Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Modification of microclimate by the blueberry leaf-tier, Cheimophila salicella (Hbn.) (Lepidoptera: Oecophoridae) Contant, Hélène 1988

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

Item Metadata


831-UBC_1988_A6_7 C66_7.pdf [ 5.8MB ]
JSON: 831-1.0097662.json
JSON-LD: 831-1.0097662-ld.json
RDF/XML (Pretty): 831-1.0097662-rdf.xml
RDF/JSON: 831-1.0097662-rdf.json
Turtle: 831-1.0097662-turtle.txt
N-Triples: 831-1.0097662-rdf-ntriples.txt
Original Record: 831-1.0097662-source.json
Full Text

Full Text

MODIFICATION OF MICROCLIMATE BY THE BLUEBERRY LEAF-TIER, CHEIMOPHILA SALICELLA (HBN.) (LEPIDOPTERA: OECOPHORIDAE). by HELENE CONTANT B.Sc, Universite Laval, 1982 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDD3S Department of Plant Science We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL 1988 ° Helene Contant, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at The University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Plant Science The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: APRIL 1988 ABSTRACT The ecology of Cheimophila salicella Hbn. (Lepidoptera: Oecophoridae), a blueberry leaf-tier was studied on high-bush blueberry, Vaccinium corymbosum L., in Richmond, British Columbia. The females frequently laid their eggs in the lichen Xanthoria sp., an oviposition site not previously reported for this species. The possible microclimatic advantages of such behaviour are discussed. In the Field, females required longer than males to complete their 6th instar, so females were usually bigger than males in that instar. The leaf shelter made by the larvae modified their microenvironment in the field. On clear and sunny days, measurements of shelter temperature were 6-7°C above those of ambient air. The shelter temperature remained warmer than the air as long as the incoming radiation levels were high. As the radiation levels dropped, the shelter temperature fell to, or a little below, air temperature. On cloudy days, there was no significant difference between the daily maximum shelter and air temperatures. Under clear skies, the daily amplitude of temperature fluctuations was greater inside the shelter than outside. A laboratory investigation of the effects of such fluctuations on development showed that a large amplitude increased the developmental rate of the lst-4th instars. This increase in rate of development was probably due to an accumulation of extra thermal units (Yeargan 1980) occurring in the large-amplitude regime. However, the high temperature of this regime retarded pupation, and the later instars required longer to complete their ii development. Overa l l , larvae in the smal l and large amplitude regimes required the same amount of time to develop from hatching to pupation. A third regime, "medium ampl itude", slowed l a rva l development, probably because the length of its thermophase was longer than that which the insect normal ly encountered in the field. F i f t h - and sixth-instar females took longer than males to complete their development, both in the laboratory and in the field. The larger amplitude regime produced heavier pupae; females were, on average, 12.7 mg heavier than males. The microclimate of the shelter provides the larvae with more degree-days than i f they were subjected to ambient air and therefore promotes faster development. Without the extra degree-days provided by the shelter, C. salicella would not be able to complete its l a rva l development before the f irst lethal autumn frost. i i i TABLE OF CONTENTS ABSTRACT ii Table of Contents iv List of Tables vi List of Figures vii ACKNOWLEDGMENTS ix I. INTRODUCTION 1 II. CHEIMOPHILA SAUCELLA: DESCRIPTION AND LIFE HISTORY 3 A. DESCRIPTION OF THE INSECT 4 B. LIFE HISTORY 7 III. BIOLOGICAL NOTES 10 A. METHODS 10 B. RESULTS 14 1. Notes on the life history and ecology of C. salicella 14 2. Larval density in the field 21 3. Proportion of males and females 24 4. Density of individual instars 27 C. DISCUSSION 30 1. Life history and ecology of C. salicella 30 2. Larval density in the field 32 3. Proportion of males and females 32 4. Length of development of individual instars in the field 34 IV. INSECT THERMOREGULATION 35 A. LITERATURE REVIEW 35 1. Regulation of metabolic heat 36 2. Heat exchange with the environment 37 3. Microclimate 42 B. MATERIALS AND METHODS 50 C. RESULTS 53 1. Clear conditions 55 a. Solar radiation 55 b. Shelter and air temperatures 55 c. Differences between shelter and air temperatures 63 2. Cloudy conditions 69 a. Solar radiation 69 b. Shelter and air temperatures 69 c. Differences between shelter and air temperatures 72 3. Other results 72 iv D. DISCUSSION 75 V. EFFECTS OF THERMOPERIOD ON INSECT DEVELOPMENT 83 A. INTRODUCTION 83 B. MATERIALS AND METHODS 88 C. RESULTS 94 1. Mortality 94 2. Developmental time 94 a. Early instars 94 b. Middle instars 97 c. Late instars 97 d. 1st- to end of 6th-instar 98 3. Pupal weight 99 D. DISCUSSION 99 1. Mortality 99 2. Effect of temperature regimes on larval development 102 3. Importance of the shelter on larval development 106 VI. CONCLUSION 108 LITERATURE CITED I l l APPENDIX I 118 APPENDIX II 119 APPENDIX III 120 APPENDIX IV 121 APPENDIX V 122 APPENDIX VI 123 APPENDIX VII 124 APPENDIX VIII 125 v LIST OF TABLES TABLE 1. Weather during the days when microclimatic measurements were taken in a high-bush blueberry field in Richmond, B.C. in 1985 54 TABLE 2. Mortality for each laval period of Cheimophila salicella larvae reared at daily average temperatures of 12° and 16°C 95 TABLE 3. Time required by Cheimophila salicella larvae to complete each larval period in 3 temperature regimes having a daily average temperature of 16°C 96 TABLE 4. Weights of 14 day-old Cheimophila salicella pupae from larvae reared in three temperature regimes having a daily average temperature of 16 °C 100 vi LIST OF FIGURES FIGURE 1. Cheimophila salicella female (a) and male (b) resting on a branch of high-bush blueberry in Richmond, B.C. Note the atrophied wings of the female and her extruded ovipositor 5 FIGURE 2. Two typical leaf-shelters made by the larvae of Cheimophila salicella on high-bush blueberry plants in Richmond B.C. The first one (a) consists of 2 green leaves tied together, and the second (b) is made of one leaf severed at the petiole and tied to a green leaf. 12 FIGURE 3. Xanthoria sp., a lichen abundant on high-bush blueberry in Richmond, B.C. is hiding a small egg mass of Cheimophila salicella. The eggs are dark pink, indicating that the embryonic development is well underway 15 FIGURE 4. Frequencies of different categories of shelter made by the larvae of Cheimophila salicella throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C 19 FIGURE 5. Changes in Cheimophila salicella larval density per leaf throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C 22 FIGURE 6. Proportions of Cheimophila salicella male and female larvae throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C 25 FIGURE 7. Densities of Cheimophila salicella larvae in each instar throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C 28 FIGURE 8. Average radiation levels (a), air and shelter temperatures (b), and T e x c e s s (c) measured on August 3, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 56 FIGURE 9. Average radiation levels (a), air and shelter temperatures (b), and T e x c e s s (c) measured on August 14, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 58 FIGURE 10. Average radiation levels (a), air and shelter temperatures (b), and T e x c e s s (c) measured on August 17, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 60 vii FIGURE 11. Radiation levels (a), and T e x c e s s (b) measured for a northeast-facing shelter made of one green and one red leaf in a high-bush blueberry field in Richmond, B.C. on August 17, 1985 65 FIGURE 12. Radiation levels (a), and T e x c e s s (b) measured for a south-facing shelter made of one green and one red leaf in a high-bush blueberry field in Richmond, B.C. on August 17, 1985 67 FIGURE 13. Average radiation levels (a), air and shelter temperatures (b), and T e x c e s s (c) measured on August 7, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 70 FIGURE 14. Average radiation levels (a), air and shelter temperatures (b), and T e x c e s s (c) measured on July 29, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 73 FIGURE 15. Average minimum and maximum temperatures between April and October at the Richmond Nature Park station of Environment Canada calculated over a 9-year period (1977-1985). The station was located < 1 km from the study site. (Bars represent SD.) 89 viii A C K N O W L E D G M E N T S It has been a priviledge for me to work under the guidance of Dr. W.G. Wellington. His philosophy of science and broad knowledge have been a source of inspiration. I also thank h im for his continual mora l support and for the many extras that he considered being " a l l part of the service". Dr. V.C. Runeckles has provided many helpful suggestions throughout al l stages of this thesis. I am grateful for his interest and encouragement. Dr. B.D. F razer always had time for discussion and his approach to ecological research helped me develop a new perspective towards both ecology and research. I thank Dr. D.H. Henderson for her insights into this work, for her encouragement and for editing the thesis. I am grateful to P. Therr ien for many valuable discussions and for helping wi th some of the microclimatic measurements. M r . J . Raine helped me find a study site and offered several ideas in the developing stage of this study. M y thanks go to Mr . F. Barone and Mr . P. Virg ino for al lowing me to use their high-bush blueberry field. Dr. J . H a l l patiently helped with the statistics. Dr. J.-F. Land ry confirmed the identification of Cheimophila salicella (Hbn.), and Dr. W.B. Schofield and F. Lutzoni identified Xanthoria sp. and Parmelia sulcata Tay l . R. Cheng, F. Contant, V. Erho, J . Ke l l y , and M . K i ngma provided invaluable technical assistance. F ina l ly , I want to thank L. Wi l lms for her generous hospitality and D.J. Quir ing for his help at a l l stages of this work, his understanding and encouragement. I am grateful for the f inancial help provided by a Post Graduate Scholarship from F C A C . ix I. I N T R O D U C T I O N Temperature is one of the more important factors regulating growth and development of insects (Andrewartha and Birch 1954; Chapman 1982; Wigglesworth 1972). Each species has its own range of temperatures in which it can survive and develop. Within that temperature range, development usually proceeds faster at the higher levels. Many insects have adaptations that allow them to survive and develop at temperatures outside their preferential range; e.g., at levels that may, for a time at least, accelerate their development. Faster development can benefit insects by reducing the length of time that they are vulnerable to parasitism, for example, or by allowing them to go into diapause before the onset of adverse conditions. Such adaptations can be behavioural (e.g., basking), physiological (e.g., supercooling), or physical (e.g., a hairy body). Shelters made by some leaf miners, leaf rollers, tent makers, and bag makers can increase the temperatures and/or humidities present in these insects' immediate surroundings (Barbosa et al. 1983, Henson 1958a, Henson 1958b, Henson and Shepherd 1952, Sullivan and Wellington 1953). Cheimophila salicella (Hbn.) (Lepidoptera: Oecophoridae), a leaf-tying caterpillar commonly found on high-bush blueberry, Vaccinium corymbosum L., constructs shelters which probably modify surrounding temperatures. This microlepidopteran makes a shelter by tying or rolling together one or more leaves of a shoot. It has a long larval stage, lasting from May to October. Without 1 2 the occasional increases in temperature created by the shelters, it is conceivable that this insect could not pupate early enough in the autumn to escape the first lethal frost. Although C. salicella is considered a pest on high-bush blueberry in Br i t i sh Columbia (Raine 1966), very little is known about its ecology. Accordingly, the f irst part of this thesis outlines the insect's life history (chapter II) and some observations on its ecology (chapter III). One a im of this thesis was to determine the effects of the shelter on temperatures experienced by C. salicella larvae. Field temperature measurements inside and outside the shelter at different times of the day and under various cl imatic conditions were used to assess the effect of the shelter on microclimate (chapter IV). Chapter V deals wi th the potential effects of large fluctuations of temperature on the development and surv iva l of C. salicella larvae. II. CHEIMOPHILA SALICELLA: DESCRIPTION AND LIFE HISTORY Cheimophila salicella (Hbn.) (Lepidoptera: Oecophoridae) is a palearctic species found from England to Siberia (Meyrick 1927) as well as in Japan (Hodges 1974). In North America, it was reported for the first time in 1955 by Andison (Raine 1966) in the Lower Fraser Valley of B.C. By 1962, the insect was considered a nuisance pest in commercial fields of high-bush blueberries, Vaccinium corymbosum L. Wherever the crop is hand-picked, larvae feeding among the berries can be picked along with them. Where machine picking is used, larvae feeding on the leaves as well as those on the berries are dislodged into the picking crates. In both instances, additional sorting is required. Although the insect's main host plant is the high-bush blueberry, it is also commonly found on Salix sp. (Salicaceae) and Spirea sp. (Rosaceae) and may occur on Betula sp. (Betulaceae), Alnus sp. (Betulaceae), Acer sp. (Aceraceae), Prunus sp. (Rosaceae), Myrica sp. (Myricaceae), Berberis sp. (Berberidaceae), Cornus sp. (Cornaceae), Potentilla sp. (Rosaceae), Ledum sp. (Ericaceae), Kalmia sp. (Ericaceae), Vaccinium sp. (Ericaceae), and Rubus sp. (Rosaceae) (Raine 1966; Gillespie 1981), all of which may grow in the acidic peat bogs in which commercial blueberries are planted. It has also been recorded on Rosa sp. (Rosaceae) in Holland (Reichert 1932). 3 A . D E S C R I P T I O N O F T H E I N S E C T 4 The adult female (8-13 mm long) is the only brachypterous oecophorid found in North America north of Mexico (Raine 1966, Hodges 1974). The body (Fig. la) is grey speckled with white. The atrophied fore wing is light grey and a horizontal dark band divides the wing approximately in half. A dark section is also found at the apex of the wing. The fore wing is 3-5 mm long (Hodges 1974). The ovipositor, visible when extruded, is orange. The legs have alternating white and black bands. This camouflage conceals moths resting on old bark. The adult male has fully developed wings (Fig. lb) and is 10-12 mm long with the wings folded (Raine 1966). The colour and pattern of the forewing varies from one individual to another. Most are light brown with 2 transverse brown bands separating the wing in 3 sections. On the side of the band closest to the body, there is a very light patch, varying in colour from white to beige. In some individuals this patch is very small, whereas in others it is larger and resembles a triangle with its base resting on the dark brown band. The hindwing is brown and lacks a distinct pattern. A detailed taxonomic description of the insect is given by Hodges (1974). The eggs are smooth, oval and about 0.75 mm long (Raine 1966). Opaque, white, and sticky when first laid, the eggs later become reddish (Raine 1966). The larvae are about 1.2 mm long when they hatch and about 24 mm 5 FIGURE 1. Cheimophila salicella female (a) and male (b) resting on a branch of high-bush blueberry in Richmond, B.C. Note the atrophied wings of the female and her extruded ovipositor. 