CO <••"» •' X) 3 C O fD rt rt fD p-CL fD CD H- CD 3 TO fD fD 3 fD Co H- CL 3 Hi CD H-o i3 CO fD 3 H TO 3 H fD N3 Ul o pa Hi fD h-1 ;<=i co • rt Co O Hi 3 Hi CO H- 3* 3 H-H- T) CO cr1 Hi fD H- rt H C CO fD X NUMBER OF U. AFFINIS FIRST-INSTAR LARVAE PER HEAD co Co fD rt fD 3 i fD H o CL Hi • CO HJ CL fD 3 rt 3* fD 3 fD CO 3 3 C §• fD H XI rn 73 IX) m m o to •u m to m m o zz N3 OJ -4 X to . tO x ^° X NUMBER OF U. AFFINIS FIRST-INSTAR LARVAE PER HEAD _JL_ to i -P-I U l X v O X t O cn X 031 121 was not s i g n i f i c a n t for d i f f u s e knapweed (Fig. 25). However the c o r r e l a t i o n c o e f f i c i e n t on d i f f u s e knapweed was high (-.992), i n d i c a t i n g thatTJ. a f f i n i s probably does decrease average numbers of seeds per seed head._ on t h i s s i t e as w e l l . There was no s i g n i f i c a n t c o r r e l a t i o n between the mean number of TJ. q u a d r i f a s c i a t a f i r s t - i n s t a r larvae per head and the mean number of seeds per seed head on either s i t e . In 1976 for d i f f u s e , and by 1977 for spotted knapweed, seed y i e l d had de-creased to 80% of the expected y i e l d , but i n 1977 about 800 to 900 d i f f u s e , and 500 to 600 spotted knapweed seeds were s t i l l being produced per p l o t (Table XXIII). From 1975 to 1977 the number of t o t a l heads and seed heads produced per plo t on the spotted knapweed release s i t e declined, but on the d i f f u s e knapweed release s i t e head numbers were lowest i n 1976. For a l l years the mean number of seeds per seed head decreased s i g n i f i c a n t l y with increasing d e n s i t i e s of U. a f f i n i s g a l l s per seed head (Figs. 27 to 32) on the d i f f u s e and spotted knapweed release s i t e s . Numbers of seeds per seed head varied considerably on both s i t e s . For example, i n 1975, when g a l l ' counts were s t i l l low, there were as few as-one and as many as twenty-four seeds i n unattacked d i f f u s e knapweed heads, and there was was a range of seven to thirty-one seeds i n unattacked spotted knapweed heads. S i m i l a r i l y , there was great v a r i a t i o n in.the r a t i o of seeds produced to numbers of g a l l s w ithin one seed head. In 1977 some d i f f u s e knapweed seed heads contained only one TJ. a f f i n i s g a l l , but produced no seed, but one seed head was found with s i x JJ. a f f i n i s g a l l s and eleven seeds. On spotted knapweed i n 1977, a seed head was found with one g a l l and no seeds, but another was found containing fourteen g a l l s and two seeds. Table XXIII. Expected arid actual seed y i e l d s per plot on d i f f u s e and spotted knapweed release s i t e s , and the estimated reduction i n y i e l d from 1975 to 1977 t o t a l number A expected A * expected •. actual actual-number V act u a l % reduction knapweed year of heads number of number of number of of. seeds / ;per number of i n seed-, counted 1 seed heads seeds seed heads seed head seeds- y i e l d •cu,. CD - 1975 651 - 175 560 - 270 5832 - 2970 532 - 149 5.37 - .03 2857 - 828 51% ffu 1976 299 - 29 257 - 128 2754 - 1455 140 - 36 4.1 ± .5 574 - 219 81 •H 13 1977 537 - 126 462 - 45 4941 - 2322 230 - 47 3.2 +- .3 852 - 287 83 13 1975 306 - 22 256 - 21 6482 - 633 251 - 28 17.8 - .3 4468 - 13 31 tte 1976 275 - 43 192 - 32 4869 - 881 180 - 26 11.8 - .9 2124 - 459 56 ods 1977 170 - 42 119 - 31 3011 - 813 68 - 14 7.8 ± .9 530 - 32 82 pl o t s i z e wass50 X 50 cm A + assumption: the % aborted heads of the t o t a l heads produced = 1 4 - 3 % for d i f f u s e knapweed (Table XIV) and 30 - 1% for spotted knapweed (Table XVI) A A _ + + assumption: ;the x number of seeds per seed head = 10.7 - .3 for d i f f u s e and 25.3 - .4 for spotted knapweed (see text f o r source of data) NUMBER OF SEEDS PER SEED HEAD I—' I—' I—' cn oo o ho -F" >< x x t e * 7 ^ ^ g# x xx xx * * 00 1 O _1_ to to to 4^  _J J X X X-a * * a * X X NUMBER OF SEEDS PER SEED HEAD 00 c l-( fD to 00 h-1 cr W fD Hi fu Hi ON Cu fD • O o rt 3 —s O CO rt Hi H- Cr1 00 fD a 3 H- Cu Hi H- fo H- Hi Hi n Hi Hi fa C H-3 co 3 rt fD H-cn i-i 7T fD 3 O oo fo 3 i-i T) fD s! cn CO fD fD CO fD fD H- Cu Cu o 3 i-i 3 fD C fu H 3 rt fD cr fu fD u> CO ON fD cn a cn T3 • H" fD rt H • fD cn H- fD 3 fD Cu £31 to cn _ L _ (-> to CO o _] L_ to to _L_ 18 o 16 -k I U J LU * 12 + LU CL g 10 + UJ LU CO u_ o LU CD *x "*x tx X X , 6 X 3 H 5 NUMBER OF U. AFFINIS GALLS PER SEED HEAD 10 Figure 29. E f f e c t of U. a f f i n i s on seed numbers per seed.head on the d i f f u s e knapweed release s i t e i n 1977 ( s i g n i f i c a n t regression at 151 D.F.) Symbols for number of data points: X 10 points; X 5 points; x 1 point. h-1 ho 4>-125 35n Q UJ UJ CO OL UJ D_ U~l Q UJ UJ to U_ O C£ UJ CD 30H 25' x x G X X X E X c 2?X X X X CX X X X X X X X 1?*xx Pxxxxxxx r ~ x x k x xxx xx 1 x kxxx X J 3x X X X X X 2 0"F Kxx x x x xx 15-i 10' X X X ( XX bcX Xx X x xxxxx: OC.XX X X X XX C X cxxxxxx X c xxx X x xxxx x X XXXx x X X X X x X X K X x X X X X X 4 6 8 10 12 14 16 1 NUMBER OF U. AFFINIS GALLS PER SEED HEAD — r -16 20 — r -22 i 24 Figure 30. E f f e c t of IJ. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1975 ( s i g n i f i c a n t regression at 252 D.F.) Symbols for number of data points: X 5 points; x 1 point 126 Q LU LU CO CL LU O. CO Q LU LU CO CL LU CD 35H 3CH 25 20 15 H i o H 5 H x X X X X X X X X * X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 4 6 8 10 12 14 16 18 20 NUMBER OF U. AFFINIS GALLS PER SEED HEAD 22 24 Figure 31. E f f e c t of JJ. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1976 ( s i g n i f i c a n t regression at 113 D.F.) Symbols for number of data points: X 5 points; x 1 point 127 Q UJ UJ If) OL UJ CL I/J Q UJ UJ I/O U_ O ry UJ CO 35' 30i 25-20H 15' 10H xx X X X X X X X x ^ v y y y v y VTrffy -y YY y ^ 4 6 8 10 12 14 16 18 20 NUMBER OF U. AFFINIS GALLS PER SEED HEAD ~22~ —i 24 Figure 32. E f f e c t of JJ. a f f i n i s on seed numbers per seed head on the spotted knapweed release s i t e i n 1977 ( s i g n i f i c a n t regression at 65 D.F.) 128 The percentage of seed heads attacked per sample decreased from 1975 to 1977 on the d i f f u s e knapweed release s i t e (41%, 18% and 11% for 1975, 1976 and 1977, respectively) and decreased from 1975 to 1976 on the spotted knapweed release s i t e (22% and 2%, respectively) and remained low on spotted knapweed i n 1977 (4%) . The maximum number of U_. a f f i n i s g a l l s per seed head increased on both release s i t e s from 1975 to 1977: from four to nine on d i f f u s e knapweed, and from nine to eighteen on spotted knapweed. Y - i n t e r -cepts on d i f f u s e and spotted knapweed regressions of numbers of seeds per seed head on numbers of TJ. a f f i n i s g a l l s per seed head were greater i n 1975 (7.2 - .3 and 19.7 - .7 for d i f f u s e and spotted knapweed, respectively) than i n 1977 (4.5 - .4 and 9.2 - .2 for d i f f u s e and spotted knapweed, respect t i v e l y ) , i n d i c a t i n g that numbers of seeds i n unattacked heads declined from 1975 to 1977 on both s i t e s . When only those heads containing U. qua d r i f a s c i a t a and unattacked seed heads are considered, there was a s i g n i f i c a n t decrease i n numbers of seeds per seed head with increasing numbers of TJ. qua d r i f a s c i a t a g a l l s per seed head i n 1977 on the d i f f u s e knapweed release s i t e (Fig. 33). The slope of the regression l i n e was s i g n i f i c a n t l y l e s s (-1.37) than f or U. a f f i n i s on d i f f u s e knapweed i n the same year (-0.87), but the Y-intercept was higher for U_. quad r i f a s c i a t a (7.67) than f o r U. a f f i n i s (4.54). Average numbers of U. qua d r i f a s c i a t a g a l l s per seed head i n 1977 on the Vernon s i t e (0.5 - .2) did not vary s i g n i f i c a n t l y with the average count i n 1979 (0.46 - .06), but the mean number of seeds per seed head i n unat-tacked heads was s i g n i f i c a n t l y l e s s i n 1979 (7.5 - .4) than i n 1977 (12 -1). There appeared to be a decrease i n numbers of seeds per seed head with increasing d e n s i t i e s of U. q uadr i f as c i a t a i n Vernon i n 1977 (Fig. 34), but 15H Kx Q UJ UJ 00 ry UJ a. oo a UJ UJ 00 UJ CO 10 + Figure 33, 5 + x NUMBER OF U. QUADRIFASCIATA GALLS PER SEED HEAD E f f e c t of U. quadrifasciata on numbers of seeds per seed head on the d i f f u s e knapweed release s i t e i n 1977 ( s i g n i f i c a n t regression at 32 D.F.) 22T c 20-5C lex 184 fx 16-f 14 12.^x 8 - & h~ ex 2-< x X X X X Figure 34. NUMBER ..OF U. QUADR IFASCI ATA GALLS PER SEED HEAD E f f e c t of U. q u a d r i f a s c i a t a on d i f f u s e knapweed seed numbers per seed head i n Vernon,. 1977 (no s i g n i f i c a n t regression) K3 VO 130 Q UJ UJ 00 oc U J 0_ 00 Q UJ UJ 00 DC UJ co 17 4CX 16 154c 14-KX 12 i r - k x x i o i < x 8+ x 1 2 3 NUMBER OF U. OUADRIFASCIATA GAULS PER SEED HEAD Figure 35. E f f e c t of U. quadrifasciata on d i f f u s e knapweed seed numbers per seed head i n Vernon, 1979. ( s i g n i f i c a n t regression at 146 D.F.) Symbols for number of data points: X 10 points; X 5 points; and x 1 point. 131 the sample s i z e was small (40 heads), so the regression was not s i g n i f i c a n t . However, i n 1979, when sample s i z e was increased to 148 heads, the same re l a t i o n s h i p was s i g n i f i c a n t (Fig. 35). V a r i a t i o n i n numbers of seeds per unattacked seed heads was great i n both years: i n 1977 there were as few as two and as many as twenty-two seeds per seed head, and i n 1979 the var-i a t i o n ranged from one to seventeen seeds per seed head. 4.Discussion: The decrease i n average numbers of seeds per seed head from 1975 to 1977 as average numbers of f i r s t - i n s t a r IJ. a f f i n i s larvae per head increased may be a t t r i b u t e d to an e f f e c t on d i f f u s e and spotted knapweed by t h i s g a l l -f l y species. Evidence for an average reduction i n seeds per seed head from 1975 to 1977 i s that f or each year numbers of seeds per seed head decreased with increasing number-s-of JJ. a f f i n i s g a l l s per seed head. U. qu a d r i f a s c i a t a also reduces average seed numbers, but i t s e f f e c t i s minimal compared with TJ. a f f i n i s , because of lower numbers of JJ. qu a d r i f a s c i a t a and because JJ. qu a d r i f a s c i a t a larvae are smaller and form thinner g a l l s , thereby using l e s s plant reserves than do JJ. a f f i n i s larvae. At d e n s i t i e s of f i v e or more g a l l s for e i t h e r species, few seeds escaped destruction, but at lower g a l l d e n s i t i e s more seeds escaped in^'heads with TJ. qu a d r i f a s c i a t a than with JJ. a f f i n i s . This may explain why the absolute slope of the regression l i n e was less with JJ. a f f i n i s than with JJ. q u a d r i f a s c i a t a on d i f f u s e knapweed i n 1977. Harris (1980b) found that on d i f f u s e knapweed JJ. a f f i n i s destroyed an average of 2.4 seeds for every g a l l and JJ. q u a d r i f a s c i a t a reduced y i e l d by an average of 1.9 seeds for every g a l l , and on spotted knapweed JJ. a f f i n i s reduced y i e l d by 1.1 seeds per g a l l . The e a r l i e r assumption that d i f f u s e and spotted knap-132 weed heads are superparasitized at an average of four and eight JJ. a f f i n i s larvae per head, res p e c t i v e l y (section II B i i ) appears v a l i d , because i n 1977, when head s u p e r p a r a s i t i z a t i o n was most severe, few seeds were produced at higher TJ... a f f i n i s 1 "densities than those estimated f o r superparasitized v. heads. This study supports the theory that g a l l s act l i k e sinks, u t i l i z i n g energy reserves from other parts of the plant other than i n heads were the g a l l s are located. As f l y d e n s i t i e s i n heads on both release s i t e s increased, average seed numbers i n unattacked heads decreased. Possibly competition f o r nu-< t r i e n t s also e x i s t s among heads with g a l l s , and heads with IJ. a f f i n i s may be robbing energy from heads with JJ. q u a d r i f a s c i a t a , as w e l l as from unat--tacked heads. This could create error i n estimating the actual e f f e c t JJ. qua d r i f a s c i a t a had on seed y i e l d on the d i f f u s e knapweed release s i t e . How-ever a s i g n i f i c a n t regression was obtained i n Vernon, where JJ. quadrifasciata occurred alone, the e f f e c t appears to be much the same as for JJ_. a f f i n i s ex-cept that i t i s not as severe. As i n previous sections, t h i s study shows the great v a r i a t i o n i n carrying c a p a c i t i e s among knapweed heads for g a l l - f l y attack. Some heads support not only large numbers of g a l l s and also produce v i a b l e seeds, but other heads y i e l d no seed even though there may be only one or two g a l l s i n them. Due to-t h i s v a r i a t i o n sample sizes had be large to show any e f f e c t on seed reduction by the f l i e s . I t i s probable that the regression equation of JJ. quadrifasT.. c i a t a on d i f f u s e knapweed on the Vernon s i t e i n 1977 would have been s i g n i -f i c a n t had more than 40 seed heads been dissected. Environmental conditions may also a f f e c t the average number of seeds per seed head. This i s w e l l - i l l u s t r a t e d by the s i t u a t i o n i n Vernon, where aver-age IJ. qu a d r i f a s c i a t a counts per seed head were the same i n 1977 and 1979, but average numbers of seeds per seed head were lower i n 1979 than i n 1977. Therefore fewer g a l l f l i e s could reduce seed y i e l d to n i l i n years when head sizes are small. However, from 1975 to 1977, when samples were taken on release s i t e s , head sizes on con t r o l s i t e s did not appear to vary. There-for the reason f or the reduction i n seed y i e l d on the release s i t e s from 1975 to 1977 was probably due to f l y attack alone. Seed head numbers could have increased from 1976 to 1977 on the d i f f u s e knap-weed release s i t e because of the apparent resurgence i n head production l a t e i n the growing season of 1977 (section II B i v ) . H arris (1980b) a t t r i b u t e d the decline i n average numbers of heads on the spotted knapweed release s i t e to environmental conditions. Wadsworth (1914) found that U_. s o l s t i t i a l i s ( i . e . , U. jaceana; Varley, 1947) reduced the seed y i e l d of C. nigra ( i . e . , C_. nemoralis; Varley, 1947) by 50% of the expected y i e l d , and f e l t that i f t h i s f l y were encouraged to increase i n density, i t could be used as a natural c o n t r o l agent against t h i s weed. If TJ. a f f i n i s and U. qua d r i f a s c i a t a reduce knapweed seed production by 50% i n countries of o r i g i n , where t h e i r e f f e c t i s l i m i t e d by natural enemies, then these b i o l o g i c a l c o n t r o l agents achieved about a 30% increase i n seed destruction i n B r i t i s h Columbia. But release s i t e populations produced such copious amounts of seed that i n 1977, at peak f l y d e n s i t i e s , an average of 800 d i f f u s e and 500 spotted knapweed seeds per 50 X 50 cm were being shed. Whether or not these seed den s i t i e s w i l l be s u f f i c i e n t to maintain knapweed 134 population d e n s i t i e s on release s i t e s depends on knapweed s u r v i v a l from seed to mature-plant stages, and w i l l be discussed i n a l a t e r section. 5.Conclusions: 1. G a l l f l i e s achieved about an 80% reduction i n p o t e n t i a l seed y i e l d on both d i f f u s e and spotted knapweed release s i t e s at peak population d e n s i t i e s i n 1977. 2. Fewer seeds are produced by unattacked heads as g a l l d e n s i t i e s i n attacked heads increase. 3. U. q u a d r i f a s c i a t a i s not as e f f e c t i v e as U. a f f i n i s i n reducing seed. 135 II I KNAPWEED DEMOGRAPHY A. EFFECTS OF SEED REDUCTION ON KNAPWEED POPULATIONS 1.Introduction: Population s i z e and ground cover of plant populations that produce greater; numbers of seeds than required for the maintenance of e x i s t i n g densities,' may be regulated by one or more of three f a c t o r s : 1.mortality: an i n -creasing proportion of plants die as density increases, and a decreasing proportion die as density decreases; 2 . p l a s t i c i t y : as plant density changes, i n d i v i d u a l plant s i z e , seed y i e l d per plant, seed y i e l d per pod or capsule, or type of reproduction (eg., vegetative as opposed to sexual) may change; and 3.controlled germination: as numbers of seeds i n the s o i l increase fewer germinate. (Harper, 1960; Harper and Gaj i c , 1961; and Palmblad, 1968). Most weedy species studied produce seeds far i n excess of the numbers needed for the maintenance of populations; e.g., sowing seeds into populations of the same species did not increase population s i z e (Putwain et^ a l . , 1968; Harper, 1960). Knapweed populations commonly produce over 1,000 seeds per 50 X 50 cm (section II B v i i ) , and therefore produce more seed than i s needed to replace i n d i v i d u a l s which died. G a l l f l i e s can reduce seed y i e l d by 80% (section II B v i i ) , but i s that s u f f i c i e n t to a f f e c t plant stands? This question was examined by a r t i f i c i a l l y thinning seedlings. 