7 long when mature (Raine 1966). There are 6 instars. The abdomen is whitish-green. Following a moult, the head capsule is smooth and yellowish, but becomes dark brown as the cuticle hardens. The prothoracic shield is dark brown and the legs are black. The metathoracic legs are lobate and enlarged, and they fibrillate when the larva is disturbed (Reichert 1932). The pupa is obtect and yellow when first formed. Later, it becomes light brown with darker wing pads. The pupa of the male is generally smaller in size (both in length and in width) than that of the female. Raine (1966) reported average lengths of 10 mm for the male and 13 mm for the female. The male and female pupae can also be differentiated by the ratio of wing-pad length to pupal length, as the wing-pads of the males extend at least two-thirds of the pupal length, whereas those of the female extend less than one-half (Raine 1966). B. LIFE HISTORY In the field, adults' emerge from the ground between mid-March and mid-April. Males emerge a few days before females in a ratio of 1 male to 2 females (Raine 1966). Emergent females climb up the stems of blueberry bushes and begin "calling" with their abdomen and ovipositor extruded. A female-produced pheromone has been isolated and identified by Gillespie et al. (1984). Mating may last between 20 min. and l h and males mate with more than one female (Raine 1966). 8 Unmated females may lay eggs, but if males are present the females rarely oviposit before they mate. Mated females lay an average of 440 eggs, whereas unmated ones lay an average of 382 (Raine 1966). The eggs are placed under loose bark, under the scales of flower buds and in the axils of new leaves (Raine 1965, 1966). According to Raine (1966), the most favoured site for oviposition is under the shreds of loose bark at the base of the bush. In the field, eggs may incubate for 8 weeks and all the viable ones hatch during a 2-week period starting in the middle of May (Raine 1965, 1966). The larvae emerge in numbers on the first warm day after the eggs have matured (Raine 1966). Larvae emerging from eggs laid under flower bud scales enter the flowers and eventually feed on the fruit as it develops. Infested berries often ripen prematurely and have dead flower parts clinging to them (Raine 1966). Larvae emerging from eggs laid in the axils of leaf buds enter the bud and feed on the innermost leaf, usually cutting it at the base so that ultimately it dries and turns black. Larvae emerging from eggs laid under the loose bark walk upwards to the leaf and flower buds. When the leaf buds have opened, the larvae that were feeding among them either fold over the edge of one leaf or tie 2 leaves together. Either way, the larvae produce shelters for themselves. One of the two leaves is often severed at the petiole and soon turns red. Later, as the leaf dries, it turns brown. Such leaves are very conspicuous in the summer when all other leaves 9 are still green. Larvae may feed on their own shelter or browse outside it (Raine 1966). Pupation occurs in October. Most larvae pupate within the leaf shelters, which later fall to the ground. Occasionally they remain attached to the branches (Raine 1966). Larvae have also been reported to pupate in the litter (Raine 1966). Parasites of the larval, pupal and adult stages of C. salicella have been recorded. Gillespie (1981) found one Apanteles sp. (Braconidae) and one Glypta sp. (Ichneumonidae) parasitizing the larvae. Raine (1966) reported that Macrocentrus iridescens French (Braconidae) parasitized the larval stage and in turn was parasitized by a secondary parasite, Habrocytus sp. (Pteromalidae). Dipterous parasites of the pupal stage, Compsilura concinnata (Mg.) (Tachinidae) and of the adult stage, Tomosvaryella sp. (Pipunculidae) have also been recorded (Raine 1966). III. BIOLOGICAL NOTES Cheimophila salicella has been studied very little in North America. The only investigators, Raine (1965, 1966) and Gillespie (1981, 1984), were mainly concerned with the economic impact of the insect on the blueberry industry in B.C. Their studies therefore concentrated on obtaining basic information on the life cycle, assessing the relative abundance of the larvae and the amount of damage, identifying parasites of C. salicella and their potential for controlling the insect, and determining the effectiveness of various insecticides and the proper timing of their application, if required. This chapter presents some new information on the behaviour and ecology of C. salicella, including observations on the types of shelter made by the insect throughout its larval stage. Since there was no information on the duration of the various instars of C. salicella in the literature, this aspect of its life cycle was also recorded in conjunction with data on larval density and sex ratio. A. METHODS The work was done in a high-bush blueberry field located in Richmond, B. C. The 1.8 hectare field was bounded on the north by a house and small garden, on the east by a cranberry field, on the south by a small patch of young blueberry bushes about 1.2 m tall, and on the west by another high-bush field. The blueberry bushes had been planted at least 10 years ago, but the new owner did not know the exact planting date. The field contained several 10 11 blueberry cultivars. Spraying for mummy berry was done, but no insecticide was used in this field. Pruning was done in the spring. Data on larval development, density, and preferred shelter types were collected between June 18 and October 28, 1984. Every bush in the field was given a number so that those to be sampled could be selected each week from a random number table. During the picking season, however, bushes that had already been picked were not sampled. After a few weeks, harvesting was completed, and random sampling of all bushes was resumed. An average of 6.5 bushes were sampled each week (excluding the 2 weeks when there was no sampling). Bushes were not re-sampled. From each bush sampled, one branch on each of the cardinal quadrants was chosen and all the leaves on this branch and its ramifications were examined for the presence of C. salicella larvae. The total numbers of leaves with and without larvae were recorded and, for each larva, the instar, type of shelter and sex, when known, were noted. When larval age could not be accurately determined, such individuals were recorded as "unknown instar". Similarly, when their sex could not be distinguished, individuals were classified, "unknown sex". Shelters were divided into the following categories: 1) green shelter: consisted of 2 green leaves tied together (Fig. 2a); 2) semi-dried shelter: consisted of 1 green and 1 dead leaf tied together (Fig. 2b), or 1 partially dried leaf folded or rolled on itself; 3) dried shelter: consisted of 2 dead leaves tied together or of 1 dried leaf folded or rolled on itself; 4) "other" shelter: any shelter which did not fit the previous categories. 12 FIGURE 2. Two typical leaf-shelters made by the larvae of Cheimophila salicella on high-bush blueberry plants in Richmond B.C. The first one (a) consists of 2 green leaves tied together, and the second (b) is made of one leaf severed at the petiole and tied to a green leaf. 14 B . R E S U L T S 1 . Notes on the life history and ecology of C. salicella Two species of lichen, Xanthoria sp. and Parmelia sulcata Tayl. were abundant on the blueberry bushes. Xanthoria sp. occurred in the upper section of the bushes, whereas P. sulcata was located mainly in the lower section. Females of C. salicella frequently oviposited in Xanthoria sp. (Fig. 3) but never laid their eggs on the other kind. Some eggs were laid between the lichen and the branch surface, and were well hidden However, when a large mass of eggs was laid in the lichen, some were deposited on the upper surface of the lichen and could be seen more easily, especially as they developed to the stage when they turned pink. (Infertile eggs remain white and do not hatch.) Most eggs were located under loose bark, but contrary to Raine's findings (1966), they were not at the base of the canes. The loose bark at the base was usually grey, thin and easily peeled. The loose bark which the 1984-1985 generations seemed to prefer was located mainly in the top half of the bushes, where the branches were reddish-brown. This bark was thicker than the bark at the bottom of the canes, and not so easily peeled. Raine (1966) mentioned that larvae foraged outside their shelters. I have also seen evidence of this. Leaves close to C. salicella leaf shelters showed increasing damage from one day to the next. Rarely, however, have I seen a larva eating leaves outside the shelter during the day. The only larvae I have 15 FIGURE 3 . Xanthoria sp., a lichen abundant on high-bush blueberry in Richmond, B.C. is hiding a small egg mass of Cheimophila salicella. The eggs are dark pink, indicating that the embryonic development is well underway. 16 17 seen outside shelters during the day were feeding on the berries or occasionally resting on a twig or leaf. The amount of damaged leaves suggested that many larvae must feed outside their shelters. I checked for possible nocturnal feeding at 2300h (standard time) and at OlOOh, but in those periods, at least, the majority of the larvae were in their shelters. C. salicella larvae initially occupy flower or leaf buds (Raine 1966). However, as the leaves open and expand, a greater variety of larval constructions appear. Apart from the categories listed previously, the larvae sometimes loosely tie a leaf, dead or alive, to dried flowers or fruit. A leaf may also be tied to a small branch. Occasionally, more than 2 leaves are tied together by 1 larva. I have also seen 2 different shelters having 1 leaf in common. The larva uses silk to tie leaves together. It does so by moving back and forth from one leaf to the other, each time pulling the leaves closer to each other, and lining the developing shelter with a layer of silk. Silk is also added whenever the larva enters or leaves its shelter. Eventually, there is a silk tunnel inside the shelter, and this is where the insect usually rests. The ends of the tunnel remain open, serving as entrance and exit holes for the occupant. Larvae not only foraged outside, they also consumed their own shelter, eating both the green and the dry leaves. Frequently, shelters made of 1 live and 1 dry leaf showed more evidence of feeding on the dry than on the live one. Sometimes, when the cover offered by one leaf was much reduced by feeding activity, the larva tied a new leaf over the small one. However, larvae often 18 seemed to tolerate shelters in which 1 of the 2 leaves was just big enough to cover the insect. I have also frequently seen shelters made of one rolled, dried leaf scarcely long and wide enough to contain the larva. These dried leafrolls were not usual ly found on the periphery of the bush, but rather in the inside layer where there was more shade. One la rva made a shelter of blue "f lagging tape" and had, as holes in the tape testified, eaten some of it. Raindrops and morning dew affected live and dead leaves differently. The brown leaves absorbed the water, whereas the green ones were less permeable. The noise produced by the f ibri l lat ion of the metathoracic legs could be heard >1 m from the larva. This behaviour occurred at night as wel l as in the day. The abdominal colour of the larvae varied according to their size and their diet. In the field, the abdomen of early- instar larvae were yellow, whereas older larvae feeding on leaves had a green abdomen. Their translucent skin could appear blue-grey when they were feeding on berries. Later, in the fa l l , their skin took a pink colour as they reached their pre-pupal stage. In the pre-pupal stage, the insect did not feed or move very much. The thorax and the head compacted together, reducing the body length. The legs were no longer functional and the body turned whitish-yellow. Comparisons of the types of l a rva l shelters that appeared during the summer and fa l l (Fig. 4) showed that the green and semi-dried shelters were abundant at the beginning of the sampling period, each comprising 3 5 % of the total number of shelters in June and Ju l y . Af ter the week of August 6, the green shelters were more abundant (55%) than any other type of shelter, but 19 FIGURE 4. Frequencies of different categories of shelter made by the larvae of Cheimophila salicella throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C. JULY A U G U S T S E P T E M B E R O C T O B E R 1984 D A T E to o decreased rapidly after the week of September 10. The frequency of semi-dried shelters declined steadily from the beginning to the end of the sampling period. The dried shelters were never very abundant, except during the week of October 8, when they represented about 30% of all shelters. "Other" shelters were relatively abundant (30%) at the beginning of the sampling period, but were very scarce during July, August and September. Thereafter, they became most abundant as the number of green shelters decreased. In fact, during October the "other" shelters represented between 45-65% of the total shelters present. 2. Larval density in the field With the exception of a 2-week period between Aug. 13 and Aug. 27, the density of C. salicella larvae on the bushes (Fig. 5) decreased gradually from the beginning of sampling on June 18 to its end on Oct. 28. The initial density on June 18 was =0.10 larva/leaf. By the week of July 2, the average larval density had decreased to 0.07 larva/leaf, and remained relatively constant during the following 5 weeks. A subsequent increase in density occurred during the week of Aug. 13, and densities of 0.15 larva/leaf were found until the end of the following week. This increase was quickly followed by a pronounced decrease in population, and by the week of Sept. 10 the density was less than 0.04 larva/leaf. Few insects were found between the weeks of Sept. 24 and Oct. 10, when densities ranged from 0.01 to 0.02 larva/leaf. 22 FIGURE 5. Changes in Cheimophila salicella larval density per leaf throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C. 0.125-1 0.000 1 1 1 1 1 1 1 1 1 T 1 1 1 1 1 1 1 1 r— 13 26 2 9 16 23 SO 6 13 20 27 3 10 17 24 1 8 16 22 JUNE JULY AUGUST SEPTEMBER OCTOBER 1984 D A T E to 24 3. Proportion of males and females Pre-gonads were not apparent in half-grown larvae early in the summer, and it was not until July 2 that the sex of some older larvae could be identified. The proportion of males (Fig. 6) fluctuated around 0.45 for most of the summer until the week of Sept. 17. A sharp decrease in the week of Sept. 24 brought the proportion of males down to 0 by the week of Oct. 8, followed by an increase to 0.20 in the week of Oct. 15, before the ultimate decrease during the last week of sampling. The proportion of females (Fig. 6) remained between 0.20 and 0.30 up to the week of Aug. 13, when it rose to 0.40. A steep increase in the proportion of females corresponding to the decrease in males occurred during the week of Sept. 24, and the proportion of females remained high for 4 weeks, attaining a peak value of 0.88 during the week of Oct. 1. Summations of the proportions of individuals of the 2 sexes do not equal 1.00 because of the presence of larvae of unknown sex. The proportion of larvae of unknown sex was relatively high (about 0.25) from the week of July 2 to the week of Aug. 13. A decrease in the proportion of these individuals followed, and it remained less than 0.10 between the weeks of Aug. 27 and Oct. 1. At the end of the sampling period, during the week of Oct. 22, their proportion rose to a very high value of 0.67. 