2.Method: This study was begun i n 1976 and continued u n t i l 1977, using the same test p l o t s . Work i n 1976 was l a r g e l y preliminary. Some of the l i f e h i s t o r y d e t a i l s found i n 1977 were used to construct the survivorship and f e r t i l i t y 136 schedules described i n the next section. The d i f f u s e knapweed study s i t e was located on f l a t rangeland, 380 m i n elevation, near P r i t c h a r d , B.C. Dif f u s e knapweed invaded the s i t e c i r c a 1965 (Robertson, pers. comm.). The study area had been enclosed and pro-tected from grazing since 1970. The l o c a t i o n of t h i s s i t e was 3 to 4 km northeast from the d i f f u s e knapweed release s i t e . The spotted knapweed • s i t e was located i n Westwold, B.C., approximately 48 km south from the d i f f u s e knapweed s i t e , at an elevation of 610 m. Knapweed invaded the area c i r c a 1970 (Buff, pers. comm.). The study s i t e was not grazed f o r four years p r i o r to 1976, and was enclosed with a fence to prevent further grazing. There was a very small g a l l - f l y population on both s i t e s because caging studies had been ca r r i e d out nearby, and some f l i e s had escaped. In 1976 t h i r t y 50 X 50 cm square plots were established on the d i f f u s e and spotted knapweed s i t e s by throwing a square, and delineating square p e r i -meters with baler twine. Corners were held i n place with fencing staples. Three strands of baler twine were stretched across each p l o t , subdividing the p l o t s into four 50 X 12.5 cm rectangles, to f a c i l i t a t e seedling counting. The treatments were 0%, 25%, 50%, 75% and 100% removal of the t o t a l number of seedlings which appeared. Six r e p l i c a t e s were chosen at random for each treatment. In 1976, the seedlings were thinned once every two weeks, when seedlings were removed at the rate of each treatment, 0%, 25%, 50%, 75% and 100% of the number present by using forceps. A seedling was defined as a plant having f i v e or fewer true leaves, none of which were dissected. A rosette was defined as-a plant which had not previously bolted, and which had s i x or more true leaves i n a sin g l e "rosette" arrangement. I t was assumed that i f a seedling was present during one sampling time, i t would reach the 137 rosette stage by the next sampling time, two weeks l a t e r . In 1977 seedlings were subdivided into two classes: small and large. Small seedlings were defined as plants which ranged i n maturity from the dicotyledon stage to the three-true-leaf stage, with no leaves longer than 2 cm, and none dissected. Large seedlings were c l a s s i f i e d as plants with four or f i v e true leaves, with at l e a s t one l e a f greater i n length than 2 cm, but none dissec-ted. Small rosettes were c l a s s i f i e d as plants with s i x or more true leaves, of which at le a s t one was longer than 2 cm, and dissected. Small and large seedlings were counted once every two weeks, and at the same time any plant i n the small rosette stage was marked by i n s e r t i n g a l e t t e r e d p l a s t i c p i n d i r e c t l y behind i t into the ground. Then the number of seedlings which had appeared i n the preceding two-week i n t e r v a l , i . e . , "newly-appeared seedlings", was calculated as follows: i f there was an increase i n the number of large seedlings during the previous  two weeks, i t was assumed that i n d i v i d u a l s which had entered the large seedling class i n the previous two weeks had been counted as small seedlings at the l a s t count. Small seedlings i n the previous count which did not enter the large seedling stage e i t h e r died or remained i n the small seedling stage. These l a t t e r i n d i v i d u a l s were c a l l e d "old small seedlings". Newly-appeared small seedlings were claculated by the equation NA = SS^ - (SS^_^ -A LS), where NA = the number of newly-appeared small seedlings, SS^ = the number of small seedlings i n the current week's count, SS. ^ = the number of small seedlings i n the previous week's count, zjk LS = the number of large seedlings i n the current week's count - the number of large seedlings i n the l a s t count, 138 SS. ., - A LS = the number of old small seedlings, l - l i f there was a decrease i n the number of large seedlings during the previous^ two weeks, i t was assumed that no small seedlings, which had been counted two weeks e a r l i e r , grew-enough f o l i a g e to enter the large seedling c l a s s , and the number of newly-appeared small seedlings was calculated from the equation NA = SS. - SS. , . 1 l - l The number of small seedlings to be removed from each pl o t was then c a l c u -lated from the number of newly-appeared small seedlings and the designated removal rate. The small seedlings were removed by p u l l i n g them out with forceps. More small seedlings were removed from thicker than thinner • ... patches of seedlings and, given a choice of two seedlings, the youngest-appearing seedlings was always removed because i t soon became apparent that some small sedlings could survive i n t h i s stage for longer than the two-week sampling i n t e r v a l . Rosettes were defined as plants whichhad not previously bolted, and there-fore consisted of only one "rosette" formation from a s i n g l e root crown. Rosettes had at l e a s t s i x true leaves i n May, at which time a l l the rosettes were tagged by i n s e r t i n g numbered p l a s t i c pins d i r e c t l y behind them. Rosette diameters were measured to the nearest cm, and the number of leaves greater than two cm were counted. When d i f f u s e and spotted knapweed reached maturity (September and August, r e s p e c t i v e l y ) , fates of each numbered rosette were recorded as e i t h e r : "dead" i f a l l the leaves had been l o s t , and no new ones appeared; "regrowing" i f a l l the leaves had been l o s t , but short, dissected leaves with 139 short p e t i o l e s had sprouted afterwards; "vegetative" i f the rosette had not attempted to b o l t ; "arrested" i f the rosette had bolted, but had not produced any heads; "unsuccessful" i f the rosette had bolted but produced' only aborted heads; or " s u c c e s s f u l " i f the rosette had bolted and produced at l e a s t one mature seed head. Diameters of vegetative rosettes were measured to the nearest cm, and the number of leaves longer than two cm were counted. Stalk heights of bolted rosettes were measure to the nearest cm, and numbers of seed and aborted heads per bolted rosette were counted. Perennial d i f f u s e knapweed was defined as any plant which had previously bolted. Perennial plants usually had numerous"rosettes at the root crown, each of which could b o l t . Those perennial plants with si n g l e rosettes had the remains of dead st a l k s from previous years when rosettes, coming from the same root, had bolted. Perennial plants were tagged i n May by i n s e r t i n g numbered p l a s t i c pins d i r e c t l y behind them. In the f a l l , t h e i r fates were recorded as: "dead" i f a l l leaves had been l o s t from a l l rosettes; "regrowing" i f a l l leaves from a l l rosettes had been l o s t , but short, dissected leaves had sprouted from the centre of at l e a s t one of the "dead" rosettes; "vegetative" i f none of the rosettes had bolted; "arrested" i f at l e a s t one rosette had bolted, but no heads were produced; "unsuccessful" i f at l e a s t one rosette had bolted, but a l l heads produced were aborted; or 140 " s u c c e s s f u l " i f at l e a s t one rosette had bolted and produced at l e a s t one mature seed head. Numbers of seed and aborted heads per bolted perennial plant were counted. Fates of perennial spotted knapweed plants were c l a s s i f i e d s i m i l a r l y to per-ennial d i f f u s e knapweed plants. Spotted knapweed was further s u b c l a s s i f i e d on the basis of numbers of rosettes attached to the root crown: a "small, perennial" plant had one or two; a "medium perennial" plant had three or four, and a "large perennial" Riant had f i v e or more rosettes per crown. In May, each perennial spotted knapweed plant was tagged with a numbered p l a s t i c p in, inserted d i r e c t l y behind i t into the ground. In August, fates of spotted knapweed perennial plants were recorded and head production was counted the same as for d i f f u s e knapweed. 3.Results: In 1976 i t appeared that seedlings remained i n the seedling stage longer than two weeks because they developed thick root crowns and dissected leaves when le a f numbers were fewer than f i v e . Therefore more than the designated percentage of t o t a l seedlings which appeared was probably removed i n 1976. If a seedling broke at the crown when an attempt was made to p u l l i t out, i t often regrew a s i n g l e rosette, or numerous i n d i v i d u a l rosettes from the crown. These regrowing rosettes were soon distinguished from other plants i n that they consisted of short, very dissected leaves, with short p e t i o l e s , and were' l a t e r removed.. Both seedlings and rosettes could lose a l l t h e i r leaves, thereby appearing dead, and l a t e r sprout new, healthy leaves from the centre of the crown. 141 In 1977 numbers of small d i f f u s e and spotted knapweed seedlings per pl o t decreased s i g n i f i c a n t l y with removal rate (Table XXIV). The percentage of small d i f f u s e and spotted knapweed seedlings which died decreased with i n -creasing thinning, but the decrease was s i g n i f i c a n t only for spotted knapweed. Similarly," there-.was a decreasing trend i n the number of d i f f u s e and spotted knapweed large seedlings and small rosettes with increasing thinning rate, but the decrease was s i g n i f i c a n t only f o r large d i f f u s e knapweed seedlings and small spotted knapweed rosettes. No plants i n the large seedling or small rosette stages were found on the plots where a l l seedlings were r e -moved because a l l i n d i v i d u a l s had been removed while they were s t i l l i n the small seedling stage. There was no s i g n i f i c a n t d i f f e r e n c e i n the percentage of large d i f f u s e knapweed seedlings which died, or small d i f f u s e and spotted knapweed rosettes which died with increasing thinning rate, but the per-centage of large spotted knapweed seedlings which died varied s i g n i f i c a n t l y with thinning rate. There was no s i g n i f i c a n t d i f f e r e n c e between thinning rates i n the number of d i f f u s e knapweed plants i n rosette or perennial classes, nor i n the number of spotted knapweed plants i n small, medium or large perennial classes i n the spring of 1977 (Table XXV). However, there was a s i g n i f i c a n t d i f f e r e n c e with thinning rate i n the number of spotted knapweed plants i n the rosette stage. But, i f one pl o t with unusually high spotted knapweed rosette counts-(85 compared with the average of 2 8 - 3 for a l l t h i r t y plots) i s omitted, than there i s no s i g n i f i c a n t d i f f e r e n c e with thinning rate. The percentage of d i f f u s e or spotted knapweed perennial plants which entered -.dead, regrowing, vegetative, arrested, unsuccessful or successful fate categories by the f a l l of 1977 did not vary s i g n i f i c a n t l y with thinning rate. Numbers of seed, Table XXIV. Average numbers of d i f f u s e and spotted knapweed plants per p l o t f o r each thinning rate i n 1977 tr e a t -ment small seedlings large seedlings small rosettes d i f f u s e knapweed spotted knapweed di f f u s e knapweed spotted knapweed d i f f u s e knapweed spotted knapweed number % dead number % dead number % dead number % dead number % dead number % dead 0% ' 25 50 75 100 . SIG. 1 337 J 38 147 J 15 124 I 9 85 - 10 5 0 - 2 * 73 6 63 T 8 72 - 5 46 --10 376 J 90 192 - 43 120 - 28 91 - 19 63 - 16 * 84 t 3 8 9 - 3 7 6 - 4 6 0 - 6 * 76 J 10 37 | 8 1 6 - 3 1 2 - 1 * 18 J 6 33 - 10 34 - 7 2 3 - 9 56 t 21 2 1 - 5 1 7 - 7 1 1 - 2 76 J 5 62 | 5 8 3 - 8 6 3 - 7 * 53 t 7 2 9 - 9 1 0 - 1 1 0 - 1 21 J 5 27 - 5 27 J 6 2 9 - 9 13 t 5 6 - 2 2 + 1 4 - 2 * 36 ~t 10 52 - 12 67 T 24 57 - 14 s i g n i f i c a n t difference among thinning rates denoted by * treatments are 0%, 25%, 50%, 75% and 100% removal of small seedlings Table XXV. Average numbers of d i f f u s e and spotted knapweed plants per p l o t f o r each si z e c l a s s i n the spring of 1977, and t h e i r fates by the f a l l of 1977. Average head production for each s i z e c l a s s i n the f a l l of 1977. knap-weed i s i z e class number of plants % plants i n fate category i n f a l l of 1977 heads per bolted plant dead re-growing vege-t a t i v e arrested unsuc-c e s s f u l l ' suc-c e s s f u l l seed aborted t o t a l dif-fuse r P 123 J 8 1.6 - .7 1 8 2 1 5 - 7 2 - 2 66 t 6 3 8 - 7 0 0 1 - 1 1 * 1 2 4 4 - 2 13.2 J .9 1 1 - 3 .8 t .1 .8 - .2 14.3 ^ .8 1 1 - 4 spotted r sp mp IP + * 2 8 - 3 9 + 1 6.7 - .9 2.3 - .3 28 J 4 1 5 - 4 8 - 2 2 - 1 0 0 6 0 46 J 4 1 3 - 3 9 - 2 1 - 1 8 ± 1 3 1 - 5 2 0 - 4 3 2 - 7 1 7 - 4 1 8 - 5 7 - 3 l l | 2 2 4 - 5 4 5 - 6 5 8 - 8 .49 | .05 1.0 - .4 1.7 7 .4 4.7 - .9 . 7 ± . l .7 - .2 1.3 - .2 2.8 - .6 1.2 | . l 1.7 - .5 3.0 - .6 8 - 1 1r=rosette; p=perennial; sp=small perennial; mp=medium perennial; lp=large perennial s i g n i f i c a n t difference between treatments due to one plot with exceptionally high rosette counts; i f t h i s p l o t i s omitted there i s no s i g n i f i c a n t difference between treatments and the x No. of rosettes per p l o t i s 26 - 2 143 undeveloped and t o t a l heads per bolted d i f f u s e or spotted knapweed rosette, or d i f f u s e or spotted knapweed perennial plant, also did not vary s i g n i f i -cantly with thinning rate. Neither d i f f u s e nor spotted knapweed rosette diameters varied s i g n i f i c a n t l y with thinning rate i n May, 1977. Again, i n the f a l l of 1977, there was no s i g n i f i c a n t d i f f e r e n c e with thinning rate among spotted knapweed rosette width classes. But for d i f f u s e knapweed there was a s i g n i f i c a n t difference with thinning rate f o r rosette diameters i n the f a l l of 1977 (Fig. 36). There was, however, no s i g n i f i c a n t d i f f e r e n c e between average spring and average f a l l d i f f u s e knapweed rosette diameters for p l o t s where a l l seedlings were removed. The following i s based on a measure of rosette s i z e , where the s i z e i s rosette diameter (cm) X the number of leaves greater than 2 cm long. The percentage of d i f f u s e and spotted knapweed rosettes which bolted of the number of rosettes i n the spring of 1977, increased s i g n i f i c a n t l y with i n -creasing rosette s i z e (Figs. 37 and 38, r e s p e c t i v e l y ) . The percentage of d i f f u s e and spotted knapweed rosettes which died or remained vegetative during the summer of 1977 decreased s i g n i f i c a n t l y with increasing rosette s i z e . There was no s i g n i f i c a n t d i f f e r e n c e with thinning rate for d i f f u s e knapweed rosettes which regrew. Bolted spotted knapweed rosettes had three f a t e s : arrested, unsuccessful, and successful. I f only spotted knapweed rosettes which bolted are con-sidered, then the percentage which was arrested decreased s i g n i f i c a n t l y with increasing rosette s i z e (Fig. 39). The percentage of rosettes which UJ cn O OL UJ O LU 8 T LU LU Q LU 7H P 6H ,1 0 0 — r — 25 T" 50 0 75 144 0 100 % SEEDLINGS REMOVED OF TOTAL WHICH APPEARED Figure 36. Average d i f f u s e knapweed rosette diameter i n the spring and f a l l of 1977 for each thinning rate spring 0; f a l l X were unsuccessful did not vary s i g n i f i c a n t l y with rosette s i z e . I f only d i f f u s e knapweed rosettes which bolted i n 1977 are considered, then average s t a l k height and numbers of seed heads per bolted s t a l k increased s i g n i f i -cantly with rosette s i z e (Fig. 40). There was a s i g n i f i c a n t p o s i t i v e . c o r r e l a t i o n (r = .907; 4 D.F.) between s t a l k height and the number of seed heads per s t a l k . I f only spotted knapweed rosettes which bolted i n 1977 are considered, there was no s i g n i f i c a n t difference i n the mean number of seed heads produced with increasing rosette s i z e . % 100-1 ROSETTE SIZE CLASS (NO. LEAVES X DIAMETER) SPRING, 1977 . Figure 37. The percentage of d i f f u s e knapweed ro-settes i n each fate class related to spring s i z e classes Fates: bolted 0 0; vegetative X X; dead and regrowing A A %100'i 0-20 21-40 41-60 61-80 81-100 +100 cm ROSETTE SIZE CLASS (NO. LEAVES X DIAMETER) SPRING, 1977 Figure 38. The percentage of spotted knapweed r o -settes i n each fate class r e l a t e d to spring s i z e classes Fates: bolted 0 0; vegetative X X; dead and regrowing A* • • - A 4>-Hi TO e H fD % IN FATE CLASS OF TOTAL NO. ROSETTES WHICH BOLTED X Hi r1- r h CO 1 Co 3 c fD OJ rt rt vO 1 fD fD h-1 rt . CO CO fD 1 . O CD X P* H o C P* CO cn H P 4 fD Co c H-3 r N Si O -3 Cu ce fD er P4 er CO CO O fD cr o H cn M o fD H l-h CO CO r-1 3 fD C CO H rt rt CO M CD H fD Co rt fD Cu TO fD O Z CO fD Cu 3* It CD fD 3 O • n Cu O r h T p* O ce CO * O Co l-h CD X) rt rt rt sf Ot • c ffi P^ e rt 3 a fD M fD cn xi H Cu C rt rt V! o fD O >« 7? o Cu rt 3 fD Co c Co cn rt M 3 X) cn O cn K H 1 3 3 fD e cr 3 o fD h-1 o 3 n CU H cr fD rt er cn I ro 70 O CO m rn co N rn o co $> -o co ?o co rn CO x o m H m 70 TO 3 H fD AVERAGE STALK HEIGHT 0- -0 Cu Co O H- <! • r h fD r h 3 Co cn TO fD fD fD TO l-i 5^ 3 fD 3 cj CD fo 3 CD cr H-fD o fD ri 3 fD CO CD CU o o H r h l-h o cn CO Co fD fD < rt (D fD rt CU l-l fD Co P 4 TO cn fD fD H< CO N CU CD fD CO rt cn Cu -d M fD !^  i-t 3* CD fD rt r 1-CO TO r- 1 P 4 ?T rt o CU 3 3 Cu 70 O CO m m co HH N m o CO J> "0 CO 70 CO i—t AVERAGE NUMBER OF SEED HEADS PER STALK X X o I o I -p o I O H-• <£> I—• > I _ < oo m o CO X oo H-• 1 -HH t—» o m o m 70 o _ l _ O O o o 9n o o o 3 - X -147 Galls were found only on the d i f f u s e knapweed s i t e . There were averages of 0.03 - .02 and 0.44 - . 09 U_. a f f i n i s and U. quadrif a s c i a t a g a l l s per seed head, r e s p e c t i v e l y , i n 1976, and 0.07 - .02 and 0.14 - .03 U. a f f i n i s and U. q u a d r i f a s c i a t a g a l l s per seed head i n 1977. 5.Discussion: Diffuse and spotted knapweed populations are regulated to some extent by density-dependent mortality at the smallest plant stage measured: the small seedling stage. On the average fewest small seedlings died on plots where thinning rates were highest short of complete seedling removal. The d i f f e r -ence i n average mortality between thinning rates for d i f f u s e knapweed small seedlings was not s i g n i f i c a n t because of unforseen patchiness of the seed-l i n g s , and the influence of plants i n rosette and perennial plant stages. Milthorpe (1961) states that "Competition a r i s e s when one i n d i v i d u a l i s s u f f i c i e n t l y close to another to modify i t s s o i l or atmospheric environment and thereby decrease i t s rate of growth". Where seeds had been washed into hollows and seedlings* grew-as thick as moss, the mortality rate was probably greater with a thinning rate of 75% than on other p l o t s where seedlings were d i s t r i b u t e d more evenly , but where the thinning rate was lower. Harris (1973b) suspected that "most of the natural mortality of the weed [spotted knapweed] i n B r i t i s h Columbia i s from i n t r a s p e c i f i c competition between seedlings and rosettes. I f t h i s i s so, then the most e f f e c t i v e b i o l o g i c a l c o n t r o l agent would be one that attacked immediately a f t e r t h i s m o r t a l i t y . The e f f e c t of agents that feed on seeds or seedlings, even though a p o t e n t i a l plant i s e a s i l y destroyed at t h i s stage, i s merely to reduce the mortality from i n t r a s p e c i f i c competition". Here, rosette numbers 148 were not affected by seedling removal, even on plo t s where a l l seedlings had been removed for two years. The s i g n i f i c a n t d i f f e r e n c e between rosette d e n s i t i e s with removal rate f o r spotted knapweed i s att r i b u t e d to unusually high rosette d e n s i t i e s on one p l o t . Therefore s u r v i v a l of ei t h e r d i f f u s e or spotted knapweed rosettes i s not affected by plants i n the small r o s e t t e , or younger stages of development. However, older plants could a f f e c t the s u r v i v a l of the younger ones i n populations where de n s i t i e s of rosettes and older plants are high. But, most small seedlings are surrounded only by small seedlings, and small seedling s u r v i v a l probably depends mostly on i n t r a s p e c i f i c competition between i n d i v i d u a l s i n t h i s stage. I t appears that seedling removal, even at the lowest rate, can decrease the number of plants surviving to the large;seedling and small rosette stages i n two years, and i f thinning continued, rosette and older plant d e n s i t i e s could be thinned. However, knapweed would:probably respond to a'decrease i n several:ways. Diffuse knapweed already 1 showed signs of an increase i n ro-sette s i z e a f t e r two years of seedling removal. Larger rosettes would be more l i k e l y to bolt"and/produce more seed heads. Although spotted knapweed did not show these: r e l a t i o n s h i p s i n t h i s study, preliminary work on other s i t e s showed that the p r o b a b i l i t y of b o l t i n g and producing more numerous seed heads increases with spotted knapweed rosette s i z e . On s i t e s where spotted knapweed has existed for many years, crowding by perennial plants could decrease average rosette diameter, and root s i z e could be a better i n d i c a t i o n of spotted knapweed rosette vigour rather than the number of leaves and l e a f length. Werner (1975) found;that the,probability of b o l t i n g for the b i e n n i e l , 149 Dipsacus fullonum ( t e a s e l ) , increased with increasing rosette diameter, but was not r e l a t e d to the age of i n d i v i d u a l rosettes. Werner.felt that b i e n n i e l s which grow i n suboptimal conditions or where competition i s great ei t h e r from the same, or other species, remain i n the rosette stage longer than plants growing i n good conditions. Often rosettes growing i n the more marginal locations died before reaching maturity. A c r i t i c a l s i z e , based on food reserves, was probably reached before b o l t i n g . Diffuse and spotted knapweed appear to be s i m i l a r to Werner's (1975) observations. Because only about one-third d i f f u s e , and one-half spotted knapweed rosettes bolted i n 1977, these weeds spend at least one growing season, probably two, i n the f rosette stage. Some i n t r a s p e c i f i c competition between plants i n rosette and older stages occurs, but competition at t h i s stage does not a f f e c t s u r v i v a l . Instead average plant s i z e and average seed y i e l d per i n d i v i d u a l plant would decrease with increasing i n t r a s p e c i f i c competition. If g a l l f l i e s do achieve a reduction i n knapweed density-,-\t>a4-l they may do i s a l l e v i a t e i n t r a s p e c i f i c competition among rosettes, and the end r e s u l t s would be fewer, but larger plants, with no decrease i n knapweed cover on rangeland. Diffuse knapweed, because of i t s shorter l i f e s p a n , would respond to a decrease i n density with a p l a s t i c response f a s t e r than would spotted knapweed. A de-crease i n plant density may also decrease the length of time spent i n the rosette stage because " c r i t i c a l s i z e " may be reached f a s t e r due to l e s s i n t r a s p e c i f i c competition among rosettes. Controlled germination may play a part i n knapweed population regulation. In 1976, although the d i f f e r e n c e was not s i g n i f i c a n t , i t appeared that as more seedlings were removeda-greater percentage of seeds germinated to take t h e i r place. In 1977 the opposite trend occurred: s i g n i f i c a n t l y fewer small 150 seedlings appeared with increasing"thinning rates on both d i f f u s e and spotted knapweed s i t e s . The seed bank seems to have been depleted on p l o t s with highest thinning rates. This may have occurred because the s o i l i s s l i g h t l y disturbed when seedlings are removed, or because seedlings continued to germinate i n 1976 as long as there was space for them. Controlled germi-nation i s studied i n a l a t e r section where seeds are a c t u a l l y sown on the surface of the s o i l , and germination rates over a range of sowing den s i t i e s are compared. This study demonstrates several ways i n which knapweed i s drought- and _' damage-tolerant. Plants i n any stage of development, even i n the small seedling stage, can lose**teheir leaves during dry periods of the growing season, and regrow them when there i s s u f f i c i e n t moisture. Small and large seedlings can remain i n t h e i r respective stages for nearly one month, and continue growth i n more favourable conditions. When seedlings are damaged they often survive i f some of the root remains i n t a c t . Although spotted knapweed bolt s more r e a d i l y than d i f f u s e knapweed, bolted spotted knapweed plants may not produce any heads, or they may develop heads only to the . small-bud stage, and then abort them. The following spring these bolted plants produce new rosettes at t h e i r bases, and usually b o l t and bear mature seed heads. This i s an example of an advantage which spotted knapweed, a s h o r t - l i v e d perennial, has over d i f f u s e knapweed. If adverse conditions set i n i n midsummer, i . e . a hot, dry s p e l l , spotted knapweed can cease growth and bear seed the following year, but for d i f f u s e knapweed i t i s an " a l l or nothing attempt" at reproduction for most bolted plants. Usually, however, d i f f u s e knapweed has s u f f i c i e n t stored root reserves to produce some mature seed heads even i n very dry years. 151 5. Conclusions: 1. Diffuse and spotted knapweed populations are c o n t r o l l e d to some extent by density-dependent s e l f - t h i n n i n g at the small seedling stage. 2. Because at l e a s t one-half d i f f u s e and spotted knapweed rosettes remain vegetative, these weeds probably spend one year, or more, i n the rosette stage i n addition to the year i n seedling and small rosette stages, before b o l t i n g . 3. The p r o b a b i l i t y of b o l t i n g increases and the p r o b a b i l i t y of dying decreases with increasing rosette s i z e for d i f f u s e and spotted knapweed. 4. Diffuse and spotted knapweed are so drought tolerant that they can lose a l l t h e i r leaves even at young stages of development, and regrow them when conditions are favourable. 5.Spotted knapweed rosettes bolt'more r e a d i l y than do d i f f u s e knapweed .. • rosettes, but spotted knapweed often does not -produce seed heads even when i t b o l t s . 6. D i f f u s e knapweed may show a p l a s t i c response to thinning by an increase i n rosette s i z e . B. SURVIVAL AND FERTILITY SCHEDULES 1.Introduction: Demography i s the branch of ecology which describes and explains the changes i n numbers of populations (Harper and White, 1974). Demographers measure b i r t h , death, immigration and emigration rates of populations. The measure-ment of death rates (or survivorship rates) can be tabulated into survivor-ship schedules. Deevey (1947) c a l l e d survivorship schedules " l i f e tables". He reviewed both v e r t i c a l ( s t a t i c ) and h o r i z o n t a l (cohort) l i f e tables, with examples from animal ecology. Horizontal l i f e tables follow the s u r v i v a l 152 of a cohort of i n d i v i d u a l s from b i r t h to death, whereas m o r t a l i t i e s of i n d i - \ viduals of a known age, and s p e c i f i c time, are measured for s t a t i c l i f e tables. Insect ecologists have found that developmental stages are the most convenient time units f or l i f e tables. For example, Dempster (1975) compiled a s t a g e - s p e c i f i c l i f e table f or the moth, T y r i a jacobaea. There are numerous examples of animal l i f e tables. Most plant e c o l o g i s t s , however, have studied the s u r v i v a l of only s p e c i f i c stages i n the l i f e h i s t o r y of a plant: for example, seed (Sarukhan, 1974; Harper, Williams and Sagar, 1965), seedling (Amor and H a r r i s , 1975; Piggin, 1976; Ross and Harper, 1972), and shoot (Hutchings and Barkham, 1976; Bernard, 1976). Harper and White (1970) found that although "Charting the fate of i n d i v i d u a l plants (to grow, to flower, to remain vegetative, to die) has been undertaken by few people" some work has been done by Tamm (1956), Sagar (1959) and by the Moscow School of Rabotnov and co-workers from 1940, onwards. Recently the population dynamics of some annuals (Watkinson and Harper, 1978; Regehr and Bazzaz, 1979) and a perennial (Baskin and Baskin, 1979) have been studied. "Observations on the longevity of populations, rates of turnover i n t h e i r vegetative and reproductive composition, etc. are extremely laborious" (Harper and White, 1970). The term "population turnover" has been loosely used by several plant ecologists (for example, Antonovics, 1972, i n h i s study of the population dynamics of Anthoxanthum odoratum), and'is a measure of the average l i f e s p a n of i n d i v i d u a l s i n the population. The average l i f e s p a n , or longevity of plants, may vary greatly from one habitat to another. R botnov (c i t e d by Harper and White, 1970) found that T r i f o l i u m pratense (red clover) had a short l i f e s p a n ; i . e . , a rapid population turnover, on a f l o o d p l a i n , where 153 the plants grew from seedlings to maturity i n two to three years, and flow-ered once before t h e i r death. In contrast, red clover which grew i n a sub-alpine meadow, did not flower u n t i l i t was 5 to 10 years old, and reproduced repeatedly for 10 years. Thus some i n d i v i d u a l s reached 20 years i n age, and the turnover, or replacement, i n t h i s population was slow. The rate of population turnover for both d i f f u s e and spotted knapweed i s an important population a t t r i b u t e in* t h i s study because g a l l f l i e s a f f e c t the plants by decreasing the number of v i a b l e seeds, and thus the p o t e n t i a l \ number of seedlings and mature plants. Diffuse and spotted knapweed r e -produce only from seed. I f the f l i e s destroy a l l the seeds every year, then the minimum length of time required f o r eradication of the knapweed popula-,. t i o n would be the turnover rate of the population. Due to lack of time d i f f u s e and spotted knapweed turnover rates on rangeland are found by con-st r u c t i n g v e r t i c a l l i f e tables, rather than cohort l i f e tables. This section also compares the population dynamics of d i f f u s e and spotted knapweed i n small seedling, large seedling and small rosette stages during the 1977 growing season. 2.Method: Except where noted, control p l o t s ( i . e . , 0% removal of seedlings) from the previous study (section I II A) on the e f f e c t s of thinning rates on d i f f u s e and spotted knapweed were used. The s u r v i v a l rates and f e r t i l i t y schedules for plants i n small rosette and older stages were determined for each i n d i ^ v i d u a l plant because they were tagged with a number. But s u r v i v a l rates for plants i n small and large seedling stages were determined from the t o t a l numbers which appeared i n each stage and the number which died during the 1977 growing season. 154 Survival and f e r t i l i t y tables were formulated a f t e r Sharitz and McCormick (1973), where x = plant stage, D^ = duration of stage (months), Months = calendar months spent i n the stage, A = t o t a l age of the s t a r t i n g cohort (months), 1 = number of i n d i v i d u a l s entering the stage, d = number of i n d i v i d u a l s dying during the stage, q = the percentage of i n d i v i d u a l s dying during each stage, L = the proportion of i n d i v i d u a l s surviving to the beginning of a stage, = 1 /1000, X m^  = the average number of seeds produced per plant f o r each stage, V = L m , X X X R 0 = 2V , i . e . , the average number of seeds produced by each plant that sur-vived to maturity out of the s t a r t i n g cohort of seeds. This i s c a l l e d the net reproductive rate (adapted from Wilson and Bossert, 1971), G = A ^ /R 0» i - e « 5 t n e mean generation time, which i s the mean period of time elapsing between the b i r t h of parents and the b i r t h of t h e i r o f f -spring (Krebs, 1972). Dif f u s e and spotted knapweed plants were assumed to increase i n age as thev increased i n s i z e . Therefore d i f f u s e knapweed grew consecutively through small seedling, large seedling, small rosette, rosette and perennial plant stages, where each stage i s defined as i n the previous section (III A). Likewise, spotted knanweed grew from small seedling to rosette stages, and then developed through small, medium, and large perennial plant stages, where these stages are defined as i n the previous section (III A). 155 Although preliminary work showed that both species could germinate ei t h e r i n the spring or i n the f a l l , here they were assumed to germinate i n the spring. Spotted knapweed seeds are shed from seed heads upon maturity, but d i f f u s e knapweed seeds could remain ins i d e seed heads u n t i l the following spring, when they f a l l to the ground, and germinate. Watson (1972) found that d i f f u s e knaoweed seeds, c o l l e c t e d from seed heads which had remained i n the f i e l d during the winter, were 88% v i a b l e . Therefore the d i f f u s e knapweed survivorship schedule was started with a cohort of 1000 seeds, which had a death rate of 12% from November u n t i l A p r i l . Because spotted knapweed seed has a higher percentage v i a b i l i t y than does d i f f u s e knapweed (section II B v i ) , the death rate for the s t a r t i n g cohort of 1000 spotted knapweed seeds was ar b i t r a r i l y assigned as 10% from November u n t i l A p r i l . Diffuse and spotted knapweed seeds were assumed to germinate at the beginning of May and to spend one month (May) i n the small seedling stage, and the following month i n the large seedling stage. The re s t of the growing season (July to October) was spent i n the small rosette stage. M o r t a l i t i e s f o r each stage were estimated from control p l o t s . In May, 1978, tagged small rosettes were c l a s s i f i e d as a l i v e or dead to f i n d m o r t a l i t y rates f o r the small rosette stage from November u n t i l A p r i l . Bv the beginning of A p r i l a l l surviving small rosettes were assumed to enter the rosette stage. Previously (section I I I A) i t was found that 66% of d i f f u s e and 46% of spotted knapweed plants i n the rosette stage remained vegetative. Therefore t h i s stage was subdivided into vegetative and bolted stages. Both species were assumed to remain i n each of these sub-stages f o r one year. The mortality rate f o r d i f f u s e knapweed i n the vegetative rosette stage was 156 estimated for two periods of time, May to October and November to A p r i l , by observing the fates of tagged rosettes i n October and May. The death rate of d i f f u s e knapweed i n the bolted rosette stage from May to October was found by observing the fates of bolted rosettes i n October. Most bolted d i f f u s e knapweed rosettes die during the winter, but some may produce mature stalks i n following years. Preliminary analysis of the data presented i n the previous section (III A) showed that 8 - 3% of the bolted plants (bolted rosettes + perennial plants) on the d i f f u s e knapweed s i t e were i n the per-ennial stage. Therefore i t was assumed that from November u n t i l A p r i l 92% of the bolted d i f f u s e knapweed rosettes died. The surviving plants were assumed to enter the perennial stage i n May, b o l t during the summer, and die by October. The mortality rates f o r spotted knapweed i n vegetative and older plant stages were estimated from May to October, and from November u n t i l A p r i l , by recording fates of tagged plants i n May and October. These estimates were obtained by considering a l l t h i r t y plots i n the previous section, be-cause seedling removal had had no e f f e c t on spotted knapweed s u r v i v a l or s i z e for vegetative rosette and older stages. A l l spotted knapweed plants were assumed to die a f t e r b o l t i n g i n the large perennial plant stage. In September, 1977, d i f f u s e knapweed seed heads were a r b i t r a r i l y c o l l e c t e d from plants outside the study p l o t s . The heads were opened and the number of seeds i n heads containing no g a l l s were counted. I t was assumed that the average number of seeds per seed head, thus found, was not affected by g a l l - f l y attack at low d e n s i t i e s . The average number of seeds produced per plant for each stage was found by multiplying the average number of seed 157 heads per plant by the average number of seeds per seed head. Because d i f f u s e knapweed perennial plants were so few (only 1.6 - .7 per plo t ) and seedling removal had not affected them, a l l t h i r t y plants i n the previous section were used to estimate the average number of seed heads per plant f o r t h i s stage. Watson's (1972) estimate of 26.6 seeds per seed head was taken to be the average number of seeds per seed head on the spotted knapweed s i t e because no counts were made on the study s i t e i n 1977. Few spotted knapweed plants were found i n the older stages and seedling removal had had no e f f e c t on seed y i e l d , therefore a l l t h i r t y plots i n the previous section were used to e s t i -mate the average number of spotted knapweed seed heads per plant f o r bolted rosette and older stages. Numbers of small seedlings, large seedlings, and small rosettes had been counted once every two weeks on control p l o t s i n the previous section (III A). These data are presented here as average numbers of plants i n each stage throughout the growing season. 3.Results: Survivorship of d i f f u s e and spotted knapweed follows a Deevey type I I I curve (Deevey, 1947) where mortality i s high during early l i f e , but the few i n d i -viduals that do survive to older stages of development have a r e l a t i v e l y constant expectancy of further l i f e (Tables XXVI and XXVII). On the average, more spotted than d i f f u s e knapweed plants died during the f i r s t three stages of development. For both species the mortality rate at the small seedling stage was greatest (73 - 6% and 84 - 3% for d i f f u s e and spotted knapweed, r e s p e c t i v e l y ) , but by the small rosette stage m o r t a l i t y rates had dropped 158 Table XXVI. Survivorship and f e r t i l i t y of d i f f u s e knapweed X D Months A 1 d x X X X S 6 Nov-Apr 6 1000.00 120. 