25 F I G U R E 6. Proportions of Cheimophila salicella male and female larvae throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C. I .O - i » • ' 2 0 2 7 3 1 0 1 7 2 4 1 8 1 6 2 2 J U L Y A U G U S T S E P T E M B E R O C T O B E R DATE to 27 4. Density of individual instars At the time sampling began (June 18), the larvae were already in their 2nd and 3rd instar (Fig. 7). All larvae had completed their 2nd instar by July 2 (no sampling was done in the week of June 25). The 3rd instar was completed during the week of July 9. The density of 4th instar larvae rose from 0 to 0.059 larva/leaf between the weeks of June 18 and July 2. It peaked at 0.064 larva/leaf during the week of July 9 and, as the density of 4th-instar larvae decreased to 0.010 larva/leaf in the week of July 23, the density of 5th-instar larvae increased to 0.056 larva/leaf and reached a maximum value near 0.07 on the following week. The 5th-instar density remained high for a longer period of time than did that of the 4th-instar (about 4 weeks compared to 2 weeks for the 4th instar). The density of 6th-instar larvae increased slowly at first, then more rapidly in the week of August 13. It reached a maximum on August 20, when the density of 5th-instar larvae was still relatively high, decreased between the weeks of August 27 and September 10, increased again in the week of September 15 and slowly decreased until the week of October 22. A size discrepancy between the sexes of both 5th- and 6th-instar larvae made it difficult to differentiate the two age groups. The density of larvae of unknown instar remained very close to 0 up to the week of Aug. 6. At that point, it increased with the density of the 6th-instar larvae and reached a maximum value of 0.031 on the week of Aug. 27. In that week, the density of larvae of unknown instar was greater than that of recognizable 6th-instar individuals. The density of larvae of unknown instar gradually decreased 28 FIGURE 7. Densities of Cheimophila salicella larvae in each instar throughout the 1984 summer in a high-bush blueberry field in Richmond, B.C. D E N S I T Y O F L A R V A E / L E A F 30 thereafter to a low value during and after the week of Sept. 24. Larvae took longer to complete their 5th and 6th instar than to complete their 4th instar. The first pre-pupae were seen during the week of Sept. 10 and in the week of Oct. 8 the first pupae were recorded on the bushes. C. DISCUSSION 1. L i f e h i s tory and eco logy o f C. salicella Species survive by spreading the risks of mortality through their populations (Wellington 1977). One way to achieve this goal is to use different oviposition sites. The females of C. salicella seem to use 2 kinds of sites. The first consists of open areas, such as flower buds. There the eggs have the advantage of being well exposed to radiant heating during the day, thereby appreciably increasing the rate of embryonic development. During a calm and sunny spring this advantage might outweigh the dangers of greater exposure to parasites and predators. During a windy spring, convective cooling would reduce the advantage of radiant heating in these exposed places. The second kind of oviposition site is more protected. It includes patches of lichen and loose pieces of bark. Both sites minimize exposure to wind and rain and thus reduce heat loss due to convection and evaporation. Both lichen and loose bark absorb moisture from the air and may provide the eggs they shield with a relatively high level of moisture for longer periods than the exposed eggs would experience. Since bark and lichen also can be heated by solar radiation (Lewis 1962; Kershaw 1985), they also provide higher than ambient temperatures for incubation. Protected sites have some disadvantages, however. Larvae hatching from eggs laid in loose bark or lichens have to travel farther to their first feeding sites than larvae from eggs laid on leaf and flower buds, the initial source of food. Small larvae often fall from twigs, which increases the risk they face after hatching. Nevertheless, the females prefer these protected oviposition sites and use them when they are available. Larvae from eggs laid in protected sites are probably more prone to disperse than those from eggs laid on the food source. As the newly emerged females walk up the stems, they first encounter the protected sites and later, as they continue their ascent toward the end of the branches, reach the flower and leaf buds. It is therefore possible that most of the eggs first produced by the females are laid in the protected sites. The first eggs laid by the females are also those that have received the greatest amount of nutrients during their development. Wellington (1959, 1965) has shown that the amount of nutrients received by the eggs of the western tent caterpillar, Malacosoma californicum pluviale (Dyar), determines individual behaviour which in turn can affect survival. This is an aspect of the ecology of C. salicella that could be studied further. Green and semi-dried shelters seemed to be preferred by most larvae over dried ones, but the reason for the greater abundance of green shelters in mid-August and early September is not clear. In contrast, the relative abundance of "other" shelters in mid-June was due to the presence of dried flowers at that 32 time. The next increase in "other" shelters at the end of September was almost entirely due to the high proportion of shelters made of yellow or red leaves. 2. Larval density in the field As sampling was not started until mid-June, I missed the first peak in larval density. A high mortality was probably sustained by early instar larvae as they often have to walk toward their food source or disperse to other branches. The slow decrease in population between June 18 and Aug. 6 reflected a steady mortality during that period. The peak in density encountered during the weeks of Aug. 13 and Aug. 20 was unexpected, however. Hand picking of the berries had started before then and by Aug. 9 one half of the field had already been picked. Since disruption from picking can disturb the population of C. salicella on a bush, it seemed preferable to sample bushes only in the "unpicked" half of the field. Perhaps cultivars more susceptible to C. salicella were sampled in this section of the field and caused the higher densities recorded during the weeks of Aug. 13 and Aug. 20. After the week of Aug. 20, when all the bushes had been picked, the whole field was sampled as before. The low density level obtained after the end of August reflected natural mortality (and eventually pupation), but also mortality caused by picking. 3. Proportion of males and females Males and females were easier to differentiate when the larvae were well grown, as evidenced by the high proportion of larvae of unknown sex during 33 earlier sampling. Since the proportion of each sex during the early instars was probably close to that in the later instars, it can be assumed that many of the "unknown" larvae were, in fact, female. That assumption brings the proportion of males and females in the early stages to =50%. A major change occurred after the week of Sept. 17, when the proportion of females increased dramatically, while that of males decreased nearly to 0%. This change was due to the shorter developmental period of the males. After most of the males had pupated, the females still required at least 3 more weeks to complete their larval development. The last increase to 20% in the proportion of males shown in Fig. 6 was an artifact caused by the very small number of larvae left in the field. The increase in larvae of unknown sex that occurred at the end of the sampling period represented the high proportion of pre-pupae then present in the field whose sex could not be determined because their opaque colours (pink or yellow) masked the pre-gonads. Male and female larvae were present in these samples in a 1 to 1 ratio despite the fact that adults occur in a ratio of 1 male to 2 females (Raine 1966). A more detailed study than mine should include assessments of mortality factors affecting late-instar larvae as well as pupal mortality. My preliminary observations showed a high rate of pupal parasitism, so there may be some selection of potential hosts by the parasites that discriminates against the early-maturing males. 34 4. Length of development of individual instars in the field The progression of instars in the field was mainly predictable. However, the density of 6th-instar larvae was lower than expected. The higher number of "unknown" instars during the weeks of Aug. 27 and Sept. 10 could have included a large proportion of 6th-instar larvae which would account for the lower than expected densities during these weeks. In fact, adding the totals of these unknown instars to that of those for the 6th instar (Fig. 7) provided a more realistic pattern for the latter during the period from Aug. 27 to Sept. 15. The second peak in the density of both 5th- and 6th-instar larvae seems due to the lack of synchrony between male and female development that was noted earlier. The more prolonged development of females also explains their heavier weight. In future work, accurate- determination of duration of the 5th and 6th instar would require separate information for each sex. Similarly, head-capsule width, which has been used previously to determine each instar (Raine 1966), may be misleading in the last 2 instars; it should also be measured for each sex, to improve the accuracy of the determination. IV. INSECT THERMOREGULATION A. LITERATURE REVIEW Insects have a relatively large surface to volume ratio and therefore may lose heat and water to their environment in amounts that can be detrimental. Often, therefore, their survival depends on their ability to keep their body temperature and water content within safe limits by balancing the losses against the gains. Several characteristics of both the environment and the insect affect its ability to achieve these thermal and water balances. Since an insect's water content constitutes 50-90% of its body weight (Chapman 1982), it is faced with the problem of retaining water or replacing the losses. Relative humidity of the air surrounding an individual is modified by air temperature, wind speed and solar radiation. The effects of these are in turn dependent on the individual's size, the permeability of its cuticle, and other factors. In general, larger insects lose less water than smaller ones. The permeability of the cuticle is well correlated with the habitat (Bursell 1974a) as the less permeable ones are found in xeric environments and the more permeable ones in more humid milieux. Thermal and water balances are not independent of each other; many mechanisms used to regulate high body temperature will at the same time decrease the water losses. Insect body temperature (T D) is dependent on two factors: (1) changes in the production and internal distribution of metabolic heat, and (2) alterations of 35 36 heat exchange with the environment. 1. Regulation of metabolic heat When the air temperature (T a) is low, winged insects increase their body temperature by producing metabolic heat, often by contracting their flight muscles. Since the ability to fly depends on muscle temperature, when T a is low a warm-up period (shivering) during which flight muscles are contracted is necessary to increase thoracic temperature (T^); e.g., the hawk moth, Deilephila nerii L., "shivers" to achieve the narrow range of 32°-36°C needed for flight (Dorsett 1962). Similarly, Heath and Adams (1965) reported that the sphinx moth, Celerio lineata Fabr., stabilizes its thoracic temperature during flight over a range of ambient temperatures; i.e., it achieves some endothermic regulation. Bees also increase by shivering. The bumblebee Bombus vagans Smith can forage for nectar at temperatures ranging from 5°C in the shade to 31°C in sunshine (Heinrich 1972a). While sitting on flowers at T a<24°C it maintains a thoracic temperature close to 32°-33°C, the minimum temperature required for flight. When T a>24°C, regulation of thoracic temperature ceases and T^h rises. Similar endothermy has been demonstrated in other bumblebee species, notably B. terricola Kirby, B. edwardsii Cresson and B. uosnesenskii Radoskowski (Heinrich 1972b,c). Heinrich (1972c) found that in some species T ^ was regulated, but the abdominal temperature was not and varied with T a. A queen, however, seems to control heat transfer from the thorax to the abdomen, as she heats her brood primarily through her abdomen even though heat is produced exclusively in the 37 thorax (Heinrich 1972d). Heat transfer to the abdomen also takes place in the dragonfly Anax Junius (Drury) (May 1976b) and modification of abdominal haemolymph circulation seems to be involved in the control of heat loss by the abdomen. Thermoregulation is a complex process determined by the interaction of several variables. Insulation or high metabolic rates can compensate for small size (McNab 1970) but in general the rates of cooling of a small insect are too high to allow substantial endothermy to occur (Heinrich 1974). Therefore for most insects, body temperature, T D, is dependent on T a (depending on the behavioural mechanisms used to control T D) and habitat selection becomes important in determining the conditions of temperature and humidity to which the insect will be subjected. 2. Heat exchange with the environment Insects can exchange heat with the environment through conduction, convection, radiation and evaporation. In most environments, insects gain heat mainly through radiation (Parry 1951) and body temperature increases linearly with increases in solar radiation (May 1979). Insects lose heat principally by convection (Digby 1955). However, the effects of various sources of heat gain and loss depend on characteristics of the insect itself, such as its size, surface area, colour, geomet^ and the microsculpture of its surface area. For a given size (volume), insects with greater surface area will have a 38 greater exchange of heat and water with the environment, whereas for a given surface area, large insects (bigger volume) will have a greater rate of heat and water exchange than smaller insects. Large insects attain a higher temperature excess (T^ - T a) than small insects but take longer to reach it under constant conditions (Digby 1955, May 1976a, Willmer and Unwin 1981). Large insects, because their body temperature changes more slowly than that of small insects, can tolerate less stable environments and intermittently higher radiation than the latter. Furthermore, large individuals can become active at cooler temperatures than smaller ones can because of the greater temperature excess the former can achieve. However, large insects must avoid constant high radiation which could lead to overheating (Willmer 1982). Colour and texture of an insect's surface also influence its thermal balance. In fact, insects of higher reflectivity heat more slowly than dark forms of the same size (Willmer and Unwin 1981), the latter being better "black bodies". Surfaces bearing hairs, bristles, scales, etc., reduce the loss of heat and water by holding an insulating layer of air adjacent to the body. In dragonflies, the layers of air-sacs at the surface of the thorax provide such an insulation device (Chapman 1982). Many insects can exert some control in maintaining their body temperature within the preferred range by means of behavioural mechanisms (see review by Casey 1971) Several species need to attain a minimum body temperature in order to start activity. Body temperature can be raised above air temperature through radiative heating by exposing the largest surface of the body perpendicular to the sun's rays. At low air temperatures the desert locust, Schistocerca gregaria (Forsk.), lies on its side so that its lateral surface is 39 perpendicular to the sun's rays (Chapman 1982). The tiger beetle, Cicindela hybrida L., rests on a slope with its back perpendicular to the sun and the underside of its body on the warm sand, thus gaining heat by conduction as well as by radiation (Dreisig 1980). Insects can also orient themselves to the wind to increase or decrease heat and water losses. As the body warms up, the insect needs to minimize heat input and adopt different postures; e.g., S. gregaria then faces the sun, thereby minimizing the surface area intercepting radiation. At higher temperatures, both S. gregaria and C. hybrida extend their legs vertically under the body, thus lifting it away from the hot surface and increasing heat loss by convection. If the temperature excess is still too great, they move to a cooler microhabitat. For example, when overheated, C. hybrida buries itself in the sand, whereas S. gregaria seeks shade or moves above the ground onto vegetation where convective and evaporative heat losses are increased by stronger wind. In the evening when temperature falls, S. gregaria crouches to the ground, gaining heat by conduction (Waloff 1963). This last behaviour can also serve to lose heat to a cooler soil (Gamboa 1976). Evaporation may be the greatest source of heat loss in stationary insects (Chapman 1982). This heat loss, due to withdrawal from the body of the latent heat of vaporization, can be increased by low humidity, high temperature and wind. Individuals of some species tend to aggregate, decreasing their effective surface to volume ratio and thereby minimizing water and heat loss. This clustering permits some increase in body temperature without detriment since metabolic heat is shared and conserved while convective cooling is reduced (Willmer 1982). In winter, bees cluster together when T a<15°C and maintain 40 the temperature of the cluster at 20°-25°C via metabolic heat. Another mechanism used to regulate water loss consists of closing the spiracles when conditions become potentially too harmful. Another type of posture is exhibited by some dragonflies. Males of Dythemis cannacrioides Calvert decrease the incidence of solar radiation by pointing the abdomen toward the sun in a position referred to as "obelix posture" and shading the thorax with the wings (Gonzalez 1987). Similarly, Pachydiplax longipennis (Burmeister), a dragonfly that spends, most of its active period on perches, relies on the obelix posture to minimize solar heating. In addition, it perches horizontally with the long axis of the body perpendicular to the sun (May 1976b). This species does not usually seek the shade and thermoregulates simply by adopting different postures. In sunlight, its body temperature is relatively independent of ambient temperature during most of the day. An interesting mechanism exists in larvae of the tent caterpillar, Malacosoma disstria Hbn., that insures, among other things, that the larvae will not be overheated (Sullivan and Wellington 1953). Below a certain range of temperatures, the larvae are photopositive and therefore sit in the sunlight on the leaves. When their internal temperature is raised above this "reaction" temperature, the larvae cease to be photopositive and seek the shade on the undersides of the leaves. If their body temperature in the shade drops below their reaction temperature, they become photopositive again. In fact, in marginal situations - e.g., hot sun and cold air, - larvae may move continually between 41 the two sides of the leaf until they locate an appropriate zone in which they can rest. Basking is common in butterflies because they need a minimum thoracic temperature for flight. Basking allows the butterfly Argynnis paphia L. to maintain a body temperature of 32°-37°C. The position of its wings plays an important role in regulating the absorption of solar energy. To gain heat, the insect exposes itself to the sun with opened wings. When its body temperature is too high, it closes its wings completely or else seeks shade (Vielmetter 1958). Heodes virgaureae L. exhibits a similar behaviour (Douwes 1976). In sunny weather, it directs the upper surface of its body toward the sun and the angle between its wings is inversely related to air temperature and solar radiation. The butterfly Calisto nubila L., a lateral basker, keeps its wings dorsally appressed while basking and exposes the side of its body to incident solar radiation (Shelly and Ludwig 1985). By tilting its wings to keep them perpendicular to the incident solar radiation, it elevates its thoracic temperature faster and reduces the time spent basking. Females of the speckled wood butterfly, Pararge aegeria (L.), bask for longer periods than their male conterparts, thereby accelerating egg maturation (Shreeve 1984). Flight in this species results in significant convective cooling and can be used as a method of reducing T D. Flight also results in cooling in Precis villada F. (Heinrich 1972a). Church (1960) showed that forced convection is the primary avenue of heat loss in flying insects. Kingsolver (1985a,b) described a new basking posture, called reflectance basking, in which the wings are used as reflectors to reflect solar radiation to the body instead of absorbing it. This posture was described for four species of Pieris. 42 Immature dragonflies, having an incompletely hardened cuticle and thus subject to excessive water loss, seek shady and cool areas (May 1976b). Similarly, in some butterflies, survival of eggs and larvae is decreased in sunny habitats (Rausher 1979) as compared to shady ones. Many insects regulate their activity in such a way as to avoid extremes in temperature and water loss (Willmer 1982). The tenebrionid beetle, Eleodes obscura (Say), decreases water stress by restricting its activity to periods when the heat load is minimal or to microclimates that are moist and cool (Marino 1986). Flower mites exploit thermally favourable floral microclimates and avoid thermally adverse conditions within inflorescences, and they may disperse to different flowers by hopping on the bill of hummingbirds visiting the flowers (Dobkin 1985). 3. Microclimate Until Uvarov's major review of climate and insects (1931), the temperatures and humidities in insect habitats were often assumed to be the same as the ambient air. Uvarov showed that standard meteorological data were of little use to ecologists, as they bore only a distant and indirect relation to the microclimates where insects lived. He encouraged entomologists to study insects' microclimates. Wellington (1950) was one of the first entomologists to investigate environmental differences among small-scale microclimates. He studied the temperature of different components of a conifer and a broadleaf tree, measuring the surface temperatures of foliage and bark of twigs as well as the internal temperatures of vegetative buds, expanding shoots and staminate cones under different weather conditions. During a clear day, the temperature of 43 exposed foliage is higher than the air temperature. Radiation plays a major role since foliage exposed to clear skies but shaded from the sun has a lower temperature than the air. At night, the exposed foliage loses its heat by radiating to the sky, and becomes colder than the surrounding air. High thin overcast does not change foliage temperature significantly but a heavy overcast may decrease the difference between the foliage and the air. Dense clouds reflect and re-emit enough radiation to keep the temperature of the exposed foliage slightly above air temperature. Whenever clouds affect foliage temperature, a small variation in cloud thickness overhead provokes an immediate change in foliage temperature. In winter, a snow-covered branch reflects part of the incoming radiation, thereby preventing the branch from becoming warmer than air temperature. At night, this snow-covered branch, being insulated, does not lose as much heat as exposed surfaces and often stays warmer than the surrounding air. However, if there is room for passage of air currents underneath it, its temperature can fall to the air temperature. (Wellington 1950). Just as the angle at which an insect is exposed to solar radiation determines its T D, the angle at which a surface is exposed to solar radiation is also important. Wellington (1950) found that, at air temperatures varying from 15°-26°C, aspen leaves hanging vertically had a temperature of 0.5°-1.6°C higher than the air temperature, while coniferous foliage exposed more directly to the incoming radiation reached temperatures between 1.6°-8°C higher than that of the surrounding air. Furthermore, an aspen leaf rolled by a tortricid larva 44 and maintained at an angle of 40° to the sun, thereby intercepting more radiation than an undamaged leaf, had a temperature comparable to that of conifer foliage (Wellington 1950). Similarly, Henson and Shepherd (1952) found, at air temperatures varying from 25.3° to 24.6°C, that the temperature of the flat lower surface of a lodgepole pine needle was 4.5 °C higher than the air when the needle was parallel to the sun's rays, and 7.9°C higher when it was perpendicular to the incoming radiation. In contrast, outgoing radiation did not vary significantly with the angle of the exposed foliage; e.g., normal aspen and coniferous leaves, which are at different angles, have about the same temperature during the night. Rain does not directly alter foliage temperature. However, when the water present on the foliage starts evaporating, the foliage temperature drops. This decrease in temperature is due to the loss of heat necessary to vaporize the water. This effect of rain is common to both day and night conditions. During a clear night, coniferous staminate cones are from 5°-8.4°C warmer than vegetative buds, and by day can frequently be as much as 10°-14.5°C above the air temperature. Under a polar air mass, or under broken clouds, the flowers have a slower cooling rate than the buds. Although the temperature differences between the stem, flowers, vegetative buds and ambient air vary for each tree species, Wellington (1950) showed that the tendency of the flowers to be warmer than the buds was consistent for white spruce (Picea glauca Voss), balsam fir (Abies balsamea Mill.) and Jack pine (Pinus Banksiana Lamb.). 45 The length of time during which one part of the tree is exposed to the sun's rays has an important impact on the temperatures reached by the various sides of the tree throughout the day. Peterson and Hauessler (1928) and Haarlov and Petersen (1952) determined that temperature differences as great as 15°-20°C could be observed between the north and the south sides of a tree trunk. Lewis (1962) reports similar findings. Such studies (Lewis 1962, Peterson and Haeussler 1928; Haarlov and Petersen 1952; Wellington 1950) showed how different the microclimates in a very small space could be. Immature insects are often unable to move great distances or disperse from the site where they hatch. Their survival therefore depends largely on the choice of a proper habitat by ovipositing females (Inoue 1986). One insect that seeks out a favourable microclimate to oviposit is the montane butterfly, Euphydryas gillettii Barnes (Williams 1981). Its egg masses are located on the undersides of the uppermost southeast-facing leaves of its host shrub, Lonicera involucrata (Rich.). These leaves are perpendicular to the morning sun and are warmer at that time of day than differently oriented leaves. Williams demonstrated that the egg masses present on the leaves perpendicular to the morning sun hatched on average 6.1 days earlier than those on leaves parallel to the sun's rays. As the mornings in the study area were usually clear and cold and the afternoons warmer and partly or entirely cloudy, it was advantageous for the eggs to be subjected to higher temperatures early in the day. This thermal advantage is important for this species since the larvae must develop to their second instar (their diapausing stage) when temperatures can be as low as -5°C. The tiger swallowtail, Papilio glaucus L., also selects sites 46 exposed to the sunlight for its egg masses (Grossmueller and Lederhouse 1975). Larvae in sunny exposures developed 15-35% faster than laboratory-reared larvae lacking radiant input. Increased developmental rates allowed the completion of a second generation in some areas. Instead of seeking out favourable climatic conditions, some insects have survived detrimental or unfavourable temperatures and humidities by creating shelters which provide a favourable microclimate. The microclimate present in bags built by the evergreen bagworm, Thyridopteryx ephemeraeformis (Haw.), was investigated by Barbosa et al. (1983). It had been previously suggested that the role of the evergreen bagworm's shelter was to offer protection against predators and parasites (Sheppard and Stairs 1976, Davis 1964). On the exposed side of trees where the average air temperature was 34.34°C during the period of study, the temperature within a bag was on average some 3.29°C higher than the ambient air. On the shaded side, the internal bag temperature was 0.79°C higher than the average air temperature (31.57°C). Therefore, on the exposed side of the tree, the shelters have a temperature 4.67°C higher than those on the shaded side. Barbosa et al. (1983) did not find any temperature difference between the surface of the bag and its interior. Furthermore, there was no evidence that metabolic heat might be partly responsible for heating of the bags. It seems that the higher temperature inside the bags was due solely to their physical properties; i.e. their dark color. 