00 SS 1 May 7 880.00 642. 40 LS 1 June 8 237.60 42. 77 SR 4 Jul-Oct 12 194.83 40. 91 SR 6 Nov-Apr 18 153.92 21. 55 R 6 May-Oct 24 132.37 25. 15 R V 6 Nov-Apr 30 107.22 12. 87 6 May-Oct 36 94.35 1. 89 6 Nov-Apr 42 92.47 85. 07 P b 6 May-Oct 48 7.40 7. 40 X m V A V X X X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 23 13.96 502. 70 +> 0 0 1 37 .84 40. 14 Ro = 14.80 542. 84 12 73 18 21 14 19 12 2 92 100 A A + + + + + + + A A A L.00000 .88000 .23760 .19483 .15392 .13237 .10722 .09435 .09247 .00740 148 113 -G = 36.68 months A plant stages: S=seed; SS=small seedling; LS=large seedling; SR=small ro-sette; R =vegetative rosette; R^=bolted rosette; P=perennial plant A A V B seed mortality estimated f rom Watson (1972) A A A Bolted rosette m o r t a l i t y estimated to be the percentage of bolted rosettes (R^) of the t o t a l number of bolted plants (R^ + P) Table XXVII. Survivorship and f e r t i l i t y of spotted knapweed x D Months A x x m x x A V x x x 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 .23 44.28 0 0 9 2 .26 108.48 0 0 10 3 .01 180.60 0 0 25 7 .23 520.56 Ro = 13 .73 853.92 S 6 Nov-Apr 6 1000. 00 100 .00 SS 1 May 7 900. 00 756 .00 LS 1 June 8 144. 00 109 .44 SR 4 Jul-Oct 12 34. 56 12 .44 SR 6 Nov-Apr 18 22. 12 6 .64 R R V 6 May-Oct 24 15. 48 5 .73 6 Nov-Apr 30 9. 75 .28 % 6 May-Oct 36 9. 47 .38 6 Nov-Apr 42 9. 09 .73 SP 6 May-Oct 48 8. 37 1 .25 SP 6 Nov-Apr 54 7. 11 .57 MP 6 May-Oct 60 6. 54 .52 MP 6 Nov-Apr 66 6. 02 .24 LP 6 May-Oct 72 5. 78 5 .78 A A 1 0 + 84 J 3 7 6 - 5 36 - 10 30 - 16 37 - 5 2.9 J .9 4 + 2 8 ^ 2 1 5 - 4 8 + 3 8 - 2 4 - 4 100 1.00000 .90000 .14400 .34560 .22120 .15480 .09750 .09470 .09090 .08370 .07110 .06540 .06020 .05780 0 0 0 0 0 0 1 3 -27 -46 125 -G A 62.19 months plant stages: S=seed; SS=small seedling; LS=large seedling; SR=small ro-sette; R^=vegetative rosette; R ^ b o l t e d rosette ^ gP, MP, and LP=small, from Watson (1972) medium and large perennial plant, r e s p e c t i v e l y ; 4001 300' 200J 159 1001 AUGUST KEPTFMBFR OCT. Figure 41. Average numbers of d i f f u s e and spotted knapweed small seedlings per p l o t i n 1977 d i f f u s e knapweed 0 0; spotted knapweed X X 160 + + considerably (21 - 5% and 36 - 10% for d i f f u s e and spotted knapweed, res-pectively) . When d i f f u s e and spotted knapweed reached rosette and older plant stages, mortality rates were less than 20%, except f o r spotted knap-weed vegetative rosettes, of which 37 - 5% died during the summer. Spotted knapweed had a longer average l i f e expectancy (62.19 months, or approxi-mately f i v e years) than d i d d i f f u s e knapweed (36.68 months, or approximately three years). The mean number of seeds per seed head on the d i f f u s e knapweed s i t e was 10.7 - .3. Given that the mean number of spotted knapweed seeds per head was 26.6 (Watson, 1972), reproductive rates f o r d i f f u s e and spotted knapweed populations were s i m i l a r : 14.8 and 13.73, r e s p e c t i v e l y . Diffuse and spotted knapweed small seedlings were present on the s i t e s i n A p r i l (Fig. 41). These were i n d i v i d u a l s which had germinated during the previous f a l l because cotyledones were lacking, or withered, and seedling hue was dark green or red rather than the bright green t y p i c a l of newly-germinated plants. Diffuse and spotted knapweed small seedlings appeared at almost a l l times during the growing season, but t h e i r numbers were s i g -n i f i c a n t l y greater i n May, a f t e r which time they declined to n i l i n the middle of August. In the f a l l there was a f l u s h i n seedling germination. Diffuse and spotted knapweed could also survive the winter i n the large seedling stage because numerous large seedlings were already present on the plots i n A p r i l , before any newly germinated small seedlings had appeared (Fig. 42). There was no s i g n i f i c a n t difference between average large seedling counts throughout the growing season, but, l i k e small seedlings, greatest numbers of large seedlings appeared to occur i n May. Average large seedling counts, however, did not drop to n i l i n August as d i d small seedling 60-i 50 40-30-20-10-161 4- ' " MAY | JUNE JULY AUGUST SEPTEMBER OCTOBER Figure 43. Cumulative numbers of d i f f u s e and spotted knapweed small rosettes per p l o t i n 1977 d i f f u s e knapweed 0 0; spotted knapweed X X counts. The f i r s t d i f f u s e knapweed small rosettes appeared i n the middle of Mav and the f i r s t for spotted knapweed appeared i n the middle of June (Fig. 43) . The cumulative t o t a l d i f f u s e and spotted knapweed small rosettes increased u n t i l the middle of August, a f t e r which few a d d i t i o n a l small rosettes appeared. The rate of appearance of d i f f u s e compared with spotted knapweed small r o -settes was s i g n i f i c a n t l y greater, and by the end of the growing season more d i f f u s e than spotted knapweed small rosettes had been produced (Fig. 44). 5.Discussion: Krebs (1972) compared a t t r i b u t e s of r - ("big-bang") and K- (repeated r e -producers) selected organisms. The terms r - and K-selection are r e l a t i v e , with r-selected organisms having a more rapid developmental time, reproducing once, reproducing early i n l i f e , having a smaller s i z e , competing l i t t l e i n t r a s p e c i f i c a l l y and having a low density-dependent m o r t a l i t y compared with LOG 1 Q C T I M E ) Figure 44. Diffuse and spotted knapweed regressions of the cumulative number of small rosettes produced per plo t i n 1977 (Y), where X = log^ (time). For d i f f u s e knapweed time = 1, 2, 3...11, where 1 = May 18th; 2 = June 1; 3 = June 15th ... 11 = October 5th. For spotted knapweed time = 2.5, 3.5, 4.5...10.5, where 2.5 = June 8th; 3.5 = June 22nd; 4.5 = July 6th ... 10.5 = September 28th. Both regre-ssions are s i g n i f i c a n t , and there i s a s i g n i f i c a n t d i f f e r e n c e be-tween "b", where Y = a + bX. d i f f u s e knapweed 0 0; spotted knapweed X - X 163 K-selected organisms. In the Plant Kingdon, r-selected organisms are usually annuals, which must recolonize every year, and have Deevey type III s u r v i v a l curves ( i . e . , m o r t ality i s greatest during early stages of l i f e , but older i n d i v i d u a l s have r e l a t i v e l y constant expectancies of s u r v i v a l ) . K-selected organisms are perennials and have Deevey type I s u r v i v a l curves ( i . e . , mor-t a l i t y i s low during early l i f e , but increases with age), or Deevey type II s u r v i v a l curves ( i . e . , m o r t ality i s constant throughout l i f e ) . D i f f u s e and spotted knapweed are neither annuals, nor perennials, and have a t t r i b u t e s of both r - and K-selected organisms. They both have Deevey type I I I s u r v i v a l curves, but spotted knapweed, with a longer l i f e expectancy, repeated r e -production, and a larger plant s i z e i s more t y p i c a l of a K-selected organism than i s d i f f u s e knapweed. Harper and White (1974) c l a s s i f y b ienniels as "big-bang" reproducers: "winter annuals, also b i e n n i a l s , have a more or l e s s long vegetative stage that changes under p e r i o d i c s t i m u l i to a flowering phase; vegetative apices are converted to a flowering condition and a "big-bang" of reproduction i s followed by death." Harper and White (1974) c i t e several examples where plants may grow vegetatively i n the rosette stage f o r several years before flowering, but c l a s s i f y them as b i e n n i e l s because death follows b o l t i n g and copious seed production. They maintain that the b i e n n i e l habit i s p a r t i c u -l a r l y advantageous, compared with the annual habit, i n environments where seedling establishment i s r i s k y . Even i f a l l seedlings p e r i s h during one growing season, rosettes are l i k e l y to survive and produce seed i n a l a t e r year. In t h i s study i t was assumed that d i f f u s e knapweed spent only one year in the vegetative rosette stage, but because 66% of the rosettes i n 1977 did not b o l t , t h i s phase of l i f e i s l i k e l y to l a s t two or more years, and the 164 mean generation time may be greater than three years. Spotted knapweed bolted more r e a d i l y than did d i f f u s e knapweed because only 46% of the rosettes remained vegetative i n 1977. However, seed y i e l d during the f i r s t year of maturation was much less (13 - 1 seeds per i n d i v i d u a l ) than for d i f f u s e knapweed (148 - 23 seeds per i n d i v i d u a l ) . Because spotted knapweed may produce ever-increasing amounts of seed i n following growing seasons, the weak attempt at reproduction during the f i r s t year of maturation i s compensated f o r , and i n t h i s study the reproductive e f f o r t of spotted knapweed was almost i d e n t i c a l to that of d i f f u s e knapweed. However, future study may f i n d that spotted knapweed has "on" and " o f f " years of flowering, as does Centaurea j aceana (Harper and White, 1974), such that years of veg-etative growth may occur between years of sexual reproduction. "The a b i l i t y of plants to p e r s i s t for long periods as j u v e n i l e or immature plants on very low nutrient and l i g h t resources, i s considered by Rabotnov to be of great importance i n the structure of perennial herbaceous plant populations. The d i v e r s i t y of 'age stage' composition i n plant populations (and the spectrum includes the buried, v i a b l e seed populations) enable species, i n s p i t e of considerable changes of environment, to p e r s i s t as components of communities, sometimes with quite small changes i n the numbers of i n d i v i d u a l s . " (Harper and White, 1970) D i f f u s e and spotted knapweed plants, which are i n vegetative rosette stages, or which grow vegetatively between years of flowering, comprise 25% to 75% of the i n d i v i d u a l s older than the small rosette stage. Even i f they produce no seed, they occupy space, and prevent other species from e s t a b l i s h i n g . Therefore, g a l l f l i e s , which can attack only plants which flower, a f f e c t , what may be le s s than 165 half the population older than the small rosette stage. Thus the " d i v e r s i t y of 'age stage' composition" (Harper and White, 1970) may also enable the plant population to p e r s i s t i n spite of intense attack by insects of mature plants. The lengthier the vegetative stage, the slower i s the turnover rate of the plant population. Minimal turnover rates for these studies were es-timated to be three and f i v e years for d i f f u s e and spotted knapweed, r e s -p e c t i v e l y . Therefore the f l i e s would have to eliminate a l l seeds for those periods before the plant populations could be eradicated. In other habitats, where longevity i s even greater, complete seed destruction would have to occur for even longer periods. Werner (1975) found that Dipsacus fullonum rosettes could remain vegetative for as long as f i v e years. D i f f u s e knapweed could conceivably p e r s i s t for f i v e years i n a vegetative stage, but on the average i t has a shorter l i f e s p a n than does spotted knapweed. Therefore a seed-reducing b i o l o g i c a l control agent, which t h e o r e t i c a l l y destroys a l l seeds, would eradicate a d i f f u s e knapweed population f a s t e r than a spotted knapweed population. The g a l l f l i e s were able to reduce the p o t e n t i a l seed y i e l d by only about 80% on both d i f f u s e and spotted knapweed release s i t e s at peak population numbers i n 1977 (section II B i x ) . At t h i s rate the reproductive e f f o r t s (R 0) for the d i f f u s e and spotted knapweed populations i n t h i s study are greater than one and the plant populations are increasing. For R 0 to be decreasing, or given that R0= 0.999, the g a l l f l i e s would have to reduce 94% and 93% of the seed y i e l d on the d i f f u s e and spotted knapweed popula-tion s , r e s p e c t i v e l y . I t i s possible that R 0 could be l e s s than 1.0 during some years even without the g a l l f l i e s . In 1977, the average number of seeds per seed head for a spotted knapweed population i n Westwold, close to t h i s 166 study s i t e , was only 1 1 - 2 (Table XXII), and i f t h i s value i s used instead of Watson's (1972) estimate of 26.6 seeds per seed head, then Ra= 0.86. Thus i f conditions that did not increase the seed y i e l d p e r s i s t e d , t h i s theore-t i c a l spotted knapweed population would decrease i n density. Possibly R0 i s l e s s than 1.0 for most years for some knanweed populations, and the persistence of these populations depends on a bumper seed crop (say, once i n f i v e years) which produces a r e s e r v o i r of seeds, rosettes and j u v e n i l e perennial plants. Because g a l l f l i e s decrease seed y i e l d i n unattacked heads i f some heads on the same plant have been attacked, the estimate of 10.7 - .3 d i f f u s e knapweed seeds per seed head i s probably low i n t h i s study, and R D may be s l i g h t l y underestimated. Although R c for d i f f u s e and spotted knapweed populations i n t h i s study were s i m i l a r , R„ values probably f l u c t u a t e greatly with vear, l o c a t i o n and knapweed species. Plant l i f e tables should be made over lengthv periods. Two to three years was considered too short to develop l i f e tables for Plantago and Ranunculus spp. (Harper and White, 1974). Antonovics (1972) studied populations of Anthoxanthum odoratum for s i x years to obtain rates of turnover on a zinc mine s i t e . "Measures of population turnover... are revealed...only i f plants are marked for repeated observations... Stable vegetation i s . . . i n a state of continuous f l u x i n which the rates of turnover are c r i t i c a l c h a r a c t e r i s t i c s of the s t a b i l i t y . " (Harper, 1967). In t h i s study, only plants i n small rosette and older stages were marked for observation, but more useful i n f o r -mation could be obtained i f seedlings were also marked and followed through the i r l i f e s p a n . Seedlings of some species develop i n t o mature plants faster 167 i f they germinate i n the f a l l , rather than i n the snring, but those that germinate i n the f a l l r i s k death during the winter. I f seedlings are not marked, then they must be counted frequently enough to e s t a b l i s h death rates for a l l cohorts which germinate (Harper and White, 1974). Intervals of two weeks appear to be short enough to determine death rates of d i f f u s e and spotted knapweed seedlings, provided the researcher can d i s t i n g u i s h between newly-germinated small seedlings and small seedlings which may be older than two weeks. Age i s usually based on plant s i z e . Plant age, however, i s often unrelated, or only weakly r e l a t e d , to seed y i e l d , e s p e c i a l l y i n b i e n n i e l s , where r o-settes must a t t a i n a c r i t i c a l s i z e before b o l t i n g . This c r i t i c a l s i z e can be attained quickly under good growing conditions, but i f competition for nutrients, l i g h t and space are great, the c r i t i c a l s i z e may take longer to a t t a i n (Werner, 1976). Thus l i f e tables for some plants may be termed " s i z e -s p e c i f i c " rather than " s t a g e - s p e c i f i c " as i n insect ecology, or " a g e - s p e c i f i c " as i n animal ecology. Diffuse and spotted knapweed mortality at the small seedling stage appears to be dependent on weather conditions to some degree. In 1977 there were two hot s p e l l s without r a i n f a l l during the growing season; the f i r s t was from July 19th to 27th (Table XXVIII), when the mean temperature was 17 - 1 °C i n Kamloops (the nearest weather s t a t i o n to the d i f f u s e knapweed s i t e ) , and 19 -l"C i n Westwold (Table XXXIX). The second hot, dry s p e l l occurred from August 5th to 17th, when the mean d a i l y temperatures i n Kamloops and Westwold were 24.3 - .5°C and 19.4 - .5°C, respectively. During the l a t t e r hot s p e l l small seedling numbers dropped to n i l for both d i f f u s e and spotted knapweed, 168 Table XXVIII. T o t a l d g i l y p r e c i p i t a t i o n (mm) for Kamloops and Westwold i n 1977 Day A p r i l May June July August September K w K w K w K w K W K w 1 T 0.8 0 0.3 0 0 0 0 0 0 0 0 2 0.2 0 4.1 3.1 T 0 1.2 0.5 0 2.8 1.2 2.3 3 0 0 0.8 2.0 T 0 2.6 2.3 5.4 0 1.2 0 4 0 0 T 0.5 0.4 3.8 0.2 1.3 T 0 1.4 T 5 0 0 T 2.5 0 0 0.5 1.8 0 0 0 0 6 0 0 0.6 0 0 0 T 0 0 0 0 0 7 0 0 0 0 0 0 0 0 0 0 4.4 2.8 8 T 7.6 0 0 T T 0 0 0 0 0 0 9 2.6 0 0 0 T 0 0 0 0 0 0 0 10 T 0.5 0.2 2.0 0 0 0 9.9 0 0 0 0 11 0 0 1.1 0 0 0 7.1 1.5 0 0 0 0 12 0 0 0 0 0 0.7 1.4 2.8 0 0 0 0 13 5.8 0 0.2 T 6.7 0 0.7 2.0 0 0 0 0 14 0.7 0.5 T 0 2.5 2.0 0 0 0 0 2.4 0 15 0.7 0.3 0 T T 0 0 0 0 0 0 0 16 2.7 0.5 T 0 0 0 1.9 T 0 0 0 0 17 0 0 0 0 0 0 0 0 0 0 T 0.8 18 T 0 0 0 0 0 T 0 3 0 T 1.5 19 T 2.0 T 0 6.9 1.5 0 0 0 0 1.4 4.3 20 0 0 T 0 T 0 0 0 0 0 T 2.3 21 0 0 0 0 0 0 0 0 1.5 T 0.9 0 22 0 0 0 0 0 0 0 0 0.6 0.8 0.8 5.8 23 0 0 5.7 11.4 0 0 0 0 1.7 2.5 12.1 1.0 24 0 0 8.9 0 0 0 0 0 3.4 T 0.7 0.5 25 0 1.0 0 1.3 0 0 0 0 T 0.5 0 0 26 1.8 0 0.6 0 0 0 0 0 12.5 6.1 0 0 27 0 0 0 1.0 T 0 0 0 T 0 0 0 28 0 0 T 0.3 0 0 9.0 10.9 2.2 7.4 0 0 29 T 0 0 0 0 0 0 0 0.5 2.5 0 0 30 0 0 0 0 0 0 0 0 4 T 0 0 31 - - 0.4 0.8 - - 0 0 0 0 - -Climate of B r i t i s h Columbia, pub. by the B.C. M i n i s t r y of A g r i c u l t u r e T = trace; K= Kamloops, W= Westwold but i f the summer had been cooler and wetter, such as i n 1976, large numbers of replacement rosettes could have been produced. Thus i n t r a s p e c i f i c com-p e t i t i o n may play a part i n the regulation of knapweed populations only when weather conditions favour germination and seedling s u r v i v a l . Seedlings may appear at any time during the growing season, but a great deal of r a i n may have to f a l l i n midsummer a f t e r a drought to r a i s e the s o i l moisture l e v e l 169 Table XXIX. Mean d a i l y temperature (°C) for Kamloops and Westwold i n 1911* A p r i l K W May K W June K W July K W August K W September K 5.5 5.1 12.6 13.2 9. 9. 15. 13. 8. .8 .9 ,3 6.8 5.8 8.3 9.9 5.9 10.1 7.4 4.0 7.1 5.0 5.4 8.0 17.1 17.6 18.3 18.7 14.7 10.5 12.7 14.2 12.6 2.0 2.0 8.6 8.9 7.5 8.3 10.3 10.0 5.0 3.1 4.7 5.3 8.6 4.1 4.4 1.7 2.8 4.0 2.8 4.4 4.2 15. 12. 