47 Shelters built by insects may also modify the moisture content of the air surrounding the occupant. In its natural setting, the microhabitat of the casemaking clothes moth larva, Tinea pellionella L., is composed of a hygroscopic case made out of silk and muskrat hair. This kerophage larva is able to develop at any humidity levels including 0%. Chauvin et al. (1979) placed both case bearing and caseless T. pellionella larvae at various humidities and recorded their weight change. At 100% relative humidity, the weight of the larva with its case increased and stabilized at a level 30% higher than its initial weight, whereas in the absence of the case, the gain was reduced so that the weight remained fairly stable. At 95% R.H., the same phenomenon occurred for larvae in cases; but for exposed larvae, weight fluctuated more than for larvae at 100% and did not stabilize but tended to drop slowly due to evaporation of body water. At 55% R.H., the rate of weight loss of larvae - without cases was twice as great as for those with cases. The larval case provides a more humid environment than that of the ambient air and thus is a major adaptation, especially under arid conditions. In very moist conditions however, its beneficial role may be out weighed (literally) by its heaviness. In a permanently saturated atmosphere, many larvae cannot carry their heavy water-logged cases, so they desert them to build new ones. Among insect-created microhabitats, the webs of tent caterpillars provide very sophisticated climatic control. The tent of Malacosoma pluviale (Dyar) acts as a greenhouse raising the enclosed air temperature much higher than the outside air temperature (Sullivan and Wellington 1953). The tent is continually expanded by the addition of new rooms, all connected by small holes. The temperature 48 may vary widely from one pocket to another depending on the amount of radiation or shade that each receives. The larvae may therefore choose the most comfortable temperature. The silk walls of the tent slow the passage of water vapor, thereby ensuring that the inside of the tent is always moist. This moisture comes from water that has evaporated from the insects' bodies as well as from their frass and exuviae. On a cloudy day, after their first morning feeding period, the larvae may return to the tent or stay at their feeding site where the}' remain inactive. If they return to the tent, they rarely enter it immediately, but rather rest on its surface or near it. On a clear day, however, all the larvae return rapidly to re-enter their tent after their first feeding period. This behaviour seems to be due to a response to evaporation rather than to temperature (Sullivan and Wellington 1953). During dry clear days, the larvae need the tent moisture, whereas on cloudy days, outside humidity is sufficient. These different behaviours do not seem to be temperature-related, since many cloudy mornings are as warm or warmer than clear ones. As long as the humidity is high, the larvae do not enter the tent. Solar heating reaches a maximum at midday and decreases afterward. However, the air usually becomes drier and hotter for a few hours after noon. During that period, the larvae that are in the tent move to cooler pockets until mid-afternoon, when even the coolest may become too hot. Then the overheated larvae emerge quickly to rest on the tent's outer surface. By that time the outside air temperature has begun to drop and is more tolerable than any of 49 the temperatures inside the tent. Henson (1958b) studied the thermal effect of insect-caused structural modifications of poplar leaves. He measured the temperature of a normal, unmodified poplar leaf and compared it to that of leafrolls, leaf mines, galls, etc. that had been made by insects using poplar leaves. He found that the shape of the structure, air circulation at its surface or within, surface-volume ratio, colour and absorptivity all affected the structure's temperature. In almost every microhabitat studied, the temperature never reached the extreme values found on the exposed surface of an undamaged leaf. For example, in the field most Choristoneura conflictana (Wlk.) rolls were built in such a way that they provided a strong heating at lower levels of solar radiation but a relatively stable temperature at high levels of radiation. This allowed the insects to be active under moderate insolation when the ambient temperature was often considerably below their "physiological zero", but kept them from being overheated when the radiation intensity was very high. Maximum radiant heating was found in structures such as leaf mines, where air is stabilized in thin layers between thin sheets of tissue and where the air spaces are comparatively small. Minimum radiant heating was found in galls, where a large volume of air is partially enclosed by tightly packed or dense tissue (Henson 1958b). Since Uvarov (1931) pointed out that temperatures where insects live differ from the ambient air temperature, microclimatic studies have become 50 increasingly important in entomological research (Peterson and Wolfe 1956; Ruesink 1976; Ferro et al. 1979). Yet many entomologists still neglect microclimates and microclimate-related behaviours when they design programs for sampling, monitoring, spraying or modelling populations. Especially with shelter-building insects, one should not neglect possible links between the behaviour of the insect, the characteristics of its life cycle, and the microclimate where it lives. The study reported here is an attempt to describe the microclimate present in the leaf shelters built by the larvae of Cheimophila salicella (Hbn.) and its relation to the insect's developmental rate. B. MATERIALS AND METHODS The microclimate study was conducted in the high-bush blueberry field described in Chapter II. Leaf shelters on three bushes (Rubel variety) in the last southern row of the field were used, as their south sides were well exposed to the sun and any temperature differences between the inside of the shelter and ambient air would be evident. Sixteen shelters were chosen from all cardinal quadrants of the bushes (north, east, south, west). Each shelter was assigned to one of the following categories: a) two green leaves tied together (G/G); b) a reddish-brown dried leaf tied onto a green one (R/G); and c) a green leaf tied onto a dried one (G/R). Air and shelter temperatures were measured hourly over a 24h period, whereas hourly measurements of photosynthetically active radiation (PAR) were taken only during daylight. PAR is directly related to total solar radiation (Howell et al. 1983) and therefore is a good estimate of solar radiation. For a detailed study of a particular shelter, temperature measurements were taken in rapid succession concurrently wi th radiation. Measurements were made on 5 days from Ju l . 29 to Aug. 18, 1985 under weather conditions ranging from overcast to clear and warm. These days were divided into 2 categories for comparison: clear (Jul. 29, Aug. 3 (day only), Aug. 14, and Aug. 17) and cloudy conditions (Aug. 3 (night only), and Aug. 7 (day only). Measurements were taken from about 0800h (standard time) one morning to about 0700h the following morning. In one instance, Aug. 7, measurements were made only from 0900h to 2200h. The dates given mark the beginnings of each 24h period when the measurements were taken. The temperature measurements were made wi th copper-constantan thermocouples (wire diameter = 0.0 76 mm) connected v i a a switch box to a portable potentiometer (Leeds & Northrup 8692-2) reading in degrees Celsius. The thermocouples were arc-welded and calibrated against a mercury thermometer in a bath of water held at different temperatures. Two thermocouples were assigned to each shelter. One was inserted inside the shelter, in the silk tunnel, through one of the openings left by the 5th- or 6th-instar larva. The other thermocouple was positioned as close as possible outside the leaf shelter. They were held in place throughout the study with hairpins and clothespins. The thermocouples were left in place throughout the investigation a) to minimize disturbance of the larva; b) to minimize the risk of breaking open or modifying the characteristics of the shelter by moving the thermocouple; and c) to minimize handl ing time and to standardize the data-collecting procedure. Radiation was measured with a L i -Cor L i -188B quantum/radiometer/photometer reading in / u E m _ 2 s - 1 . The instrument was generally held perpendicular to the shelter, but was held perpendicular to 52 the sun when at least a portion of the shelter was directly facing the sun. The results from 3 sets of thermocouples which did not give accurate measurements at low temperatures had to be discarded. Consequently, 13 shelters provided the bulk of the readings; 3 in the north quadrant, 5 in the east, 4 in the south and 1 in the west. Two or three thermocouples functioned improperly for short periods of time (from 2h to 1 day) but seemed in order later. It is possible that the insect's presence in the shelter occasionally interfered with the proper functioning of the thermocouple. Correlation analysis (Reg Procedure, SAS Institute 1985) between the amount of radiation and the difference between the shelter and the air temperatures was conducted for sunny and cloudy conditions. The effects of climatic conditions and the type of shelter on the daily maximum T e x c e s s (difference between air and shelter temperatures) of each unit were analyzed by analysis of variance (GLM Procedure, SAS Institute, 1985). A separate ANOVA was used to assess the effects of the cardinal location of the shelters on the maximum T e x c e s s. In this ANOVA, the data for the west side were not included because only one shelter was located in this quadrant. The effects of the same factors on the nocturnal minimum T e x c e s s were analyzed in the same way. Only the minimum T e x c e s s occurring between 2000h and 0200h was used in this analysis because the sky started to clear around 0200h on the "cloudy" night. Among the "clear" nocturnal readings, only 5 of the 26 minimum ^excesses occuring between 2000h and 0200h did not correspond to the T e x c e s s for the entire night. A significance level of 5% was adopted for all analyses. 53 C. RESULTS The summer of 1985 was very warm and dry. In July, only a trace of precipitation was recorded for the whole month. The warm and sunny trend set in July continued in Aug. despite the fact that some cloudy and wet days were present at the beginning of the month. Sunshine in Aug. was well above normal, precipitation was below normal and air temperatures were near normal. As a result, more clear days than cloudy ones were available for investigation. Of the 4 clear days, July 29, Aug. 3, Aug. 14 and Aug. 17, the last 2 were very warm and had very few if any clouds. On the other hand, some high thin clouds were present on July 29 and again on Aug. 3. During the evening of Aug. 3, the clouds thickened to an overcast during the night. Daytime cloud was present on Aug. 7, but the overcast began to break up around 2200h, so no further nocturnal readings were taken past that time. Although the weather on July 29 might seem to a casual observer to be relatively clear, the results for this day will be discussed separately and were not included in the analysis. Clear conditions refer to Aug. 3 (day), Aug. 14 and Aug. 17, whereas cloudy ones refer to Aug. 3 (night) and Aug. 7 (day). A more detailed description of the climatic conditions prevailing in the field during the measurements is presented in Table I. 54 TABLE I. Weather during the days when the microclimatic measurements were taken in a high-bush blueberry field in Richmond, B.C. in 1985. DAY WEATHER CONDITIONS REMARKS JL 29 Sunny and warm 21h00 1 inversion High thin clouds OlhOO wind Next morning: cloudy 05h00 wind 06h00 clouding over 07h00 clouds Aug. 3 Sunny with high clouds 18h00 clouds (until about 02h00) Clouds thickening in evening 19h00 wind Night: generally overcast, 20h00 wind some wind 22h00 rain Next morning: sunny 02h00 clearing up slightly, no more wind 07h00 cloudy 08h00 some sun Aug. 7 Cloudy with isolated showers lOhOO overcast, some raindrops Some sun in the afternoon llhOO wind, some sun 12h00 partially sunny from here on 22h00 mainly clear, some fall-out raindrops Aug. 14 Sunny day 23h00 warm breeze Little or no clouds 04h00 breeze Clear night, warm and damp Next morning: sunny Aug. 17 Sunny and warm, no clouds 03h00 foggy Night: foggy with clear sky 07h00 clouds Next morning: sunny Standard time. 55 1. Clear conditions a. Solar radiation On clear days, the solar radiation (Figs. 8a, 9a, 10a) increased rapidly between 6h00 (STD time) and 0900h. The radiation level reached a plateau between 0900h and 1300h and subsequently decreased. The decrease did not occur as fast as the morning increase. Between 1800h and 0600h radiation levels were usually below 250 M E m _ 2s _ 1. On Aug. 14 (Fig. 9a) and 17 (Fig. 10a), the two sunniest days during which measurements were made, the average radiation at midday was between 1900 and 2000 juEm" 2 s " The maximum amount of incoming radiation measured (2510 //Em" 2 s " ') was received by a green shelter in the south quadrant at 1300h on Aug. 7. b. Shelter and air temperatures The air and shelter temperatures followed very similar patterns over a 24h period (Figs. 8b, 9b, 10b). In the morning, temperatures started to rise near 0600h. The maximum temperature inside the shelter occurred around 1300h whereas that of the air occurred 2 hours later around 1500h. In the afternoon, both temperatures decreased and reached their minimum values at night. On Aug. 14 (Fig. 9b) minimum average shelter temperature was recorded between FIGURE 8. Average radiation levels (a), air and shelter temperatures (b), and e^xcess (c) measured o n August 3, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) to o < o o ^ UJ or UJ IS 10 9-. 6H 3H -3 i—•—i 1 ' 1 ' 1 ' 1 ' 1 r- T ' 1 ' 1 ' 1 ' 1 8 10 12 14 16 18 20 22 24 2 4 6 8 TIME (HOURS) 58 FIGURE 9. Average radiation levels (a), air and shelter temperatures (b), and e^xcess (°) measured on August 14, 1985 in a high-bush blueberry Field in Richmond, B.C. (Bars represent SEM.) 60 FIGURE 10. Average radiation levels (a), air and shelter temperatures (b), and Texcess (c) measured on August 17, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) 62 2300h and OlOOh the following day, whereas the minimum average air temperature was measured at OlOOh. On Aug. 3 (Fig. 8b) and 17 (Fig. 10b) minimum average temperatures for both shelter and ambient air occurred between 0400h and 0500h. On Aug. 3, the maximum average temperature recorded inside the shelter was 33°C whereas it was 30°C in the air. On Aug. 14 and Aug. 17, the maximum average temperature in the shelter was 39°C, but ranged between 34° and 36°C in the air. Minimum average shelter and air temperatures on Aug. 14 were around 11°C and 12°C respectively. On Aug. 17, which had a cooler night, the shelter and air temperatures were 3°C and 4°C respectively. As noted, the shelter temperature increased faster than that of the ambient air in the morning and started to decrease earlier. On average, shelter temperatures were above those of the ambient air for almost 8 hours, from about 0800h to a little after the maximum air temperature was reached at about 1600h. After that, however, and until 0800h the following morning, the air was warmer than the shelter. The pattern of shelter and air temperatures resembled that for radiation. In the morning, both temperatures increased as radiation levels increased (compare Figs. 8b, 9b, 10b and Figs. 8a, 9a, 10a). In the afternoon, however, there was a time lag of 1 to 2 hours between the decrease in solar input and the subsequent decrease in shelter and air temperatures. 63 In spite of the high temperatures that were occasionally recorded in the leaf shelters, most larvae remained in their shelters at least until the end of the observations on Aug. 17. At that time, it was impossible to tell whether the occupant was still present in 1 of the 13 shelters, but another had definitely been abandoned. It was not clear whether the shelter had been deserted because of excessive heat or because extensive feeding had rendered it unacceptable. Although no mass exodus of the larvae occurred during periods of high temperatures, a larva inhabiting a south-facing R/G shelter kept its head outside the shelter during the high-temperature period on Aug. 17. c. Differences between shelter and air temperatures The difference between shelter and air temperatures (Figs. 8c, 9c, 10c) closely followed radiation patterns (Figs. 8a, 9a, 10a). The correlation between T e x c e g s and radiation (Appendix I) showed that 41.5% of T e x c e s s could be explained by radiation (F = 401.12, df= 1,563). As soon as radiation levels increased, T e x c e s s increased. There was no time lag such as the one in the response of shelter and air temperatures to a decline in radiation. T e x c e s s decreased when radiation levels decreased. As would be expected, T e x c e s s was greater during the day than at night. Daytime T e x c e s s could be as much as 16°C, although on average the maxima were between 6° and 7°C. At night, when there was no solar radiation, the air temperature was about 2°C warmer than that of the shelter. 64 The maximum T e x c e s s reached and the amount of time T e x c e s s was greatest varied between shelters. Figure 11 shows the changes in T e x c e s s for a R/G shelter facing northeast on Aug. 17. This shelter was exposed to early morning sun for a relatively short period and was shaded by an adjacent bush during most of the day. T e x c e s s increased very quickly early in the morning, peaked at 14°C around 0900h, and decreased quickly to reach a low value of 3°C by llOOh, finally reaching 0°C by 1600h. Although the shelter temperature was above that of the air for 8 hours, the values of T e x c e s g were high only during the first 3 hours. For comparison, Figure 12 shows the T e x c e s s of a south-facing R/G shelter for the same date. This shelter was not shaded by adjacent leaves or bushes and therefore was well exposed to the sun during the entire day. T e x c e s s of this south-facing shelter did not rise as fast as that of the shelter facing northeast; e.g., at 0900h the south-facing shelter had a T e x c e s s of only 7°C compared with 14°C for the northeast shelter (Fig. 11). However, the T e x c e s s of the south-facing shelter continued to rise to about 14°C at lOOOh. It reached a plateau between 1000 and 1400h, with a maximum of 16°C at 1230h, and decreased quickly after 1400h. By about 1600h the air was warmer than the inside of the shelter. 65 F I G U R E 11. Radiation levels (a), and T e x c e s s (b) measured for a northeast-facing shelter made of one green and one red leaf in a high-bush blueberry Field in Richmond, B.C. on August 17, 1985. 67 FIGURE 12. Radiation levels (a), and T e x c e s s (b) measured for a south-facing shelter made of one green and one red leaf in a high-bush blueberry field in Richmond, B.C. on August 17, 1985. 69 2. Cloudy conditions a. Solar radiation The daily pattern of solar radiation (Fig. 13a) under the cloudy pattern that occurred during the observation period was similar to that under clear skies, but levels attained were lower. The average radiation levels on Aug. 7 were below 1500 j i E m ^ s " 1 and peaked at 1387 A i E m - 2 s _ 1 at 1200h. The maximum average radiation level which occurred at 1300h on Aug. 7 was relatively high (1383 ± 251 j i E m _ 2 s " 1, mean ± SEM) due to intermittent sunlight. b. Shelter and air temperatures The patterns of shelter and air temperatures on Aug. 7 (Fig. 13b), a generally cloudy day, were similar to the patterns exhibited on clear days: a gradual increase until approximately midday, followed by a decrease, however, temperatures did not reached the high levels measured on clear days. Maximum average shelter and air temperatures were 26 °C (at 1300h) and 25 °C (at 1500h) respectively. Although these maxima were lower than those recorded on clear days, their time of occurrence remained the same. Minimum average temperatures were 13 °C for the shelters and 14°C for the air. After these minima were recorded at 2100h, no subsequent measurement was taken because the sky was clearing. The average shelter and air temperatures recorded at night 70 FIGURE 13. Average radiation levels (a), air and shelter temperatures (Jo), and ^excess (c) measured on August 7, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) I o < o o !< Cr: U J Q_ Z£ U J in v\ o X 30 20 H 10 H 0 9 6 -3 -0 -- 3 - i — - — i — • — i — • — i • i B Legend shelter olr - , — . — r — . — i — • — i — > — i — - • — i — • — i — • — i • i • — i — i — i — i — | — > — i — i — i — • — i — i — i — i — i — i — i — • — i — • — i — i — | 8 10 12 14 16 18 20 22 24 2 4 6 8 TIME (HOURS) 72 under cloudy conditions on Aug. 3 (Fig. 8b) remained relatively high (about 16°C) until the clouds started to break up around 0200h. c. Differences between shelter and air temperatures The values of T e x c e s s were significantly affected by radiation (F = 60.43, df= 1,129) (Appendix II) although radiation explained a smaller proportion of the variation in T e x c e s s (31%) than it did on clear days (41.5%). T e x c e s s during the day was significantly reduced (F= 17.88, df=l,45) (Appendix III) compared with that under clear skies. The average maximum value of T e x c e s s was only 5 ± 0.7°C compared with 10 ± 0.7°C in clear weather. The greatest difference occurring between shelter and air temperatures at night under cloudy skies was quite similar to that occurring under clear skies (F=3.22, df=l,33) (Appendix IV). 3. Other results The type of shelter and the cardinal location of the shelter on the bushes did not have a significant effect on the maximum T e x c e s s (F=1.65, df=l,45, Appendix III; and F = 2.21, df=2,41, Appendix V respectively). Similarly, these two variables did not have a significant effect on the minimum T e x c e s s (F=0.83, df=2,33, Appendix IV; and F = 0.46, df=2,30, Appendix VI respectively). Radiation (Fig. 14a), shelter and air temperatures (Fig. 14b) and 73 FIGURE 14. Average radiation levels (a), air and shelter temperatures (b), and Texcess (c) measured on July 29, 1985 in a high-bush blueberry field in Richmond, B.C. (Bars represent SEM.) ' ' 1 ' 1 ' 1 ' 1 ' I ' 1 ' 1 ' 1 ' — I ' 1 • 1 11 13 15 17 19 21 23 1 3 5 7 9 11 TIME (HOURS) 75 ^excess (Fig. 14c) obtained on July 29 - a day on which thin clouds appeared and began to thicken overnight - were quite similar to those obtained on a cloudy day (Figs. 10a, 10b, 10c) and a cloudy night (Figs. 8b and 8c) The radiation levels during the day were low (maximum mean value =1079 ± 179 j i E m ~ 2 s ~ 1 ) . Although shelter and air temperatures during the day were relatively high compared with those under cloudier skies on Aug. 7 (Fig. 13b), the T e x c e s s was similar to the pattern exhibited under those cloudy conditions (Fig. 13c). D. DISCUSSION Temperatures inside the shelter on clear days were higher than ambient air and this difference was mainly due to incoming radiation. Solar radiation had a direct effect on shelter temperature and an indirect effect on air temperature. On a cloudy day with low radiation levels, the shelter temperature did not differ greatly from the air temperature because there was little radiant heating. Conversely, on clear days, the high radiant input warmed the shelter to considerably higher temperatures. As long as incoming radiation levels were high, the shelter temperature remained higher than the air temperature. However, as solar radiation levels fell, the heat the shelters lost to the surrounding air by convection was not replaced by radiative heating, so the shelter temperature dropped. As all the irradiated surfaces, including the ground, became cooler later in the day, there was less convection, and the air temperature also decreased. Because shelter temperatures started to fall earlier than the air temperatures, 76 however, the rate of decrease in T e x c e s s was steeper than that of the air; i.e., the pattern of T e x c e s s followed that of the incoming radiation more closely than did the shelter or the air temperatures. The gradual drop in shelter and air temperatures during the late afternoon contrasts with the more sudden change that occurs in a shelter when a thick cloud passes between the sun and the tied leaves. In such instances, the insolation level drops suddenly and the shelter temperature, in the absence of any radiative heating, responds very quickly. The differences between shelter and air temperatures on cloudy and clear nights were not significantly different f rom each other, although it was expected that the differences would be smaller under cloudy conditions. The shelter, acting l ike a "black body", absorbs short-wave radiation during the day and re-emits long-wave radiation. A t night, when there is no incoming radiation, the long-wave emissions cool the shelter temperature to slightly lower values than the air, part icular ly under clear skies. It was expected that nocturnal clouds would re-radiate some of the outgoing heat and thus decrease the difference between air and shelter temperatures. Possibly the clouds on the night of Aug. 3 were not sufficiently thick to do this, but no suitable cloud occurred during the period when the measurements might have been repeated. In general, the diurnal relationships noted between air temperature, shelter temperature and radiant heating and cooling resemble those described by Well ington (1950) for the air temperature, the temperature of the different parts of a tree, and radiation. Some of the temperatures measured on clear days seemed high, perhaps 77 because of limitations of the available equipment. The thermocouples had to be fine enough to be placed in the shelter without destroying or modifying the shelter characteristics or disturbing the larva. On the other hand, they also had to be strong enough to be manipulated frequently. This problem was tackled by using very fine thermocouple wire inside the shelter and connecting it to a thicker, stronger wire which was then soldered to connectors that could be attached to a switch box. Unfortunately, the junction between the fine and the thick wires could not always be completely shielded from the sun, so that it was sometimes directly exposed to radiant heat. As a consequence, at high radiation levels some of the thermocouples indicated higher than normal temperatures. Since the thermocouples measuring air and shelter temperatures for a given shelter were close together, the same degree of error occurred in both. Therefore, even when both readings were inaccurate, the calculated values of T e x c e s s were acceptably close to the existing difference. Larval development of Cheimophila salicella is prolonged, lasting almost 5 months. As for all other poikilotherms, the larvae rely on adequate temperature levels to complete their development. It would therefore be advantageous to exploit the temperature benefits of shelters to accelerate development as much as possible. Shelter-making is a behaviour that exposes the occupants to temperatures as much as 6° to 7°C warmer than the air in clear weather. Development thus can proceed faster within shelters than outside them, especially when the air temperature is slightly lower than the larval developmental threshold. On such occasions, when skies are cloudless, the temperature in the 78 shelter would rise above the threshold and allow development to proceed. Any mechanism which can promote faster development and pupation before the first frost occurs will aid the survival of this species. As radiation explains only 40% of T e x c e s s, other factors are responsible for the rest of the variation. Air temperature, although indirectly dependent on radiation, is another important factor, especially in cloudy weather. Anything affecting the interception and absorption of incident radiation by the shelter will also affect the magnitude of T e x c e s s. ..Such variables include the physical characteristics of the shelter (colour, size, shape, etc.), its orientation (quadrants) and location on the bush (top, bottom), and its angle to the sun's rays. A shelter with a darker exposed surface should absorb more radiation than a lighter one, because it is a better "black body", and thus should be warmer in sunlight. This occurred in the mines of the lodgepole needle miner, Recurvaria milleri Busk, although the effect of colour was also related to mine size (Henson 1958). In the present study, the colour (type) of the shelter did not modify significantly the maximum value of T e x c e s s , but the experiment was only designed to show whether air and shelter temperatures differ from each other, and consequently did not allow for precise measurements of other factors that would help in understanding the variation in T e x c e s s. To detect any effect of colour, other variables, such as size of the shelter and its angular orientation to the sun would have to be held constant. The size and shape of the shelter then have to be considered separately. Henson (1958b) showed that a shelter with a greater surface to volume ratio will become warmer than one with a smaller 79 ratio. The shape of the shelter also will affect its capacity to intercept incoming radiation (e.g., a shelter with a convex surface would be able to intercept some of the sun's rays at right angles at several times of the day, whereas a "flatter" shelter would not. Thus a convex shelter could maintain a modest ^excess lasting several hours, whereas the flat shelter would have a high ^excess ^ o r a briefer period). The bushes used in this study had few shelters located on the north side and most of those were located on the upper part of the bush. Thus they intercepted the sun's rays for a much longer period than any north-facing shelter located lower on the bush. In other words, the insect did not reside (or survive for long) in the less suitable location. The side of the bush and therefore the direction a shelter faces determines when and for how long radiation will be received. Krenn (1933, cited in Geiger 1950) measured temperatures on each side of a tree during the day and found that the side facing east received radiation for 4.5h from 0700h to 1130h, the side facing south received radiation for 8.5h from 0830h to 1700h, the side facing west was heated from 1330h to 1800h (5.5 hours), whereas the side facing north received no direct radiation from the sun. The daily pattern of T e x c e s s in each of the 4 quadrants studied here were somewhat different from each other, as the comparison between the east-and south-facing shelters showed (Figs. 11 and 12), with 80 ^excess reaching a higher value earl ier on the east than on the south. Nevertheless, in a high-bush blueberry field such as the one used for this study, the overall effect of the quadrants may not be very important because of the amount of shading other bushes provide for many of the shelters. The amount of shading received by one shelter varies as the sun's position in the sky changes over time and as the bushes grow. When the berries are mature, they weigh down the branches, thus exposing to the sun shelters that would otherwise have remained in the shade. Consequently, the effect of the cardinal location of shelters is complicated by such changes, and its ultimate impact is difficult to assess. Af ter taking into account the possible impact of al l the foregoing factors on T e x c e s s , one might expect that the " i dea l " shelter would be a relatively large but thin construction located on the top hal f of the bush where shading is reduced and exposure to radiation is increased. That might be the " i dea l " shelter in part ly cloudy, relatively cool weather, but not necessarily in other kinds of weather. In fact, as climate varies from year to year, the species can insure its surv iva l only by spreading the risk among the individuals of a generation. Different types of shelters in different locations provide one method of insuring that some members of each generation w i l l be well-placed to cope with changing weather. Fo r example, dark shelters would be ideal for cool weather or for larvae l iv ing in frequently shaded areas. Dr ied shelters, which would expose the insect to relatively high temperatures i f subjected to intense radiation, would be safer in less exposed parts of a bush and, in fact, are frequently found in such 81 places (Chapter II). High temperatures did not seem to distress the larvae very much. Some of the days during which this investigation was conducted were among the hottest of a hot summer. Even on those days, there was no mass exodus. In contrast, in Choristoneura conflictana (Wlk.), the larvae drop out of their leafrolls when temperatures reach a value of 36°C (Henson 1958b). The only sign of overheating by C. salicella was when a larva partially emerged from its fully exposed shelter during the midday period. It seems that C. salicella larvae can withstand high temperatures, at least for a short period of time. More work is needed to understand how the insect copes with extreme temperatures. The shelters made of green leaves might also, through transpiration, provide a moister environment than dried leaves, which on occasion could improve the insect's development and survival (Bursell 1974b). The tightly closed leafroll of Compsolechia niveopulvella Chamb., for example, reduces water loss from both the leaves and the insects and increases the feeding rate of the occupants (Henson 1958a). As the shelter made by C. salicella is opened slightly at both ends, its ability to conserve moisture would not be so great as that of C. niveopulvella. Shelters made of one dried and one green leaf might be the most versatile, with a dark surface for rapid heating and a live leaf to provide moisture. 