15.0 14.7 10.6 14.6 8.9 14.2 9.7 13.9 13.7 12.4 11.3 8.0 12.1 15.5 16.0 14.7 12.2 11.4 11.7 13.2 12.1 9.1 14.5 14.1 14.2 15.9 15.3 18.6 15.5 13.1 12.9 12. 15. 10. 10.6 14.9 15.8 16.0 .2 ,1 .4 10. 8. 9. 8. 6. 8. 12. 8. 11. 9. 10.0 7.8 9.5 8.3 6.1 10.3 11.1 9.5 12.0 12.8 12.5 10.0 10.3 10.6 8.1 11.4 5.6 7.8 11.4 10.0 12.2 17.2 14.7 13.3 16 18 24 24 18 15 16.4 19.1 18.4 18.8 18.3 18.0 21.6 22.1 23.8 24.3 19.6 19.2 18.6 20.4 22.0 17.7 20.6 18.6 20.9 18.7 19.6 12.2 11.4 14.5 13.6 13.4 19.2 20 18 11 11 13 14 15 14 14 16 16.7 19.2 19.7 17.8 16.4 18.6 16, 19. 16. 16. 14. 16. 15. 14. 23.0 15.4 13.6 15.0 15.7 17.5 18.6 22. 20. 21. 19. 17.0 20.1 19.2 22.2 18.8 16.6 16.4 16.8 21.1 23.9 21.2 23.1 22.6 25 26 25 21 17 19.0 21.5 18.9 13.9 11.4 9.5 10.3 14.2 13 18 17 17 16 15.3 17.5 15.8 19.2 16.7 13.9 13.6 12.2 15.3 19.2 16.7 17.2 18.6 24.2 22.8 20.6 31.6 13.9 17.0 18.3 23.4 25.5 25.6 22.9 23.2 22.5 24.0 23.3 25.5 23.6 25.3 26.3 28.5 25.4 22.5 22.5 23.7 24.8 27.5 25 23 23 18 16 15 15 16.0 17.9 13.6 13.4 14.5 17.5 18.8 22 19 18, 17, 18, 20.0 20.3 19.5 20.8 22.2 22.5 20.8 17.5 16 18 19 21 21.1 20.3 21.4 17.5 15.0 13.9 13.3 12.5 12.5 11.4 9.2 10.3 15 13 17 18 15 15 14 15 15.8 15.9 15.3 16.5 17.3 13.8 12.2 ,1 ,3 ,3 ,8 ,3 .4 .4 11. 15. 14. 13. 13. 11. 12, 10.1 10.8 9.9 10.3 8.6 9.3 12.6 10.3 Climate of B r i t i s h Columbia, pub. by the B.C. M i n i s t r y of A g r i c u l t u r e K = Kamloops, W = Westwold s u f f i c i e n t l y f o r germination, because "The longer and more severe the drouth the greater the quantity of r a i n f a l l that w i l l be required subsequently to break i t . " (Daubenmire, 1959) 5.Conclusions: 1.Diffuse and spotted knapweed have Deevey type I I I s u r v i v a l curves. 170 2. Diffuse knapweed i s more t y p i c a l of an " r - s e l e c t e d " organism, and spotted knapweed i s more t y p i c a l of a "K-selected" organism. 3. Diffuse knapweed, with a fa s t e r population turnover rate, produces more small rosettes each year than does spotted knapweed. 4.In t h i s study reproductive e f f o r t s of d i f f u s e and spotted knapweed were s i m i l a r . 5.Small seedlings usually appear i n the spring or i n the f a l l . 6. Dry, hot weather i n the summer increases small seedling death, but a few large seedlings and small rosettes usually survive to take the place of mature plants which die. 7. Diffuse and spotted knapweed can survive the winter i n any stage of develop-ment . C. KNAPWEED SEED SOWING TRIALS 1.Introduction: Population d e n s i t i e s of some plant populations are regulated by " c o n t r o l l e d germination", where germination rates are suppressed with increasing seed densities (Palmblad, 1968). Diffuse and spotted knapweed produce widely varying seed d e n s i t i e s on rangeland. For example, at White Lake Observatory there were only 66 - 13 d i f f u s e knapweed seeds per 50 X 50 cm, while i n Lumby-1 there were 6628 - 1228 d i f f u s e knapweed seeds per 50 X 50 cm (section II B v i i ) . Preliminary observations i n the f i e l d showed that, even i f av-erage seed y i e l d s were low, den s i t i e s of seedlings i n small patches could be very high. One purpose of t h i s study was to determine i f d i f f u s e and spotted knapweed exhibited a con t r o l l e d germination response to an increasing seed density. 171 Watson (1972) found that spotted knapweed i s more prevalent i n mesic habitats i n the I n t e r i o r of B r i t i s h Columbia, while d i f f u s e knapweed i s more common i n d r i e r l o c a t i o n s . This study compares the success of knapweed establishment from seed i n two ha b i t a t s , one hotter and d r i e r than the other. Also, i n order to compare the e f f e c t s of weather on knapweed establishment, seed sowing t r i a l s on one s i t e were made at two d i f f e r e n t times during the 1976 growing season. 2.Method: Diffuse and spotted knapweed seeds were sown at the A g r i c u l t u r e Canada Re-search Station i n Kamloops, and approximately 70 km from Kamloops, i n West-wold. The Kamloops s i t e has a hotter and d r i e r climate than the Westwold s i t e (Table XXX). Diffuse knapweed i s more prevalent i n the Kamloops area while spotted knapweed i s most common i n Westwold. No knapweed was found on the Kamloops s i t e , but spotted knapweed was found i n sparse patches a few meters from the Westwold s i t e . Other species commonly present on the Kam-loops s i t e were S a l s o l a k a l i and Artemesia t r i d e n t a t a , whereas i n Westwold, Amelanchier a l n i f o l i a , Shepherdia canadensis, Pseudotsuga men z i e s i i , Pinus  contorta l a t i f o l i a , Poa pratensis and Calamagrostis rubescens were common. The Kamloops s i t e was located i n an open f i e l d , while the Westwold s i t e was i n a c l e a r i n g i n the midst of regrowth from a previously logged f o r e s t . Table XXX. Comparison of Kamloops and Westwold seed sowing s i t e s S i t e * t o t a l annual p r e c i p i t a t i o n (mm) * x d a i l y temperature (°C) * x No. of days with f r o s t A l t i t u d e (m) Kamloops 260 8.3 130 350 Westwold 316 6.1 192 646 Climate of B r i t i s h Columbia, standard 1941 - 1970 average 172 Both s i t e s were mechanically worked to a depth of about 10 cm and then raked smooth. Plots were established i n grids 75 cm apart. P l o t s i z e was 25 X 25 cm. Grid s i z e and shape va r i e d between s i t e s . Borders of p l o t s were d e l i n -eated with baler twine, held i n place with fencing stapes. Sowing rates were chosen at random for each p l o t . Seeds for the Kamloops 1976 sowing had been c o l l e c t e d i n 1975 i n the Kamloops area during August, and stored during the winter at room temperature. For both 1976 f a l l sowings at Kamloops and at Westwold seeds were c o l l e c t e d i n the Pritchard ( d i f f u s e knapweed) and Westwold (spotted knapweed) areas, and stored at room temperature u n t i l they were used i n October. Seeds were sprinkled evenly over the surface of the s o i l and pressed f i r m l y into the ground because Watson (1972) found that germination was best on the s o i l surface. P l o t s were not watered. The Kamloops s i t e was seeded twice, i n A p r i l and October of 1976. In the spring of 1976 sowing rates were 0, 78, 156, 625 and 1250 d i f f u s e knapweed seeds per p l o t , and 0, 39, 156 and 625 spotted knapweed seeds per plo t (be-cause there were not enough spotted knapweed seeds f or other d e n s i t i e s ) . In the f a l l of 1976 seeds were sown at rates of 0, 69, 104, 625, 1875 and 5625 seeds per plo t for both d i f f u s e and spotted knapweed. These rates corres-ponded to r a t i o s of 0, 9:1, 6:1, 3:1, 1:1, 1:3 and 1:6 square centimeters of space to the number of seeds sown. Only one sowing took place i n Westwold. This was i n the f a l l of 1976, at the same time that seeds were sown i n Kamloops. Sowing rates i n Westwold were the same as the Kamloops f a l l rates. Each sowing rate, for a l l locations and times, was r e p l i c a t e d f i v e times. Counts of knapweed were made i n the spring and f a l l of 1977, when the plants were c l a s s i f i e d according to t h e i r s i z e , as small seedlings, 173 large seedlings or rosettes (small rosettes were included i n the rosette class) as i n the previous section. Knapweed found outside the pl o t s was not counted. In September, 1977, counts of seed heads per p l o t , and seeds per head were made for knapweed which had bolted i n the spring 1976 t r i a l i n Kamloops. 3.Results: a)Kamloops, spring 1976 t r i a l : By the spring of 1977, one year a f t e r the seeds were sown, a l l surviving d i f f u s e and spotted knapweed plants were i n the ro s e t t e stage. Mean numbers of d i f f u s e and spotted knapweed rosettes increased l i n e a r l y , and at the same rate, with sowing rate (Fig. 45). When the plants were counted i n May of 1977, the rosettes had already started to b o l t . B o l t i n g occurs r e l a t i v e l y synchronously i n natural populations, therefore most rosettes which were des-tined to bolt would probably have started to do so i n May. The average num-ber of b o l t i n g rosettes per plo t for either knapweed species did not d i f f e r s i g n i f i c a n t l y with sowing rate. On the average 0.7 - .3 and 0.7 - .2 d i f f u s e and spotted knapweed rosettes per p l o t , r e s p e c t i v e l y , were b o l t i n g i n the spring, and there was no s i g n i f i c a n t d i f f e r e n c e between these means. There was no s i g n i f i c a n t d i f f e r e n c e , i n r e l a t i o n to sowing rate, i n the percentage of rosettes which died during the 1977 growing season of the number e s t a b l i -shed i n the spring of 1977 for either d i f f u s e or spotted knapweed (Table XXXI). Only a few of the bolted rosettes produced seed i n the f a l l of 1977. The rest were arrested (did not produce any heads) or bolted unsuccessfully (pro-duced only aborted heads). Most of the d i s t a l heads on plants which 174 50-i -r 200 400 600 800 .1000 NUMBER OF SEEDS SOWN PER PLOT, SPRING 1977 1200 Figure 45. Average numbers of d i f f u s e and spotted knapweed rosettes per plot i n the spring of 1977. for the Kamloops, 1976 sowing t r i a l d i f f u s e knapweed-0 0; spotted knapweed X X produced mature seed heads, had dehisced before the plants were harvested. The proximal seed heads on these plants appeared smaller than the d i s t a l heads, and probably contained fewer seeds, as shown previously f o r a natural spotted knapweed population (section II B v i i i ) . There was also considerable var-i a t i o n i n average numbers of seeds per head among i n d i v i d u a l plants. There-fore numbers of seeds for each dehisced seed head were estimated on the basis of head s i z e , p o s i t i o n on the branch, and average numbers of seeds per head i n undehisced heads on the same plant, and the t o t a l seed y i e l d per p l o t was found. Although only s i x d i f f u s e , and two spotted knapweed plo t s produced seeds, i t appears that lowest seed y i e l d s were produced on plo t s with the 175 Table XXXI. Percentage dead d i f f u s e and spotted knapweed by the f a l l of 1977, of the number counted i n the spring of 1977, on the Kamloops, spring 1976 sowing t r i a l No. seeds sown per plot knapweed^" species 2 % dead from May to Sep-tember, 1977 39 d s 10 J 10% 67 - 33 78 d s 33 - 17 156 d s 13 J 13 51 - 23 313 d s 59 - 15 625 d s 53 J 18 7 9 - 8 1250 d s 68 - 14 d = d i f f u s e knapweed; s = spotted knapweed 2 no s i g n i f i c a n t d i f f e r e n c e with sowing rate f o r ei t h e r d i f f u s e or spotted knapweed highest rosette counts i n the spring of 1977 (Fig. 46). b)Kamloops, f a l l 1976 sowing: Seeds on t h i s s i t e probably germinated i n the f a l l of 1976 because by May of 1977 some plants had reached the rosette stage. Most plants, however, were i n the large seedling stage. Large seedling numbers increased l i n e a r l y for d i f f u s e and spotted knapweed with sowing rate. There was no s i g n i f i c a n t d i f f e r e n c e i n regression c o e f f i c i e n t s between species (Fig. 47), but i t appears that at higher sowing rates more spotted than d i f f u s e knapweed large seedlings were produced. In May average numbers of spotted knapweed rosettes per p l o t (8 - 2) did not. d i f f e r s i g n i f i c a n t l y with sowing rate, but average numbers of d i f f u s e knapweed rosettes per p l o t increased s i g n i f i c a n t l y with 176 cn < U_ O LU Q. CO Q LU LU co LL. O ac LU < h-o h-500-4001 °~ 300-\ 2 0 0 i 100 0 10 20 —T" 30 T " 40 1 50 TOTAL NUMBER OF ROSETTES PER PLOT, SPRING 1977 Figure 46. T o t a l number of d i f f u s e and spotted knapweed seeds produced per plo t by the f a l l of 1977 for the Kamloops, spring 1976 sowing t r i a l d i f f u s e knapweed 0 —0; spotted knapweed X- X sowing rate (Table XXXII). In the f a l l of 1977 there was no s i g n i f i c a n t d i f f e r e n c e i n mean numbers of d i f f u s e (0.2 - .1) or spotted (0.4 - .3) knap-weed rosettes per p l o t i n r e l a t i o n to sowing rate. Diffuse and spotted knap-weed mortality during the 1977 growing season was between 87% and 100% of the t o t a l number of plants (large seedlings + rosettes) present i n the spring of 1977, and there was no s i g n i f i c a n t d i f f e r e n c e i n m o r t a l i t y with sowing rate. c)Westwold, f a l l 1976 sowing: By May 1977 knapweed had developed only as f a r as the large seedling stage, i n d i c a t i n g that the seeds had germinated i n the spring of 1977. Numbers of 177 200, 1000 2000 3000 4000 5000 NUMBER OF SEEDS SOWN PER PLOT, FALL 1976 6000 Figure 47. Average numbers of d i f f u s e and spotted knapweed large seedlings per p l o t i n the spring of 1977 for the Kamloops, f a l l 1976 sowing t r i a l d i f f u s e knapweed 0 0; spotted knapweed X -X d i f f u s e and spotted knapweed large seedlings increased l i n e a r l y , and at the same rate, with the sowing rate (Fig. 48), but d i f f u s e knapweed appeared to do better at higher seed d e n s i t i e s . In the f a l l of 1977 d i f f u s e and spotted knapweed were found i n small seedling, large seedling and rosette stages. The small seedlings had germinated i n the f a l l because they s t i l l had cot-yledons attached. I t was assumed that any large seedling or'.rosette was a survivor of the spring cohort. In the f a l l there was no s i g n i f i c a n t d i f f e r -ence between mean numbers of d i f f u s e or spotted knapweed large seedlings + rosettes per pl o t i n r e l a t i o n to sowing rate, nor was there a s i g n i f i c a n t d i f f e r e n c e between mean numbers of d i f f u s e (7 - 1) or spotted (8 - 2) knap-178 Table XXXII. Comparison of d i f f u s e and spotted knapweed rosette estalishment during the 1977 growing season for the Kamloops, f a l l 1976 sowing t r i a l No. seeds sown per p l o t , f a l l , 1976 knapweed"*" species x number of rosettes per p l o t % dead rosettes during 1977 of number i n -spring spring, 1977 f a l l , 1977 69 d 2.4 + + •£ 0 + + °^ 100 + + s 4.4 .9 .4 .2 87 9 + + + 104 d 1.6 + .6 0 + 0 100 + 0 s 4 1 0 0 100 0 + .8 0 + + 208 d 1.4 + + 0 100 + 0 s 5.0 .5 0 0 100 0 + + + 625 d 5 + 2 0 + 0 100 + 0 s 10 3 2 2 91 8 + + 98 + 1875 d 19 + 6 .8 + .5 + 1 s 17 4 0 0 100 0 18 + + + 5625 d + 7 .4 + .4 99 + 1 s 9 — 7 0 0 100 0 "d = d i f f u s e knapweed; s = spotted knapweed s i g n i f i c a n t difference with sowing rate f o r d i f f u s e knapweed (p <.01) no s i g n i f i c a n t d i f f e r e n c e with sowing rate f o r d i f f u s e knapweed 'no s i g n i f i c a n t d i f f e r e n c e with sowing rate for spotted knapweed weed rosettes alone, i n r e l a t i o n to sowing rate (Table XXXIII). The per-centage of d i f f u s e knapweed large seedlings + rosettes which died during the 1977 growing season increased with sowing rate, but the percentage of spotted knapweed large seedlings + rosettes which died during 1977 did not vary s i g -n i f i c a n t l y with sowing rate, and the average death rate was 22 - 3%. The germination rate was defined as the number of large seedlings per p l o t / the number of seeds sown per pl o t X 100%. The germination rate i n May, 1977, did not d i f f e r s i g n i f i c a n t l y with sowing rate f o r d i f f u s e knapweed (Fig. 49), but decreased s i g n i f i c a n t l y with sowing rate for spotted knapweed (Fig. 50). In the f a l l of 1977 the. germination rate, defined as the number of small '* seedlings per plot/the number of seeds sown per p l o t X 100%, decreased s i g -179 o _ i Q_ 0£ LU D_ to _1 cn O r-l LU LU to tJ2 4ooH ^ 300H LU i—i j§ QL 2001 <r to LL. o LU CD IOOH 1 1000 2000 3000 4000 . : 5000 NUMBER OF SEEDS SOWN PER PLOT, FALL 1.9.76 6000 Figure 48/ '• Average- numbers-of 'diffuse and spotted "Knapweed large seedlings per p l o t i n the-spring of 1977 for the Westwold, f a l l 1976 sowing t r i a l d i f f u s e knapweed 0 0; spotted knapweed X-•——X n i f i c a n t l y with increasing sowing rate for both d i f f u s e and spotted knapweed. 4.Discussion: During the f i r s t year of growth d i f f u s e and spotted knapweed established equally well on experimental s i t e s i n Kamloops and i n Westwold. In f a c t , at young stages of development, spotted knapweed appeared to do better i n Kamloops, where d i f f u s e knapweed was most common, and d i f f u s e knapweed appeared to do better i n Westwold, where spotted knapweed was more common. Mixed populations of d i f f u s e and spotted knapweed i n B r i t i s h Columbia are rare, so that strong i n t e r s p e c i f i c competition probably e x i s t s at older stages of development, and/or s l i g h t changes i n habitat and climate gives 180 Table XXXIII. Average numbers of d i f f u s e and spotted knapweed plants produced by the f a l l of 1977 on the West-wold, f a l l 1976 sowing t r i a l , and the death rates during the 1977 growing season No. of seeds sown per p l o t knapweed species number per p l o t , f a l l 1977 large seedlings + rosettes rosettes % dead during : the\1977 •' growing season 69 104 208 625 1875 5625 d s d s d s d s d s d s 6 ± 2 1 3 - 5 17 J 3 1 1 - 3 21 J 7 15 - 3 64 J 10 3 1 - 7 65 ^ 9 76 - 11 130 J 45 107 - 13 4 ± 2 5 - 2 11 J 2 7 - 2 6 ± 3 8 - 3 15 ± 5 1 1 - 5 6 - 3 1 2 - 8 8 - 6 % 1 4 - 4 1 4 - 4 31 ± 12 24 - 14 16 J 7 2 7 - 9 50 J 4 2 9 - 3 71 t 8 2 5 - 5 1 d = d i f f u s e knapweed; s = spotted knapweed s i g n i f i c a n t d i f f e r e n c e with sowing rate f o r d i f f u s e knapweed; no s i g n i f i c a n t d i f f e r e n c e with sowing rate f o r a l l other cases one species an advantage over the other. Consequently, i f these populations were monitored for several more years, one species may dispace the other. Since the temperature ranges f or d i f f u s e (13°C to 28°C) and spotted (10°C to 28°C) knapweed seed germination are so s i m i l a r (Watson, 1972), seedling es-tablishment i s probably equally successful f o r both species i n a wide v a r i e t y of habitats. Spotted knapweed p e r s i s t s i n the hot, dry climate of Walachin, and d i f f u s e knapweed grows w e l l i n the cool, wet climate of Boston Bar, possibly because t h e i r respective counterparts have not dispersed to these s i t e s yet. Seedling establishment and successive growth of some plants may depend on whether the seedlings germinated i n the f a l l or i n the spring (Harper and i Figure 49. T T 1000 2000 3000 4000 NUMBER OF SEEDS SOWN PER PLOT, FALL 1976 5000 6000 Percentage d i f f u s e knapweed appeared of the number of seeds sown on the Westwold, f a l l 1976 sowing t r i a l large seedlings, spring 1977 0 0; small seedlings, f a l l 1977 X X % 4 0 l 7 30' 20 10 4 " ! 