82 A n " i dea l " shelter should not only maximize the rate of development; it should also minimize mortal ity. Apa r t from exposure to lethal temperatures, other sources of mortal i ty include parasites and predators. La rvae in shelters well exposed to the sun are also wel l exposed to observant enemies, especially birds. Shelters made of one dried and one live leaf are part icular ly conspicuous. Therefore, there is probably a point at which trade-off occurs between adequate ^excess a n ( * increasing mortal ity. Consequently, an " i dea l " shelter probably does not exist. One that is " i dea l " on one day could be unsuitable the next. Shelter-making behaviour nevertheless seems to be an adaptation that allows C. salicella larvae to complete their development faster than i f they were entirely dependent on ambient temperatures. Without the extra degree-days provided by the shelter 's microclimate, the larvae would not reach the pupal stage before the temperature in this locality fell below their developmental threshold. V. EFFECTS OF THERMOPERIOD ON INSECT DEVELOPMENT A. INTRODUCTION The thermoperiod, the pattern of temperature fluctuation during the 24h day, is modified for larvae of Cheimophila salicella by the leaf shelters in which they develop. Consequently, the developmental rate of the larvae is also modified from rates that would occur outside the shelters, perhaps because the amplitude of temperature fluctuations is greater inside the shelters than in the ambient air. The amplitude of daily temperature fluctuations can affect the action of some factors involved in larval development. The proteolytic activity in the last-instar larvae of Tribolium confusum Duv. and T. casteneum (Hbst.) was increased at temperatures fluctuating between 20°C and 29°C (mean = 24.5°C) compared to the activity at a constant 28 °C. With an increase in proteolytic activity, protein nutrients were better utilized, resulting in an increase in metabolic efficiency and thus ultimately in a decrease in length of development (Birks et al. 1962). This section presents the results of a laboratory study of the development of C. salicella in three temperature regimes: one with large fluctuations, another with moderate fluctuations, and a third with a constant temperature. A l l three regimes had the same daily mean temperature. The relationship between constant temperatures and the rate of development of insects has been extensively studied. The relationship has been described as an S-curve (Wigglesworth, 1972) with a linear middle section 83 84 corresponding to the optimal range of temperature for development. Temperatures outside the optimal range retard the insect's development and increase mortality. Insects in the field are subjected to temperatures that fluctuate on a daily and a seasonal basis. Thermoperiods are analogous to photoperiods. Each thermoperiod consists of a period of high temperature, the thermophase, and a period of lower temperature, the cryophase. Both phases are defined by their duration and their temperature; e.g., a thermoperiod comprising an 8 h cryophase at 10°C and a 16 h thermophase at 25°C can be written, 8C:16T, (10:25°C) (Beck 1983). The rate of development of insects under alternating temperature regimes may be similar, accelerated or decreased compared to the rate of development at an equivalent mean constant temperature. Survival and length of development of larval and pupal stages of the tufted apple bud moth, Platynota idaeusalis (Walker), were similar in alternating and constant temperatures in the range 20-33 °C (Rock 1985). Growth rates of Pieris brassicae L. obtained under alternating temperatures conformed to values expected under equivalent mean constant temperature (Neumann and Heimbach 1975). The number of days from hatching to larval maturity in the European corn borer, Ostrinia nubilalis (Hbn.), was similar under alternating and constant temperatures (Beck 1982). And when the maxima of fluctuating temperatures were within the optimal temperature range for development, the developmental rate of Trichogramma pretiosum Riley was similar to that at comparable mean constant temperatures (Butler and Lopez 1980). No differences were found in the rate of development of eggs or nymphs of Lygus hesperus Knight when reared 85 under fluctuating temperatures within the range 15-33°C and equivalent mean constant temperatures (Champlain and Butler 1967). Other studies have shown that insects may develop faster under alternating temperatures than under the equivalent constant temperatures. Thermoperiods produced shorter stadia for the black cutworm, Agrotis ipsilon (Hufn.), than did comparable constant temperatures (Beck 1986). Exposure of the aphid parasite, Praon exsoletum (Nees), to fluctuating temperatures resulted in an increase in the rate of development relative to that occurring under constant conditions at all mean temperatures below 24°C (Messenger 1969). The same accelerative effect was shown for the spotted alfalfa aphid, Therioaphis maculata (Buckton) (Messenger and Flitters 1958, 1959; Messenger 1964). Development of the red turnip beetle, Entomoscelis americana Brown, in laboratory studies was accelerated significantly (6-9%) under alternating regimes where the difference between maximum and minimum temperature was 10°C, but not in regimes where the differences were 5 or 15°C (Lamb and Gerber 1985). This laboratory result was not supported by field data. In a few cases, development has taken longer under alternating temperatures than under equivalent mean constant temperatures. The development of the Pitcher-plant mosquito, Wyeomyia smithii Coq., at a constant 25°C was shorter than under a thermoperiod of 12T:12C, (31.5:18.5°C) lagging the 16L:8D photoperiod by 2 h (Bradshaw 1980). 86 Insects generally grow and develop at a faster rate under alternating temperature regimes in which the minimum temperature falls below the developmental threshold than under comparable mean constant temperatures. Males of Telenomus podisi Ashmead (which could not develop at all) and females (which could develop only marginally) at a constant temperature of 15.5°C successfully developed under an alternating temperature regime of 14:22°C (Yeargan 1980). At low alternating regimes of 6:20°C and 6:30°C, which had a cryophase below the estimated development threshold of 10°C, developmental rates and survival of the tufted apple bud moth increased relative to the mean temperatures of the alternating regimes. Thermoperiods with a cryophase lower than the developmental threshold tended to produce a lower developmental threshold for A. ipsilon (Beck 1986). The lowered developmental threshold under thermoperiodic conditions can be explained by the role of limiting factors and accumulated thermal units. Insects' developmental sequences involve several rate-limiting factors, particularly enzymes, whose temperature characteristics may differ. These rate-limiting factors may vary from one instar to the next and from species to species (Beck 1983). A thermoperiodic regime might satisfy all developmental rate-limiting factors, whereas a relatively low constant temperature might not (Sharpe and DeMichele 1977; Beck 1983). The eggs of Oncopeltus fasciatus (Dallas) for example need to be exposed to high temperatures (10°C above the hatching threshold) for a few hours a day in order to complete embryonic development, even though embryonic development can proceed at constant temperatures below the hatching threshold 87 (Lin et al. 1954; Richards and Suanraksa 1962). Yeargan (1980) has explained the importance of thermal units in relation to development in thermoperiods having a cryophase temperature below the developmental threshold. During such thermoperiods, the insect may accumulate more degree-days than it would at the equivalent mean constant temperature. In this situation, the thermoperiod should produce faster development. This prediction is based on the assumption that even though no significant development may occur during the time spent below the developmental threshold, neither is there any adverse effect from repeated exposures to low temperatures (Yeargan 1980). Under fluctuating temperature regimes in which the temperature of the thermophase exceeds the upper limit of the optimal range, insects have shown reduced survival and a slower rate of development than under the equivalent mean constant temperature (Yeargan 1980). Larval development of the tufted apple bud moth was retarded and survival was reduced when the temperature of the thermophase exceeded the upper limit of 33°C compared to the equivalent mean constant temperatures (Rock 1985). Similar findings were reported for Praon exsoletum (Messenger 1969). Alternating temperature regimes can also affect fecundity. Although larvae of Wyeomyia smithii took longer to develop under alternating temperatures, adults showed a 7-fold increased in fecundity (Bradshaw 1980) over those reared under the equivalent constant mean temperature. Siddiqui et al. (1973) reported faster 88 development, earlier attainment of maximum fecundity, and a shorter reproductive period for the pea aphid under fluctuating temperatures within the favourable range of temperatures, which resulted in a higher intrinsic rate of increase than that found at the equivalent mean constant temperature. Alternating temperatures can also affect larval and adult weight. Increased larval body weight (Welbers 1975, Beck 1986) and larger head capsules (Beck 1986) have been reported for some insects under alternating temperature regimes. Adult red turnip beetles weighed more when reared under fluctuating temperatures than when reared under equivalent mean constant temperature (Lamb and Gerber 1985). B. MATERIALS AND METHODS Three temperature regimes were established in Percival 1-35LL controlled environmental chambers. All treatments had a daily average temperature of 16°C, as this temperature occurs virtually throughout the whole larval stage of C. salicella (Fig. 15). The moderate fluctuating regime was represented by a thermoperiod of 8C:16T, (12:18°C) as these minimum and maximum temperatures are commonly found in the field during most of the insect's life cycle and the widely fluctuating temperature regime was represented by 18C:6T, (12:28°C), a thermoperiod simulating the greenhouse effect of the leaf shelters. These regimes will be referred to as 16°C, 12:18°C and 12:28°C respectively. The photoperiod in all growth chambers was 16hL:8hD until larvae molted to the 5th instar. At 89 FIGURE 15. Average minimum and maximum temperatures between April and October at the Richmond Nature Park station of Environment Canada calculated over a 9-year period (1977-1985). The station was located < 1 km from the study site. (Bars represent SD.) 91 this stage of their development, larvae were moved to a HhL:13hD photoperiod, which induced pupation. Fifty-one egg masses were collected from the high-bush blueberry field in Richmond, B.C. in mid-April, 1985 and randomly assigned to one of the 3 temperature treatments. Each egg mass was placed in a petri dish with moist paper filter and checked daily for hatching. Ten early-hatched larvae were used from each egg mass. As some of the egg masses were not viable, the number of masses in each treatment varied at the beginning pf the experiment (15 egg masses at 16°C, 16 in 12:18°C, and 14 in 12:28°C). In a few instances, less than 10 larvae per egg mass were available to start the experiment. Larvae were reared on blueberry leaves in 35 7-dram plastic snap-cap vials. The caps of the vials were glued onto an 18.5 x 24.5 cm sheet of 0.3 cm plastic paneling. A 0.4 cm hole was drilled in the middle of each cap through the paneling. The panel was then placed over a 15.5 x 23 x 5 cm plastic tray filled with water. Freshly cut blueberry stems bearing a few leaves were then inserted through the holes in the caps into the water. Upon hatching, larvae were placed individually on each group of leaves and then covered with a vial snapped onto the cap. Vial labels were coded so that the development of each individual could be followed separately. The leaves were replaced every 2 to 3 days as necessary. First and second instar larvae were examined daily for survival and 92 moult. Larvae that had left the leaves were returned using a fine brush. The brush was disinfected with alcohol and then rinsed in water to reduce the risk of spreading any viral or bacterial diseases. For the older instars, only the dates at which larvae moulted to 3rd and 5th instars, or pupated were recorded. To reduce the amount of disturbance, older instars were not re-examined for a few days after each moult. Thereafter, they were examined every other day until the size of the head capsule relative to that of the body indicated an impending moult to either the 3rd or 5th instar, or until the body changes indicated approaching pupation. The larvae were then examined daily to ensure that their moulting or pupation date was accurately recorded. The length of development was obtained in this way for the following periods: from hatching to the end of the 2nd instar; from the beginning of the 3rd to the end of the 4th instar; and from the beginning of the 5th to the end of the 6th instar. These periods of the larval stage will be termed "early", "middle" and "late" instars respectively. The length of development was also calculated for the entire larval stage, from hatching to the end of the 6th instar. Sex was determined by the absence (female) or presence (male) of pre-gonads visible through the skin of late instar larvae (Chapman and Lienk 1971). Pupal weights were measured on a Metier ME30 microbalance 14 days after pupation to allow time for the pupae to harden and stabilize their weight. This experiment is similar to a split-plot design. The length of development and pupal weight was analyzed by analysis of variance (ANOVA) 93 (GLM procedure, SAS Institute 1985) with the variation partitioned between the following factors: temperature regime, egg mass within temperature regime and larvae within egg mass. The variation due to larvae of a given egg mass is further partitioned into variation due to sex and that due to the interaction between sex and temperature regime. The proportion of dead larvae at the end of each period was calculated for each egg mass. Mortality for the entire larval stage was also calculated. An ANOVA (GLM Procedure, SAS Institute 1985) was performed on the arcsine-transformed values of these proportions to determine whether the mortality was similar in all temperature regimes. The analysis of mortality took into account all larvae involved in the experiment, whereas in the analysis of length of development and of pupal weight only insects of known sex that successfully pupated were included. Both mortality and length of development were analyzed for early, middle and late instars as well as for the entire larval stage. Comparisons of means were carried out by Duncan's multiple-range test (SAS Institute 1985). A significance level of 5% was used in all analyses. As the developmental threshold for C. salicella is not known, larvae were also reared at a constant 12 °C regime to determine whether the alternating temperature regimes included temperatures below the developmental threshold. The egg masses for this experiment were collected at the same time as those used in the 16 °C daily average temperature experiment, and the conditions were similar except for the mean temperature of 12 °C. Mortality was calculated using 94 the same method as in the 16 °C experiment. C. RESULTS 1. Mor ta l i t y The proportion of dead larvae per egg mass for each larval period was not significantly affected by the treatments at any stage of their development. Most of the mortality occurred during the first 2 instars (Table II). The mortality was almost 100% (n=150) during the first 2 instars in the 12°C constant regime (Table II), and only 4 individuals reached the 5th instar. 