1000 2000 I 3000 4000 5000 6000 Figure 50. NUMBER OF SEEDS SOWN PER PLOT, FALL 1976 Percentage spotted knapweed appeared of the number of seeds sown on the Westwold, f a l l 1976 t r i a l large seedlings, spring 1977 0 0 small seedlings, f a l l 1977 X X 182 White, 1974). Those which germinated i n the f a l l would be older, and better established the following summer, than those which germinated i n the spring. This phenomenon would have affected the s u r v i v a l of knapweed which was pro-duced from seed sown i n the spring of 1976 compared with f a l l of 1976 i n Kamloops. The f a l l cohort should have the advantage over the spring cohort, but the opposite to the expected occurred: the spring cohort survived the 1976 growing season, but the f a l l cohort died out during the 1977 growing season. Weather caused t h i s r e v e r s a l . The 1976 growing season was wetter and colder (the mean temperature from A p r i l u n t i l September was 15°C and the t o t a l r a i n f a l l was 221.23 mm) than the 1977 growing season (the mean temp-erature from A p r i l u n t i l September was 16.4°C and the t o t a l r a i n f a l l was 139.50 mm; Table XXXIV). The weather i n 1977 was s i m i l a r to the th i r t y - y e a r average from 1941 to 1970, for which the mean temperature from A p r i l u n t i l September was 16.3°C and the t o t a l r a i n f a l l was 140.97 mm, but the 1977 growing season was exceptionally wet. Therefore seedlings, which germinated i n the f a l l of 1976, grew into large healthy rosettes, some of which even survived the t y p i c a l l y hot, dry summer of 1977. But, although some plants from the f a l l , 1976 cohort had reached the rosette stage by the following spring, they did not survive the hot 1977 summer. Therefore knapweed may be-come established i n Kamloops only during exceptionally good growing condi-t i o n s . Once established, however, i t could withstand normal years. The Westwold 1976 f a l l cohort survived the 1977 growing season even though the Kamloops 1976 f a l l cohort did not. In Westwold the t o t a l r a i n f a l l from A p r i l u n t i l September i n 1977 (139.31 mm) was s i m i l a r to Kamloops (139.50 mm), but the mean temperature was cooler i n Westwold (12.8°C) than i n Kamloops (16.4°C). Even though the seeds i n Westwold probably did not germinate u n t i l 183 Table XXXIV.:. 1976 and 1977 weather data for Kamloops and Westwold Kamloops Westwold x temper- t o t a l p r e c i - x temper- : ' t o t a l p r e c i -month ature (°C) p i t a t i o n (mm) ature (°C) l p i t a t i o n (mm) 1976 1977 1976 1977 1976 1977 1976 1977 January -2.78 -.41 19.56 20.0 -3.33 -6.13 19.55 40.39 February .56 2.50 10.16 5.4 -1.11 -.02 10.16 30.73 March 2.22 4.70 18.80 11.7 0.00 1.96 18.80 28.70 A p r i l 9.45 10.50 3.81 14.5 5.56 7.11 13.21 13.21 May 13.89 13.50 21.08 22.6 10.56 9.72 31.50 25.15 June 16.11 19.30 33.02 16.5 12.22 15.56 44.70 25.15 July 19.45 19.90 32.26 24.6 16.11 16.33 45.97 31.85 August 17.78 21.90 122.17 34.8 15.0 17.89 110.74 22.61 September 15.56 13.50 8.89 26.5 12.78 10.45 4.75 21.34-October 8.33 8.30 20.07 8.3 5.56 5.43 25.15 21.34 November 2.78 0.30 14.22 50.1 0.56 -1.67 24.64 50.04 December 0.56 -6.60 38.10 63.3 -.56 -8.22 31.24 74.68 ft Climate of B r i t i s h Columbia, published by the B r i t i s h Columbia Mi n i s t r y of A g r i c u l t u r e the following spring, placing the seedlings at a disadvantage to those which appeared i n Kamloops i n the f a l l of 1976, the cooler summer temperatures and more runoff from the greater snow cover i n Westwold probably resulted 'in .... moister s o i l conditions i n Westwold than i n Kamloops, with better knapweed s u r v i v a l i n Westwold. E a r l i e r , i n a study of population regulation of natural knapweed i n f e s t a t i o n s (section III'A), i t was shown that density-dependent competition acted as a regulatory factor only at the small seedling stage. Here, density-dependent m o r t a l i t y f o r d i f f u s e knapweed appeared to occur both at the large seedling and rosette stages. In Westwold, the percentage of d i f f u s e knapweed large seedlings which died i n 1977 increased s i g n i f i c a n t l y with sowing rate, and with the number of large seedlings established i n the spring of 1977. In Kamloops the percentage of d i f f u s e knapweed rosettes which died during the 1976 growing season was' lower for the sparser sowing rates (10%, 33% and 13% 184 for 39, 78 and 156 seeds per p l o t , respectively) than for the denser sowing rates (59%, 53% and 68% f o r 313, 625 and 1250 seeds per p l o t , r e s p e c t i v e l y ) . There was no s i g n i f i c a n t d i f f e r e n c e i n the percentage of d i f f u s e knapweed ro-settes which died with increasing sowing rate, but the above r e s u l t s do show a trend. Spotted knapweed did not appear to be regulated by density-dependent competition-e i t h e r i n Westwold or i n Kamloops at the large seedling and rosette stages; density-dependent regulation probably occurs only at the small seedling stage. Other studies of plant population dynamics have shown that density-dependent population regulation occurs at the seedling stage. Harper and McNaughton (1962) found that m o r t a l i t y was greater on p l o t s where Papaver spp. (poppies) were sown at denser rates. S i m i l a r i l y Marshall and J a i n (1968) obtained an increase i n death rate with an increasing sowing rate for Avena spp. (oats). An exception to these studies i s Harper and Gajic's (1961) study of a weed of cereal crops, Agrostemma githago (corn cockle), where no change i n plant mortality was observed with sowing rate. In t h i s study i t appears that seed y i e l d per unit area decreases with i n -creasing rosette density. In other s i m i l a r studies, reproductive structures, such as capsules or pods which contain seeds, usually decrease i n number per i n d i v i d u a l plant as plant density increases. For example, capsules of Agrostemma githago (Harper and G a j i c , 1961) and of Papaver spp. (Harper and McNaughton, 1962) and pods of V i c i a faba (beans'; Hodgson and Blackman, 1956) decreased per plant as plant density increased. Numbers of seeds per capsule also decreased f o r the above study with Papaver spp. I f more work were done with knapweeds a decrease i n the number of seed heads per plant and per p l o t might be found. This r e l a t i o n s h i p did not occur here because few rosettes 185 bolted a f t e r only one season i n the vegetative stage, and sample sizes were small. Diffuse and spotted knapweed populations are regulated, to some degree, by con t r o l l e d germination. The decrease i n germination rate for seedlings which appeared by the spring of 1977 i n Westwold was not s i g n i f i c a n t probably be-cause of the large standard errors of the means, but there was a decreasing trend i n the percentage of seeds which germinated with increasing sowing rate. Palmblad (1968) found that c o n t r o l l e d germination was exhibited by Bromus  imermis, Conyza canadensis, Plantago lanceolata and Silene an g l i c a . In the f a l l , on the Westwold s i t e , the decrease i n the germination rate with i n -creasing sowing rate may also have been influenced by numbers of plants a l -ready established on the p l o t s . Although the diffe r e n c e among means was not s i g n i f i c a n t , numbers of large seedlings + rosettes per p l o t appeared to increase with sowing-rate, and seedlings may not have germinated on the denser p l o t s due to lack of space. Very low percentages of seed germinated at any one time - usually l e s s than 30% of the o r i g i n a l number of seeds sown. The remaining seeds probably were blown away, buried, eaten by i n s e c t s , r o-dents or b i r d s , or washed out of the p l o t s . The p r o b a b i l i t y of rosette establishment and the subsequent b o l t i n g of the rosettes i s independent of the number of seeds sown. Therefore, even i f g a l l f l i e s destroyed 80% of the seeds, as found on the release s i t e s i n 1977 (section II B ix) the chance of knapweed establishment on a new s i t e would depend on weather conditions and habitat rather than the number of seeds deposited on the s i t e . I t i s probable that many of the knapweed i n f e s t a t i o n s i n B r i t i s h Columbia started from only a few seeds because knapweed was f i r s t 186 introduced into the province as a contaminant of a l f a l f a seed. 5.Conclusions: 1. D i f f u s e and spotted knapweed rosette establishment i s independent of numbers of seeds sown, but seedling and rosette s u r v i v a l are dependent on s i t e and weather f a c t o r s . 2. When growing conditions are favourable d i f f u s e knapweed i s regulated by density-dependent mortality at the large seedling and rosette stages, but spotted knapweed does not appear to be regulated at these stages. 3. Both d i f f u s e and spotted knapweed may be regulated by co n t r o l l e d germin-ation. 4. Knapweed w i l l e s t a b l i s h i n hot, dry lo c a t i o n s , such as Kamloops, only i n cool, wet summers. 187 IV CONCLUDING DISCUSSION A. INTRODUCTION U_. a f f i n i s and IJ. quadrif a s c i a t a established e a s i l y , and reached an apparent peak i n de n s i t i e s i n 5 to 7 years a f t e r t h e i r release. The e f f e c t of these f l i e s on seed production on d i f f u s e and spotted knapweed was spectacular at peak population d e n s i t i e s , i n 1977: 80% of the expected v i a b l e seed y i e l d was destroyed on both the d i f f u s e and spotted knapweed release s i t e s . The appearance of the plants was s t r i k i n g l y a l t e r e d , with d i s t a l heads aborting and proximal heads developing to the seed stage. There was a 33% decrease i n average spotted knapweed seed weight and a 30% decrease i n d i f f u s e knapweed seed v i a b i l i t y . However, these e f f e c t s have not been severe enough to control knapweed. Like most b i o l o g i c a l control systems the knapweed/gall f l y system i s a very simple system compared with natural phytophagous insect communities and t h e i r host plants. The only source of food for g a l l - f l y larvae i s ovariole t i s s u e i n knapweed heads, and the larvae and adults have no : s p e c i f i c enemies. Knapweed i s not s e r i o u s l y thinned by other animals unless i t i s grazed. But the f a i l u r e of the f l i e s to control d i f f u s e and spotted knapweed can be a t t r i b u t e d not only to inherent properties of the g a l l f l i e s and the knapweed, but also to the nature of the responses of the -plants to attack. The system i s further complicated by the presence of two g a l l - f l y species, each attacking heads at d i f f e r e n t stages of maturity; a f a c u l t a t i v e second f l y generation; and heterogeneity i n head sizes among knapweed such that plants 183 of one species and i n one population react d i f f e r e n t l y to the same rate of g a l l - f l y attack. IJ. a f f i n i s populations appear to be regulated by density-dependent competition for food and space i n the f i r s t - i n s t a r , p r e - g a l l stage, as was predicted for IJ. jaceana i f predator pressure were to be l i f t e d (Varley, 1947). Super-numerary U. a f f i n i s larvae i n d i f f u s e and spotted knapweed heads causes the head to abort. Evidence of t h i s phenomenon are the 4 mm long heads, with hard, dark-grey outer s h e l l s observed i n the laboratory a f t e r o v i p o s i t i o n by U_. a f f i n i s . Some of these heads contained s i n g l e t h i r d - i n s t a r U. a f f i n i s larvae, but without the t y p i c a l fusiform g a l l s . About 57, of the heads iden-t i f i e d as superparasitized were of eit h e r the above type. Some larger super-p a r a s i t i z e d heads had grown to the flowering point, but could not flower be-cause the f l o r e t s had been replaced by m u l t i l o c u l a r g a l l complexes, contain^:r ing dead larvae at a l l i n s t a r s and dead adults which f a i l e d to emerge from the woody g a l l s . Most superparasitized heads (95%) were distinguishable from undeveloped heads only on the basis of t h e i r d i s t a l p o s i t i o n s on branches, where heads normally develop to the seed stage. This study did not conclusively e s t a b l i s h that d i s t a l head abortion was caused by sup e r p a r a s i t i z a t i o n by U. a f f i n i s . Probing by females before o v i p o s i t i o n may have injured the young heads, or ." the plant may'have detected probing and aborted the head (Shorthouse, pers. comm.). This i s substantiated by the fac t that d i s t a l head abortion occurs more frequently on "small-headed" than on "large-headed" d i f f u s e knapweed plants, and i t stands to reason that small heads are more e a s i l y injured than large heads, given both head types are at the same stage of development. 189 But, given equal numbers of eggs l a i d i n small and large heads, the r e s u l t i n g larvae have les s room and food i n the small heads. Small heads containing many eggs or f i r s t - i n s t a r s could abort as a r e s u l t of a "stimulation that the head could not t o l e r a t e " (Shorthouse, pers. comm.). Sometimes undeve-loped heads cannot be distinguished from heads known to be superparasitized because laboratory work, where spotted knapweed was caged with U. a f f i n i s , showed that even larger heads (4 to 6 mm long) could abort when exposed to the f l i e s . A l l of these heads shrunk to approximately 3 to 4 mm long, and resembled undeveloped heads. Injury due to probing by females probably did not 'cause these heads to abort because they were older and larger; too many eggs or larvae per head seems to be the reason for head abortion. TJ. a f f i n i s (ZwBlfer, 1970) and U. jaceana (Varley, 1947) both commonly ov i p o s i t f a r greater numbers of eggs per head i n the laboratory than there i s room and food f o r l a r v a l development. Females probably cannot determine numbers of eggs already l a i d i n heads when they probe, and many females probably o v i -p o s i t into the same head. Furthermore, i t seems a poor s u r v i v a l t a c t i c on the part of TJ. a f f i n i s to destroy future o v i p o s i t i o n s i t e s simply by t e s t i n g the heads for s u i t a b i l i t y f o r egg deposition. Possibly, some of the f l o r e t s may be damaged, as are flower heads of Sonchus arvensis when attacked by T e p h r i t i s d i l a c e r a t a (Shorthouse, 1980). In t h i s thesis i t i s assumed that d i s t a l heads which abort have been superparasitized by U. a f f i n i s . TJ. q u a d r i f a s c i a t a numbers appear to be regulated by many factors even with-out i t s natural enemies. This species never reaches high d e n s i t i e s , but p e r s i s t s i n sp i t e of severe competition by such g a l l f l i e s as TJ. a f f i n i s and U. jaceana (Varley, 1947) and other'head-feeding i n s e c t s . When JJ. a f f i n i s d e n s i t i e s are greater than 0.5 f i r s t - i n s t a r larvae per head, U. quadri-f a s c i a t a i s suppressed because U. a f f i n i s , which o v i p o s i t s i n younger 190 heads than U. q u a d r i f a s c i a t a , saturates o v i p o s i t i o n s i t e s w i t h l a r g e numbers of woody g a l l s before the heads are mature enough f o r U. q u a d r i f a s c i a t a to u t i l i z e . At very h i g h U. a f f i n i s d e n s i t i e s many of the heads i n i t i a t e d during the f i r s t f l y generation abort and o v i p o s i t i o n s i t e s f o r U. q u a d r i -f a s c i a t a are f u r t h e r r e s t r i c t e d . But U. q u a d r i f a s c i a t a f i r s t - g e n e r a t i o n progeny s u r v i v e i n large-headed p l a n t s , and emerge i n August to c a p i t a l i z e on the many heads i n i t i a t e d by small-headed p l a n t s l a t e i n the growing season. However, when f i r s t - g e n e r a t i o n U. a f f i n i s d e n s i t i e s are low and most heads attacked i n June and J u l y develop to the seed stage, few new heads are i n i t i a t e d i n August, e s p e c i a l l y on spotted knapweed. I f hot, dry weather i n August i s combined w i t h the above s i t u a t i o n , second generations of U. q u a d r i f a s c i a t a may leave very few diapausing l a r v a e . But populations of U. q u a d r i f a s c i a t a can increase r a p i d l y from only a few a d u l t s emerging i n the s p r i n g . The e a r l i e r emergence of U. q u a d r i f a s c i a t a i n the s p r i n g may a l s o be a s u r v i v a l t a c t i c by t h i s s p e c i e s , although i t i s at a c e r t a i n disadvantage because few heads are a v a i l a b l e at that time, and those that • have appeared are very young. These early-emerging f l i e s may d i s p e r s e ; U. q u a d r i f a s c i a t a was found more f r e q u e n t l y i n new l o c a t i o n s than U. a f f i n i s , ( and p r e l i m i n a r y work on f l y a c t i v i t y showed U. q u a d r i f a s c i a t a to be more a c t i v e than U_. a f f i n i s . B. REASONS FOR THE FAILURE OF THE GALL FLIES TO CONTROL KNAPWEED i . R e s i l i e n c e of knapweed: Knapweed heads are i n d i v i d u a l u n i t s i n the sense that the amount of net o v a r i o l e t i s s u e i n each head appears to l i m i t the number of g a l l s that can form. Heads which are s u p e r p a r a s i t i z e d do not y i e l d seed, but the e n t i r e p l a n t does not d i e . When the r a t e of s u p e r p a r a s i t i z a t i o n during the f i r s t 191 f l y generation i s high, the plant responds by increasing the rate of proximal head i n i t i a t i o n l a t e r i n the growing season. But, i n August TJ. a f f i n i s appears to be less e f f i c i e n t i n attacking heads, and U. quadrifasciata cannot use a l l the ovaries that U. a f f i n i s misses. Therefore seed y i e l d can be high i n these late-appearing, fly-induced heads. F l y numbers are r e s t r i c t e d by supe r p a r a s i t i z a t i o n during the f i r s t generation, but the seed density may s t i l l be 20% of the expected y i e l d . This amounted to 800 and 500 v i a b l e seeds per 50 X 50 cm on the d i f f u s e and spotted knapweed release s i t e s , r e s p e c t i v e l y , i n 1977. From l i f e tables, these numbers of seeds can produce 74 d i f f u s e and 5 spotted knapweed mature, bolted plants. Heterogeneity i n head s i z e among plants occurs i n both d i f f u s e and spotted knapweed populations, but i t i s most s t r i k i n g i n d i f f u s e knapweed. Hetero-geneity i n head s i z e accounted f o r both r e s t r i c t i o n of U. a f f i n i s d e n s i t i e s , and high seed y i e l d l a t e i n the growing season. More than 95% of the heads on small-headed d i f f u s e knapweed aborted during the f i r s t f l y generation, whereas only 53% aborted on large-headed plants. U. a f f i n i s , U. quadrifas-, c i a t a , and some seed were produced by the large-headed plants during t h i s i n t e r v a l , but fewer new heads were i n i t i a t e d by these plants when the second generation emerged and only 61% of these late-appearing heads developed to the seed stage. In contrast, as many heads were i n i t i a t e d i n August and September by small-headed plants as during June and July, and 73% of these developed to the seed stage. i i .Phenology of TJ. a f f i n i s and TJ. qu a d r i f a s c i a t a: Although U. qua d r i f a s c i a t a emerges 5 to 7 days e a r l i e r than U. a f f i n i s i n June, the second generations of both species appear to be synchronized. 192 There i s an i n t e r v a l of one week during early August when f i r s t f l y gen-erations have a l l but died o f f , and the second generation has j u s t started to emerge. The r a t i o s of f l i e s to heads at s u i t a b l e stages for o v i p o s i t i o n are so low at t h i s time that many heads escape attack e n t i r e l y , or are only l i g h t l y attacked. i i i . Longevity and aggressiveness of vegetative stages: Diffuse and spotted knapweed have been c l a s s i f i e d as b i e n n i e l s or t r i e n n i e l s (Watson and Renney, 1974; Anonymous, 1962), but t h i s study showed that rangeland d i f f u s e and spotted knapweed can remain i n the rosette stage for two or more years before b o l t i n g . Diffuse knapweed may spend only the l a s t four months out of i t s three- to four-year l i f e span reproducing sexually. Bolted spotted knapweed spends les s time then d i f f u s e knapweed producing heads, and spotted knapweed can grow vegetatively between years of sexual reproduction. The g a l l f l i e s attack only when the plants have bolted, and then only when they are i n the bud stage. Knapweed can become fi r m l y established i n new locations before the f l i e s can attack them. The f i r s t rosettes to bo l t usually produce a few heads, but these are only l i g h t l y attacked or missed because f l y populations are low. This ensures a supply of seed for coming years. Meanwhile knapweed rosettes become more vigorous, and as they develop, hundreds of heads w i l l be pro-duced before there are s i g n i f i c a n t numbers of f l i e s a v a i l a b l e , except by immigration. Storey (1976) also f e l t that "U. a f f i n i s by i t s e l f may not reduce adequately the density of spotted knapweed i n western Montana. This suggested because the type of c o n t r o l 193 effected by JJ. a f f i n i s (the reduction i n seed production) i s not a d i -rect c o n t r o l of spotted knapweed ( i . e . , i t does not control e x i s t i n g p l a n t s ) . Consequently, wlthough the number of seeds produced could be reduced, there would s t i l l be a r e l a t i v e l y large number of seeds being dispersed by the plants. Since spotted knapweed i s not an annual, any successful seedlings could be serving as population additives instead -of population replacements; thus, a spotted knapweed i n f e s t a t i o n could conceivably increase despite the presence of JJ. a f f i n i s . " The rosette stage of d i f f u s e knapweed i s the most aggressive and competitive stage of the l i f e c y cle. The c i r c u l a r arrangement of leaves shades the ground and discourages germination of other seedlings. A l l e l o p a t h i c compounds are produced not only by the growing leaves, but also by le a f L'L'zz l i t t e r (Hajak, pers. comm.). The large tap root uses most of the a v a i l a b l e s o i l moisture i n the v i c i n i t y of the plant, further discouraging growth of other species. Rosettes are very drought tolerant, even when small, and can lose a l l t h e i r leaves during hot, dry weather, and regrow new ones when conditions are favourable. Preliminary observations show that rosettes can be grazed down to the l a s t l e a f without k i l l i n g them. However, i t i s only a f t e r the competitive rosette stage the f l i e s attack, and then only when the bolted plants are i n the bud stage. Spotted knapweed i s s i m i l a r to d i f f u s e knapweed i n that the rosette stage i s very aggressive, but i t bolt s e a r l i e r i n l i f e . Probably spotted knapweed can take a chance and b o l t with a smaller c r i t i c a l rosette s i z e because i n the advent of poor growing conditions, s t a l k elongation can cease, or a l l heads may abort, with subsequent p e r e n n i a l i t y and seed production i n l a t e r years. In August head i n i t i a t i o n terminates unless d i s t a l heads are superparasitized, and vegetative basal growth of rosettes around maturing stalks commences. Again, the f l i e s attack only during the non-competitve stage of spotted knapweed. Harris (1973b) f e l t that " I t may ... be rewarding to 194 synchronize the attack by the b i o c o n t r o l agent with the period of rapid growth by competing vegetation. This should make maximum use of the compe-t i t i o n between plant species for l i g h t , water, or minerals i n suppressing the weed." Harris (1973b) c i t e d an example where the stem weevil, Ceutor-hynchus l u t t u r a F. attacks the t h i s t l e Cirsi'um arvense (L) Scop, in''early spring when grasses are most competitive with the t h i s t l e . In four years a f t e r releases were made, the t h i s t l e s declined to 3% of t h e i r former den-s i t y . The problem i s that a safe agent may not be a v a i l a b l e for importation against d i f f u s e and spotted knapweed which attacks with s u f f i c i e n t s everity i n the spring, when rosettes are growing most vigorously. i v . P l a s t i c reactions by knapweed to a decrease i n seed density: Given a decrease i n seedling density, there i s some i n d i c a t i o n that d i f f u s e knapweed rosettes increase i n s i z e , but the reduction i n seedling density must be great - probably over 95%. Thus a reduction i n seed density should eventually r e s u l t i n an increase i n average rosette s i z e i f the seed bank i s depleted and fewer seedlings survive to the rosette stage than the number of rosettes leaving the population due to death i n the vegetative stage, or death a f t e r maturation. But t h i s process would take years, probably a decade with a 95% to 100% reduction i n seed density every year. But, thinning of knapweed populations w i l l probably not decrease plant cover because larger rosettes produce larger mature plants. Furthermore, a l l e l o p a t h i c secretions may prevent the invasion of useful rangeland species even when plant cover.is reduced. Thinning of knapweed populations did not take place for a s u f f i n . c i e n t l y long period to determine the reaction by weeds. Another e f f e c t of' a l l e v i a t i n g i n t r a s p e c i f i c competition i n knapweed,: e s p e c i a l l y d i f f u s e knap-weed, could be a f a s t e r turnover rate of the population because les s time 195 would have to spent i n the rosette stage before c r i t i c a l s i z e i s reached, v. Knapweed establishment from seed: Plant establishment i n t h i s study occurred independently of seed density sown, down to the lowest seeding rate of 39 seeds per 50 X 50 cm. Harper (1960) f e l t that "The seed population bears no d i r e c t r e l a t i o n s h i p to the density of weed i n f e s t a t i o n i t w i l l produce; t h i s depends on the frequency with which i n d i v i d u a l seeds meet microsites s u i t a b l e f or establishment". Therefore, i f entire, knapweed plants are transported i n the undercarriage of a v e h i c l e , as commonly occurs (Strang, et a l . , 1979), the p r o b a b i l i t y that knapweed establishes w i l l be the same given that a l l seeds are deposited, or that g a l l f l i e s had already reduce the expected y i e l d by 80%. C. FUTURE ASPECTS OF KNAPWEED CONTROL "...the most e f f e c t i v e stage at which a weed population may be attacked i s at the early phases of i t s spread, not when i t has become an obvious i n f e s -t a t i o n . " (Harper, 1960). Although no recent surveys of the area occupied by knapweed have been undertaken ( a e r i a l photography programmes are ,~ scheduled for the near f u t u r e ) , d i f f u s e and spotted knapweed have spread to new areas since the l a s t survey i n 1972 made by Watson and Renney (1974). Diffuse knapweed i s already i n Alberta (Harris and Cranston, 1979) and i n the northwestern United States alone, d i f f u s e and spotted knapweed occupy some 1.5 m i l l i o n ha (Maddox, 1979). B i o l o g i c a l c o n t r o l i n Canada i s a . r e l a t i v e l y slow processs: on the average i t takes 20 years to achieve success for one weed (Harris, 1979). Therefore knapweed can conceivably spread to new l o c a -t i o n s , despite the e f f o r t s of spraying crews to contain t h i s pest u n t i l b i o -l o g i c a l c o n t r o l i s successful. 196 Usually f i v e b i o l o g i c a l control agents must be released, and four must be established to c o n t r o l one weed (Harris, 1979). However, because d i f f u s e and spotted knapweed are c l o s e l y r e l a t e d , the establishment of only s i x agents i n 28 years should bring about co n t r o l (Harris and Cranston, 1979). M u l t i p l e introductions against weeds must be made because, unlike insect pests which are attacked and k i l l e d by t h e i r enemies, weeds are usually only weakened by phytophagous insects u n t i l competition from other plants, secon-dary attack by pathogens, or lowered reproductive capacity u l t i m a t e l y reduces d e n s i t i e s (van den Bosch and Messenger, 1973). A moth, Metzaeria paucipunctella Z e l l e r (LepidopteratGelichiidae), whose larvae also cause seed reduction, but by d i r e c t feeding, was introduced i n 1973 on spotted knapweed, and although t h i s agent became established i n some loca t i o n s , i t has f a i l e d to become e f f e c t i v e . A fourth agent, Sphenoptera — j j u g o s l a v i c a Obenberger (Coleoptera:Buprestidae), s p e c i f i c to d i f f u s e knapweed, was introduced i n 1976, and 'survived on a s i t e near Penticton. This beetle requires a r i d habitats that arrest rosette growth i n mid-summer, because l e a f growth disloges eggs and larvae that develop between them. Galls are produced i n the roots, where development to the adult stage i s completed, but rosettes seldom die when attacked, and subsequent b o l t i n g and seed head pro-duction may not be prevented. (ZwBlfer, 1976: Myers, pers. comm.) New agents against d i f f u s e and spotted knapweed are currently being screened. Among them are the root-boring moths, ' Pseudocosma caecimaculana Dup., Agapete zoegana L., and Pterolonche inspersa Stg., and two r u s t s , Puccinia  centaureae DC. and P. jaceae Otth. (Harris and Cranston, 1979). Although U. a f f i n i s and U. q u a d r i f a s c i a t a have not, to date, c o n t r o l l e d 197 d i f f u s e and spotted knapweed, they may have weakened the plants to a degree where attack by a d d i t i o n a l agents w i l l r e s u l t i n complete c o n t r o l . The s i n k - e f f e c t , whereby f l y larvae i n g a l l s a t t r a c t nutrients which would have gone into the development of other heads, may also draw on reserves a l l o c a t e d to vegetative growth by spotted knapweed i n August, thus gradually decreasing perennial plant s i z e . Spotted knapweed seedling vigour may decrease as a r e s u l t of the 33% reduction i n average seed weight caused by the f l i e s , and th i s decrease i n seedling vigour may also have the long-term e f f e c t of r e -ducing average plant s i z e . The percentage of dormant seeds i n the s o i l may decrease as a r e s u l t of f l y attack because i t appears that seed-coat pro-p e r t i e s , e s p e c i a l l y for d i f f u s e knapweed, have been alt e r e d by the f l i e s . However, studies of seed-bank dynamics and seedling s u r v i v a l should be undertaken over long periods before i t can be determined that the g a l l f l i e s are contributing to the con t r o l of knapweed. The knapweed/gall-fly system i s a u s e f u l system to study from the point of insect population ecology because aspects of e c o l o g i c a l theory can be tested with few measurements. No other insects attack knapweed heads, and few other insects damage the re s t of the plants. Spider mites can destroy bolted plants . -• , when the plants are i n the bud stage, but t h i s i s rare. A l l developmental stages of the f l i e s can be sampled, and to date they have no s p e c i f i c enemies. Varley (1947) on the other hand, had to contend with 8 other phytophagous insects which attacked C. nemoralis (some of them were head-feeding insects) and 14 p a r a s i t i c and 1 predatory species of U. j aceana. He had to monitor'the s u r v i v a l and f e r t i l i t y of nine of the enemies of U. jaceana. I t i s from e c o l o g i c a l theory that new b i o l o g i c a l control agents are chosen, therefore i t stands to reason that as much should be learned from 198 these simple system i n f i e l d conditions as possible. D. RECOMMENDATIONS FOR THE BIOLOGICAL CONTROL OF WEEDS Early attempts at c l a s s i c a l b i o l o g i c a l c o n t r o l were based on the researcher's i n t u i t i o n , and thus c l a s s i c a l b i o l o g i c a l c o n t r o l was more of an ar t than a science (Harris, 1973a). More recently, recommendations f o r choosing agents (Watt, 1965; Turnbull, 1967; van den Bosch, 1968 and 1971; Levins, 1969; Frazer, 1972; and Ha r r i s , 1973a and 1971), the screening and releasing of agents (van den Bosch and Messenger, 1973; ZwBlfer and H a r r i s , 1971; Harris and ZwBlfer, 1968; and Beirne, 1975) and the evaluation of t h e i r e f f e c t i v e -ness (DeBach and Huffaker, 1971; and van den Bosch and Messenger, 1973) have been made that should make the process more s c i e n t i f i c . A s c i e n t i f i c processs, however, involves the p r e d i c t i o n of the best agent to contr o l the pest. To a ce r t a i n extent, non-promising agents can be shelved, e s p e c i a l l y where weedy pests are concerned, i f they attack non-vulnerable stages. But r e l i a b l e predictions of the performance of a new agent cannot be made yet. Laboratory studies indicated early i n the study of the knapweed/ g a l l - f l y stystem that the f l i e s can saturate o v i p o s i t i o n s i t e s , and superpara-s i t l z e heads. But, often insects disperse rather than deplete t h e i r food r e -source i n the f i e l d , and although Varley (1947) predicted that the g a l l f l y , TJ. jaceana, could be con t r o l l e d by density-dependent competition for food, other factors i n B r i t i s h .Columbia could have regulated U. a f f i n i s d e n s i t i e s . Even i f TJ. a f f i n i s proved to follow Varley's (1947) theory, as i t subse-quently did, the amount of seed reduction could not be f o r e t o l d i n terms of the more p r a c t i c a l , weed-control standpoint. Heterogeneity i n knapweed head s i z e , and the great e f f e c t t h i s f a c t o r has i n regulating U. a f f i n i s and U. qua d r i f a s c i a t a numbers, as well as i n the p o t e n t i a l seed y i e l d of knapweed 199 at peak f l y d e n s i t i e s , was not recognized u n t i l the f l i e s were reaching t h e i r population maxima. It remains to be seen, i f b i o l o g i c a l control w i l l ever become more of a science than an a r t . To take f u l l advantage of the unique opportunities for studying introduced agents, and the e f f e c t s these agents have on weeds, each agent should be introduced s i n g l y so that i t s i n d i v i d u a l e f f e c t s can be determined. Theo-r e t i c a l l y , wherever several species are introduced, the more e f f i c i e n t agents replace or suppress the less e f f i c i e n t ones, presumably with no reduction i n control (van den Bosch, 1968). But, i n a s i t u a t i o n where agents that attack d i f f e r e n t parts of a plant are released together the o v e r a l l r e s u l t might be counter-productive. For example, foliage-feeding insects may reduce seed weight, but flower-feeding insects may increase seed weight (Maun and Cavers, 1971a and 1971b). By studying simple systems involving one insect attacking one plant, the researcher may foresee the outcome i n a more complicated system. Harris (1973b) f e l t that gall-forming insects may be poor choices as b i o -l o g i c a l c o n t r o l agents because they have "evolved a homeostasis with t h e i r host that renders them incapable of i n f l i c t i n g serious damage to i t " . This study support's H a r r i s ' p r e d i c t i o n . However, control of the rangeland weed, Pamakani, by a t e p h r i t i d g a l l f l y , Eupatorium adenophorum, which forms g a l l s i n stems, was successful i n some parts of Maui (Bess and Haramoto, 1959), and Shorthouse (1980) c i t e d several exampls of gall-forming insects which warranted consideration as b i o l o g i c a l control agents. Harris (1973a) further predicted that flower- and bud-feeding insects would be of l i t t l e value for b i o l o g i c a l control and t h i s study supports h i s generalization. 200 The wide d i s t r i b u t i o n of b i o l o g i c a l c o n t r o l agents before t h e i r success can be demonstrated on a few t y p i c a l s i t e s , and t h e i r i n i t i a l d i s p e r s a l can be monitored i s a waste of money and may also f a l s e l y r a i s e the hopes of land-owners who may be desperate for a low-cost means of pest control.' The a b i l i t y of agents to search for t h e i r hosts i s an important a t t r i b u t e which should be evaluated. If one agent were d i s t r i b u t e d i n i t i a l l y to a l l . lo c a t i o n s , then e f f e c t s of i n d i v i d u a l agents that might be imported i n the future i s .'"difficult at best-, and'.of ten 1 impossible'to determine i f the ' . o r i g i n a l agent cannot be kept out of the study areas. There must be co n t r o l s i t e s where the ecology of the weed can be monitored without i t s enemies to e s t a b l i s h a baseline for estimating the e f f e c t s of insect attack. Caging a l t e r s the natural plant growth and the micro-ecology of the test s i t e . "The plant demographer should... i d e a l l y concern himself not only with the numbers of but also the s i z e d i s t r i b u t i o n of i n d i v i d u a l s within a population; he should d i s t i n g u i s h between genetic and p h y s i o l o g i c i n d i v i d u a l s ; he may have to make a r b i t r a r y d i s t i n c t i o n s about when a plant becomes a plant; and he should attempt to measure r e a l d e n s i t i e s within, an area rather than mean den s i t i e s over an area" (Harper, 1960). The importance of looking at the i n d i v i d u a l i n a plant population was"'shown i n t h i s study by the more frequent head superparasitism by U. a f f i n i s of "small" than "lar g e " heads, given both head types are at the same stage of maturity. Weeds may form monocultures, and although introduced b i o l o g i c a l c o n t r o l agents are usually monophagous, the environment can s t i l l be heterogeneous i n terms of the d i s t r i b u t i o n of preferred s i t e s of feeding or o v i p o s i t i o n . These s p e c i a l units can become saturated without weed' "control;-.-... . 201 Reduction of weedy pests by b i o l o g i c a l c o n t r o l agents i s more desirable than chemical treatment because p e s t i c i d e s pose dangers to human health, no matter how c a r e f u l l y they are used. The public has become s u f f i c i e n t l y educated about the dangers of p e s t i c i d e s that environmentalist groups are often successful i n stopping planned spraying programs, e.g., the spraying of Eurasian m i l f o i l i n Okanagen Lake, and the spruce budworm i n the Fraser Canyon. There are increasing demands for the supposedly safer and more e f f e c t i v e a l t e r n a t i v e , b i o l o g i c a l c o n t r o l . "But t h i s .method should involve careful'.follow-up.istudies to make i t t a r m o r e l s c i e h t i f i c process. 202 V LITERATURE CITED Amor, R. L. and R. V. H a r r i s . 1975. Seedling establishment and vegetative spread of Cirsium arvense (L) Scop, i n V i c t o r i a , A u s t r a l i a . Weed Research 15 (16): 407 - 411. Andres, L. A. and R. D. Goeden. 1971. The b i o l o g i c a l c o n t r o l of weeds by introduced natural enemies. B i o l o g i c a l c o n t r o l , C. B. Huffaker, e d i t o r . Plenum Press, New York. Anonymous. 1962. Nebraska Weeds. B u l l e t i n No. 101-R. Published by the Weed and Seed D i v i s i o n , Department of A g r i c u l t u r e and Inspection, State of Nebraska. Anonymous. 1976. Climate of B r i t i s h Columbia. Published by the B r i t i s h Columbia Min i s t r y of A g r i c u l t u r e , V i c t o r i a , B.C. Anotonovics, J . 1972. Population dynamics of the grass Anthoxanthum  odoratum on a zinc mine. Journal of Ecology 60: 35 - 365. Baskin, J . M. and C. M. Baskin. 1979. Studies on the autecology and pop-u l a t i o n biology of the weedy monocarpic perennial, Pastinaca s a t i v a . Journal of Ecology 67: 601 - 610. Benson, L. 1959. Plant C l a s s i f i c a t i o n . D. C. Heath and Company, Lex-ington Massachusetts. Beirne, B. P. 1975. B i o l o g i c a l control attempts by introductions against pest insects i n the f i e l d i n Canada. Canadian Entomologist 107: 225 -236. Bernard, J . M. 1976. The l i f e h i s t o r y and population dynamics of shoots of Carex r o s t r a t a . The Journal of Ecology 64(3): 1045 - 1048. Berube, D. E. 1980. I n t e r s p e c i f i c competition between Urophora a f f i n i s and U. q u a d r i f a s c i a t a (Diptera:Tephritidae) for o v i n o s i t i o n a l s i t e s on d i f f u s e knapweed (Centaurea diffusa:Compositae). Z. angew. Entomol. (in press). Bess, H. A. and F. H. Haramoto. 1959. B i o l o g i c a l control of Pamakani, Eupatorium adenophorum, i n Hawaii by a T e p h r i t i d g a l l f l y , Proce-cidochares u t i l i s . 2. Population studies of the weed, the f l y , and the parasites of the f l y . Ecology 40: 244 - 249. Birc h , L.C. 1971. The r o l e of environmental heterogeneity and genetical heterogeneity i n determining d i s t r i b u t i o n and abundance. i n Dynamics  of Populations, P. J . den Boer and G. R. Gradwell, e d i t o r s . Centre for A g r i c u l t u r a l Publishing and Documentation, Wegeningen. Brink, V. C. 1975. (personal communication) Professor of rangeland ecology, Department of Plant Science, U n i v e r s i t y of B r i t i s h Columbia. Buff, L. P. 1975. (personal communication) Rancher, Westwold, B.C. 203 Cameron, E. 1935. A study of the natural c o n t r o l of ragwort (Senecio  jacobaea L.) Journal of Ecology 23: 265 - 322.: Cavers, P. B. 1971 The e f f e c t s on reproduction of removal of plant parts by natural or a r t i f i c i a l means. Proc. 2nd International Symposium on B i o l o g i c a l Control of Weeds, Rome I t a l y . Daubenmire, R. F. 1959. Plants and the Environment. Johm Wiley and Sons, Inc., New York. DeBach, P. 1974. B i o l o g i c a l c o n t r o l by natural enemies. Cambridge Univ e r s i t y Press, London. DeBach, P. and C. B. Huffaker. 1971. Experimental techniques for evalu-ation of the effectiveness of natural enemies. B i o l o g i c a l Control. C. B. Huffaker, e d i t o r . Plenum Press, New York. Deevey, E. S. 1947. L i f e tables for natural populations of animals. Quarterly Review of Biology 22: 293 - 314. Dempster, J . P. 1971. Simple elementary models i n population dynamics. i n Dynamics of Populations, P. J . den Boer and G. R. Gradwell, e d i t o r s . Centre f o r A g r i c u l t u r a l Publishing and Documentation, Wegeningen. Dempster, J . P. 1975. Animal Population Ecology. Academic Press, New York. den Boer, P. J . 1971. S t a b i l i z a t i o n of animal numbers and the hetero-geneity of the environment: The problem of the persistence of sparse populations. i n Dynamics of Populations, P. J . den Boer and G. R. Gradwell, e d i t o r s . Centre for A g r i c u l t u r a l Publishing and Docu-mentation, Wegeningen. E l l i o t t , J . M. 1971. Some Methods for the S t a t i s t i c a l Analysis of samples of Benthic Invertebrates. Freshwater B i o l o g i c a l Association, S c i e n t i f i c p u b l i c a t i o n No. 25. Fletcher, R. A. and A. J . Renney. 1963. A growth i n h i b i t o r found i n Centaurea spp. Canadian Journal of Plant Science 43: 475 - 481. Frazer, B. C. 1972. Population dynamics of recognition of biotypes i n the pea aphid (Homoptera:Aphididae). Canadian Entomologist 104: 1729 -1733. Goodall, D. W. 1970. S t a t i s t i c a l plant ecology. Annual review of ecology and systematics 1: 99 - 124. Harper, J . L. 1960. Factors c o n t r o l l i n g plant numbers, i n The Biology of  Weeds. J . L. Harper, e d i t o r . Blackwell S c i e n t i f i c P u b l i c a t i o n s , Oxford. Harper, J. L. 1963. The nature and consequence of interference amongst plants. i n Genetics To-day. Proc. XI Int. Cong. Genetics, S. J . Geerts, e d i t o r . Symposium Pu b l i c a t i o n s , Pergamon Press, N. Y. 204 Harper, J . L. 1967. A Darwinian approach to plant ecology. Journal of Ecology 55: 247 - 270. Harper, J. L. and D. Oajic. 1961. Experimental studies of the mortality and p l a s t i c i t y of a weed. Weed Research 1: 91 - 104. Harper, J . L., P. H. L o v e l l , and K. G. Moore. 1970. The shapes and sizes of seeds. Annual Review of Ecology and Systematics 1: 327 - 356. Harper, J . L. and I. H. McNaughton. 1962. The comparative biology of c l o s e l y r e l a t e d species l i v i n g i n the same area. VII. Interference between i n d i v i d u a l s i n pure and mixed populations of Papaver species. New Phytologist 61: 175 - 188. Harper, J . L. and J . White. 1970. The dynamics of plant populations, i n Dynamics of Populations. P. J . den Boer and G. R. Gradwell, e d i t o r s . Centre for A g r i c u l t u r a l Publishing and Documentation, Wegeningen. Harper, L. L. and White. 1974. The demography of plants. Annual Review of Ecology and Systematics,5: 419 - 463. Harper, J . L., J . T. Williams and G. R. Sagar. 1965. The behavior of seeds i n s o i l . I.The heterogeneity of s o i l surfaces and i t s r o l e i n deter-mining the establishment of plants from seed. The Journal of Ecology 53: 273 - 286. H a r r i s , P. 1971. B i o l o g i c a l control of weeds. Environmental l e t t e r s 2(2): 75 - 88. H a r r i s , P. 1973a. Selection of e f f e c t i v e agents for the b i o l o g i c a l control of weeds. Canadian Entomologist 105: 1495 - 1503. H a r r i s , P. 1973b. Weed v u l n e r a b i l i t y to damage by b i o l o g i c a l control agents. Proceedings of the Second International Symposium on B i o l o g i c a l Control of Weeds, October, 1971. Misc. P u b l i c a t i o n No. 6, Commonwealth A g r i c u l t u r a l Bureaux, P. H. Dunn, ed i t o r . H a r r i s , P. 1979. Cost of b i o l o g i c a l control of weeds by insects i n Canada. Weed Science 27(2): 242 - 250. H a r r i s , P. 1980a. Establishment of Urophora a f f i n i s F r f l d . and U. quadrj-f a s c i a t a (Meig.) (Diptera:Tephritidae) i n Canada for the b i o l o g i c a l control of d i f f u s e and spotted knapweed. ( i n press) H a r r i s , P. 1980b. E f f e c t s of Urophora a f f i n i s F r f l d . and U. quadrifasciata Meig. (Diptera:Tephritidae) on Centaurea d i f f u s a Lam. and C. maculosa Lam. (Compositae). ( i n press) H a r r i s , P. and R. Cranston. 1979. An economic evaluation of control methods for d i f f u s e and spotted knapweed i n western Canada. Canadian Journal of Plant Science 59: 375 - 382. H a r r i s , P., D. Peschken and J . Milroy. 1969. The status of b i o l o g i c a l c o n t r o l of the weed Hypericum perforatum i n B r i t i s h Columbia. Canadian Entomologist 101(1): 1 - 1 5 . 205 H a r r i s , P. and H. ZwBlfer. 1968. Screening of phytophagous insects for b i o l o g i c a l control of weeds. Canadian Entpmolgist 100: 295 - 303. Hodgson, G. I. and G. E. Blackman. 1956. An analysis of the influence of plant density on the growth of V i c i a faba. I. The influence of plant density on the pattern of development. Journal of Experimental Botany 7: 1 4 7 - 1 6 5 . Huffaker, C. B., P. S. Messenger and P. DeBach. 1971. The natural enemy component i n natural control and the theory of b i o l o g i c a l c o n t r o l , i n B i o l o g i c a l Control. C. B. Huffaker, e d i t o r . Plenum Press, New York. Hutchings, M. J . and J . P. Barkham. 1976. An i n v e s t i g a t i o n of shoot i n t e r -actions i n M e r c u r i a l i s perennis L. a rhizomatous perennial herb. The Journal of Ecology 64(2): 723 - 743. Janzen, D. H. 1971. Seed predation by animals. Annual Review of Ecology and Systematics 2: 465 - 492. Krebs, C. J . 1972. Ecology: The Experimental Analysis of D i s t r i b u t i o n and  Abundance. Harper and Row, Publishers, Inc., New York. LeClerge, E. L., W. H. Leonard and A. G. Clark. 1962. F i e l d P l o t Technique. Burgess Publishing Company, Minneapolis. 2nd E d i t i o n . Levins, R. 1969. Some demographic and genetic consequences of environmental heterogeneity f o r b i o l o g i c a l c o n t r o l . B u l l e t i n of the Entomological Society of America 15: 273 - 240. Maddox, D. M. 1979. The knapweeds: t h e i r economics and b i o l o g i c a l control i n the Western States, U.S.A. Rangelands 1(4): 139 - 141. Majak, W. 1979. (personal communication) Research S c i e n t i s t , A g r i c u l t u r e Canada, Kamloop s, B.C. Marsden-Jones, E. M. and W. B. T u r r i l l . 1954. B r i t i s h Knapweeds. Royal Society No. 138, Bernard Quaritch Ltd., London. Marshall, D. R. and S. K. J a i n . 1968. Interference i n pure and mixed pop-ulat i o n s of Avena fatua and A. barbata. Journal of Ecology 57: 251 -270. Maun, M. A. and P. B. Cavers. 1971a. Seed production and dormancy i n Rumex. crispus. I. The e f f e c t s of removal of cauline leaves at anthesis. Canadian Journal of Botany 49(7): 1123 - 1130. Maun, M. A. and P. B. Cavers. 1971b. Seed production and dormancy i n Rumex  crispus. I I . The e f f e c t s of removal of various proportions of flowers at anthesis. Canadian Journal of Botany 49(10): 1841 - 1948. Milthorpe, F. L. 1961. The nature and analysis of competition between • plants of d i f f e r e n t species, i n Mechanisms i n B i o l o g i c a l Competition. F. L. Milthorpe, e d i t o r . Cambridge Uni v e r s i t y Press, London. 206 Myers, J . H. 1980. (personal communication). Professor of Ecology, Un i v e r s i t y of B r i t i s h Columbia. Noy-Meir, I. 1975. S t a b i l i t y of grazing systems: an a p p l i c a t i o n of predator-prey graphs. Journal of Ecology 63(2): 459 - 481. Pajunen, V. I. 1971. Adaptation of A r c t o c o r i s a carinata (Sahib.) and C a l l o c o r i x a producta (Reut.) populations to a rock pool environment, i n Dynamics of Populations. P. J . den Boer and G. R. Gradwell, e d i t o r s . Centre for A g i r c u l t u r a l Publishing and Documentation, Wegeningen. Palmblad, I. G. 1968. Competition i n experimental populations of weeds with emphasis on the regulation of population s i z e . Ecology 49: 26 -34. Piggin, C. M. 1976. Factors a f f e c t i n g seedling establishment and s u r v i v a l of Echium plantagineum L., T r i f o l i u m subterraneium L. and Lolium  rigidum Gaud. Weed Research 16(4): 267 - 272. Putwain, P. D., D. Machin and J . L. Harper. 1968. Studies i n the dynamics of plant populations. I I . Components and regulation of a natural population of Rumex a c e t o s e l l a . The Journal of Ecology 56: 421 - 231. Regehr, D. L. and F. A. Bazzaz. 1979. The population dynamics of Erigeron  canadensis, a successional winter annual. Journal of Ecology 67: 923 - 933, Robertson, J . 1975. (personal communication) Rancher, P r i t c h a r d , B.C. Ross, M. A. and J . L. Harper. 1972. Occupation of b i o l o g i c a l space during seedling establishment. The Journal of Ecology 60: 7 7 - 8 8 . Roze, L. D. and B. D. Frazer. 1979. B i o l o g i c a l c o n t r o l of d i f f u s e and spotted knapweed by Urophora a f f i n i s and U. quadrifasciata i n B r i t i s h Columbia. Proceedings of the F i r s t International Rangeland Congress. Denver, Colorado. Sagar, G. R. 1959. PhD Thesis, U n i v e r s i t y of Oxford. Sarukhan, J . 1974. Studies of plant demography: Ranunculus,? repens L., R. bulbosus L. and R. a c r i s L. I I . Reproductive strategies and seed population dynamics. The Journal of Ecology 62: 151 - 177. Sharitz, R. R. and J . F. McCormick, 1973. Population dynamics of two competing annual plant species. Ecology 54: 723 - 739. Shorthouse, J . 1980. (personal communication) Professor of biology, Laurentian U n i v e r s i t y , Sudbury, Ontario. Shorthouse, J . 1980. M o d i f i c a t i o n of the flower heads of Sonchus arvensis (Family Compositae) by the g a l l former T e p h r i t i s d i l a c e r a t a (Order Diptera, Family Tephritidae). Canadian Journal of Botany (in press). Southwood, T. R. E. 1971. E c o l o g i c a l Methods. Chapman and H a l l , Ltd., London. 207 Stevens, 0. A. 1932. The number and weight of seeds produced by weeds. American Journal of Botany 19: 284 - 794. Storey, J . M. 1976. A study of Urophora a f f i n i s (Diptera:Tephritidae) r e -leased on spotted knapweed i n western Montana. MSc Thesis, Montana State U n i v e r s i t y . Strang, R. M., K. M. Lindsay and R. S. P r i c e , 1979. Knapweeds: B r i t i s h Columbia's undesirable a l i e n s . Rangelands 1(4): 141 - 143. Tamm, C. 0. 1956. Further observations on the s u r v i v a l and flowering of some perennial herbs, I. Oikos 7: 272 - 292. Tanskiy, V. I. 1969. The harmfulness of the cotton bollworm, H e l i o t h i s  obsoleta F. (Lepidoptera, noctuidae) i n southern Tadzhikistan. Entomological Review 48: 2 3 - 2 9 . \ Thomas, A. G. and H. M. Dale. 1975. The r o l e of seed reproduction i n the ^ dynamics of established populations of Hieracium florbundum and a comparison with that of vegetative reproduction. The Canadian Journal of Botany 53: 3022 - 3031. Turnbull, A. L. 1967. Population dynamics of exotic i n s e c t s . B u l l e t i n of the Entomological Society of America 13: 333 - 337. van den Bosch, R. 1968. Comments on population dynamics of exotic i n s e c t s . B u l l e t i n of the Entomological Society of America 14: 112 - 115. van den Bosch, R. 1971. B i o l o g i c a l control of i n s e c t s . Annual Review of Ecology and Systematics 2: 4 5 - 6 6 . van den Bosch, R. and P. S. Messenger. 1973. B i o l o g i c a l Control. Intext Press, Inc., New York. Varley, G. C. 1947. The natural c o n t r o l of population balance i n the knap-weed g a l l f l y (Urophora j aceana). Journal of Animal Ecology 16: 139 -187. Wadsworth, J. T. 1914. Some observations on the l i f e - h i s t o r y and bionomics of the knapweed g a l l - f l y Urophora s o l s t i t i a l i s Linn. Annals of Applied Biology 1:" 142 - 169. Watkinson, A. R. and J . L. Harper. 1978. The demography of a sand dune annual: Vulpia f a s c i c u l a t a . Journal of Ecology 66: 1 5 - 3 3 . Watson, A. K. 1972. The biology and c o n t r o l of Centaurea d i f f u s a Lam., and Centaurea maculosa Lam. MSc. Thesis, U n i v e r s i t y of B r i t i s h Columbia. Watson, A. K. and A. J . Renney. 1974. The Biology of Canadian weeds. 6. Centaurea d i f f u s a and C. maculosa. Canadian Journal of Plant Science 54: 687 - 701. 208 Watt, K. E. F. 1960. The e f f e c t of population density on fecundity i n inse c t s . Canadian Entomologist 92: 674 - 695. Watt, K. E. F. 1965. Community s t a b i l i t y and the strategy of b i o l o g i c a l c o n t r o l . Canadian Entomologist 97: 887 - 895. Wellington, W. G. 1950. E f f e c t s of r a d i a t i o n on the temperatures of insect habitats. S c i e n t i f i c A g r i c u l t u r e 30: 209 - 234. Werner, P. A. 1975. Predictions of fate from rosette s i z e i n teasel Dipsacus fullonum L . ) . Oecologia (Berl.) 20: 197 - 201. Wilson, E. 0. and W. H. Bossert. 1971. A Primer of Population Biology. Sinaur Associates, Inc., Stamford, Connecticut. ZwBlfer, H. 1976. (personal communication) Lehrstuhl ftlr Zoologie, U n i v e r s i t S t Bayreuth, 858, Bayreuth. ZwBlfer, H. 1970. Investigations of the h o s t - s p e c i f i c i t y of Urophora  a f f i n i s F r f l d . (Dipt., Trypetidae). Commonwealth I n s t i t u t e of B i o l o g i c a l Control, Progress Report No. XXV. ZwBlfer, H. 1974. Innerartliche Kommunikations-systeme b ei Bohrfliegen. B i o l o g i e i n unserer Z e i t / 4. Jahrg. 1974/ Nr. 5. ZwBlfer, H. 1976. Investigations on Sphenoptera (Chilostetha) jugoslavica Obenb. (Co. Buprestidae), a possible b i o c o n t r o l agent of Centaurea  d i f f u s a Lam. (Compositae) i n Canada. Z e i t s h r i f t fUr angewandte Entomologie 80(2): 170 - 190. ZwBlfer, H. and P. H a r r i s . 1971. Host s p e c i f i c i t y determination of insects for b i o l o g i c a l control of weeds. Annual Review of Entomology 16: 159 -178. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0095186/manifest

Comment

Related Items