2. D e v e l o p m e n t a l t ime a. Early instars The length of development from hatching to the end of the 2nd instar was significantly different in different temperature regimes (F= 16.39, df=2,40) and between egg masses within a given temperature regime (F=1.67, df=40,122) (Appendix IX). Insects reared in the 12:28°C regime developed significantly faster than those reared in the other 2 regimes (Table III). Insects in the 12:18°C regime took the longest to reach the 3rd instar. About 2 5% of the larval stage was spent as early instars. 95 TABLE II. Mortality for each larval period of Cheimophila salicella reared at daily average temperatures of 12° and 16 °C. LARVAL PERIOD #DEAD / INITIAL # OF LARVAE 12°C 16°C 1ST AND 2ND INSTARS 3RD AND 4TH INSTARS 5TH AND 6TH INSTARS 1ST TO 6TH INSTAR 138/150 8/150 4/150 150/150 181/441 32/441 45/441 258/441 TABLE I I I . Time n q u l r a d by Che ImophlI a s a l i c e l l a l a r v a e to complete each l a r v a l p e r i o d In three temperature regimes h a v i n g a d a l l y average temperature of 1C*C. TEMPERATURE DAYS REQUIRED TO COMPLETE THE FOLLOWING PERIODS OF LARVAL DEVELOPMENT REGIME — — — • 1ST AND 2ND 3RD A NO 4TH 5TH A NO STH 1ST TO END OF INSTARS' INSTARS INSTARS STH INSTAR 16'C CONSTANT 24.5 ± 0.3 a 26.2 ± 0.4 a 44.2 1 0.9 a 95.4 ± 0.8 a 8C:16T. (12:18*C) 26.1 ± 0.3 b 27.5 ± 0.6 b 51.2 ± 1.3 b 104.5 ± 1.1 b 18C:6T. (12:28*C) 23.1 ± 0.4 c 19.5 ± 0.3 c 54.5 ± 0.9 c 97.4 1 1 .O a Means w i t h i n columns not f o l l o w e d by the same l e t t e r are s i g n i f i c a n t l y d i f f e r e n t (Duncan's n u l 1 1 p l e - r e n g a t e s t . P<0.05. SAS I n s t i t u t e 1985) 97 b. Middle instars The length of development from the beginning of the 3rd to the end of the 4th instar was significantly different among the temperature regimes (F= 131.94, df=2,39) (Appendix X). Insects reared in the 12:28°C regime developed from 3rd to 5th instar on average 7-8 days (almost 33%) faster than larvae held in the other temperature regimes (Table III). Insects in the 12:18°C temperature regimes took significantly longer to complete their development than did those in either 16°C or 12:28 °C. Females developed faster than males (F=7.54, df= 1.110) (Appendix X). This difference was not confirmed by Duncan's multiple range test on the non-adjusted means but a comparison of adjusted means showed that females required only 24 + 0.4 days to develop from 3rd- to 5th-instar, whereas males required 25 ± 0.4 days. The amount of time spent as middle instars was about 27% of the larval stage for insects reared in the 16°C and 12:18°C regimes (roughly the same as for the first 2 instars). However, insects reared in the 12:28°C regime spent slightly less that 20% of their larval stage as middle instars. c. Late instars The length of development of late instars differed significantly among the different temperature regimes (F=36.30, df=2,40) (Appendix XI). Larvae reared 98 in the 16 °C regime reached the pupal stage first, followed by those reared in the 12:18°C regime (Table III). Larvae reared in the 12:28°C regime required on average 3-10 days longer than larvae from the other regimes. There was also a significant difference (F = 51.26, df= 1,111) (Appendix XI) in the time males (46 ± 0.8 days) (mean ± SEM) and females (54 ± 0.9 days) required to complete their development. In all treatments, the period extending from the beginning of the 5th to the end of the 6th instar was the longest of all groups of instars. Larvae reared in the 12:28°C regime spent 56% of their larval development as late instars, whereas those in the 12:18°C and 16°C regimes spent 46% and 49% respectively. d. 1st- to end of 6th-instar There was a significant difference in the length of development of the larvae in different temperature regimes (F=20.96, df=2,41) (Appendix XII). Larvae reared in the 12:18°C regime took on average 6-8 days longer to develop from the 1st instar to pupae than those in the other 2 regimes (Table III). Females took significantly longer than males to go through their larval stage (F=29.92, df= 1,123), requiring on average 103 ± 0.8 days, whereas males took 95 ± 0.8 days. 99 3. P u p a l we ight There was a significant difference in pupal weight in the different temperature regimes (F=6.16, df=2,41) and between larvae from the same egg mass (F=1.65, df 41,124) (Appendix XIII). As a consequence, the difference in pupal weight between insects from different regimes must be examined with caution. Pupae of larvae reared in the 12:28°C regime weighed on average more than those reared in the other temperature regimes (Table IV). There was also a significant difference in pupal weights between males and females (F=62.41, df= 1,124) (Appendix XIII). Females weighed 42.9 ± 1.28 mg and males 30.2 ± 0.63 mg. D. DISCUSSION 1. Mor ta l i t y As Cheimophila salicella is exposed to temperatures in the field ranging from less than 12 °C to more than 28°C, it is not surprising that the different temperature regimes used here did not significantly affect mortality. Even the youngest larvae survived quite well under a daily temperature fluctuation of 16°C, an amplitude which is much greater than the air temperature fluctuations in the field during the period in which larvae are in this stage of their development. Since larvae in leaf shelters experience wider temperature fluctuations than those found in the outside air, they could be expected to 100 TABLE IV. Weigths of 14 day-old Cheimophila salicella pupae from larvae reared in three temperature regimes having a daily average temperature of 16 °C. TEMPERATURE REGIME PUPAL WEIGHT (mg) MEAN ± S E M 1 16 °C CONSTANT 8C:16T, 12:18°C 18C:6T, 12:28°C 33.9 ± 1.37 a 35.9 ± 1.48 a 40.0 ± 1.55 b Means not followed by the same letter are significantly different (Duncan's multiple-range test, P<0.05, SAS Institute 1985). 101 tolerate them as well as they did here. Most (39.8%) larval mortality occurred during early instars and was due in part to high humidity inside the vials. The 1st- and 2nd-instar larvae tended to disperse from the leaves and become trapped in condensation that accumulated on the sides of the vials. To decrease the humiditj', the plastic bottom of the vials was replaced with fine cotton mesh. This design was satisfactory for small larvae but not for later instars, which sometimes chewed holes through the mesh and escaped. Some larvae also escaped through the hole drilled in the vial cap. Wrapping "Play-doh" around the stem effectively blocked that escape route, although a few larvae occasionally found an opening near the stem to crawl through. Some mortality also was caused by handling. Although a soft brush was used to manipulate larvae, a few were injured when leaves were changed or when wanderers were put back on their leaves. Since early instars had to be checked daily for moulting, this disturbance also contributed to their mortality. The mortality sustained in the 12 °C constant regime was much greater than that in the daily average temperature of 16°C. The condensation problem was more severe at 12 °C, but does not explain such an increase in mortality. Indeed, the high mortality at 12 °C indicates that this temperature is not within the optimal range for Cheimophila salicella's development. Insects have different 102 developmental thresholds for different stages of their development (Bursell 1974b; Beck 1983). A variety of thresholds (developmental threshold, hatching threshold, developmental-hatching threshold, hatching-survival threshold, etc.) have been determined for various insects (Johnson 1940; Lin et al. 1954; Hodson and Al Rawy 1958; Richards 1959). The eggs of the hemipteron, Oncopeltus fasciatus (Dallas), require 13°C for only hatching to occur, but require 15°C for hatching and full development and over 17°C for development and the production of viable young bugs (Richards 1959). The eggs of C. salicella used in the 12°C temperature regime had been in the field for up to one month before they were brought to the lab. Diurnal fluctuation of temperatures in the field might have provided the eggs with temperatures that met the requirement for embryonic development to be completed. In the lab, the 12 °C constant regime might have been sufficient for the eggs to hatch. However, had the eggs been placed at 12°C immediately following oviposition, it is probable that hatching would not have occurred. In other words, 12 °C is probably lower than the developmental threshold of Cheimophila salicella. 2. Effect of temperature regimes on larval development Using the 2 criteria of length of time required for development and pupal weight as an indicator of fitness, larvae did best under the widely fluctuating 12:28°C temperature regime, and second-best in the 16°C constant temperature regime. There was no significant difference in the length of larval development between individuals in 12:28°C and those in 16°C. More importantly, the 103 individuals in the 12:28°C regime produced significantly heavier pupae than those from the 16°C constant regime. The 12:18°C temperature regime did not provide any advantage for larval development. Overall, larvae developed significantly slower in the 12:18°C regime than in the others and the pupae were not significantly heavier than those produced in the 16°C constant temperature. In many cases of faster insect development in fluctuating vs constant temperature regimes, the acceleration in development is due to an accumulation of extra thermal units in the fluctuating regime (Yeargan 1980). These extra thermal units are accumulated only when the following conditions are met: 1) the average daily temperature is similar in both constant and fluctuating regimes; 2) the temperature of the cryophase is below the developmental threshold. Mathematically, thermal units available to insects in a 24h period are calculated as follows: n L (Tj - a) X HTi, where i = 1 Tj = value of the ith temperature (°C) to which the insects are subjected in a 24h period; a = developmental threshold; H-p; = amount of time (h) the insects are subjected to Tj. In cases when Tj is lower than a, the value of Tj - a is set at 0. If the developmental threshold was <12°C, the thermal units accumulated by the larvae under each temperature regime in 24h would be: 104 16°C: (16°-12°C) x 24h = 96 DH (degree hours),1 12:18°C: (12°-12°C) x 8h + (18°-12°C) x 16h = 96 DH, 12:28°C: (12°-12°C) x 18h + (28°-12°C) x 6h = 96 DH, In this instance, the same number of thermal units is being accumulated every day in each of the temperature regimes. In this study, extra thermal units could be accumulated only if 12 °C was below the developmental threshold and, as previously discussed, that is probably the case. If the developmental threshold were above 12°C, e.g., 13°C, then the thermal units accumulated in a day for each temperature regime would be: 16°C: (16°-13°C) x 24h = 72 DH, 12:18°C: (12°-13°C) x 8h + (18°-13°C) x 16h = 80 DH, 12:28°C: (12°-13°C) x 18h + (28°-13°C) x 6h = 90 DH. In these circumstances, the thermal units accumulated in each regime vary. Whatever the developmental threshold (13°C, 14°C, etc.), the 12:28°C regime would always accumulate more thermal units in 24h than the 12:18°C regime, and the latter in turn would always accumulate more than the 16°C regime. The net result of these differences is that larvae subjected to 12:28°C should develop faster than those at 12:18°C, which in turn should develop faster than larvae reared at 16°C. The latter difference did not occur in this experiment (Table III). Larval development of insects subjected to the 12:18°C regime was slower than those in the 12:28°C and 16°C regimes. The significant increase in ' t o convert the amount of thermal units from DH to DD (degree days), divide the number of DH by 24h/day. 105 developmental time for larvae in the 12:18°C regime is probably attributable to the length of the thermophase (16 hours) a length not encountered in the field in spring or fall. Cook (1927), Peairs (1927) and Huffaker (1944) reported that the highest rate of acceleration in insect development under alternating temperatures is obtained when the thermophase is 6-8 hours. Higher temperatures in the fluctuating temperature regimes appeared to deter pupation in the last stage of larval development. Larvae subjected to the highest daily temperatures took longest to pupate whereas those at 16°C pupated fastest. A maximum daily temperature lower than that provided by the fluctuating regimes was probably required during the latter stages of larval development to trigger pupation. This is understandable, as air temperatures in the field in early October rarely, if ever, reach levels near 28°C. With the autumn mid-day sun at lower angles, temperatures in the leaf shelter would also usually be cooler than 28°C. Among the 3 temperature regimes tested in the laboratory, the 12:28°C regime was the most beneficial for C. salicella. In this regime, the larvae developed most rapidly through the first part of their larval stage. This accelerated rate enabled larvae to spend less time in their early instars and thus, in natural situations, would reduce larval mortality from predation. Most of the body weight was gained during the late instars, since the late instars occupied the bulk of the larval stage, both in the laboratory (Table III) and in the field (Fig 7). 106 3. Importance of the shelter on larval development The importance of the shelter on larval development of C. salicella is best shown by comparing the extra number of degree-days (DD) provided by the shelter and the total number of DD required by the insect to complete its larval development. The number of DD required for larval development can be estimated at 286 DD (from Table III and using a developmental threshold of 13 °C). Similarly, the number of extra DD provided by the shelter can be estimated using the following assumptions: 1- Developmental threshold is 13°C; 2- Thermal advantage of leaf and flower buds is similar to that of leaf shelters; 3- On cloudy or partially sunny days, the extra heat provided by the shelter ( T e x c e s s ) has a value of 0 6C; 4- On sunny days, T e x c e s s has a value of 6°C and lasts 5h. Between June 1 and August 31, air temperature is above 13 °C during the period of high ^excess a n ^ a n t n e e x t r a heat is available for larval development. During May and Sept., air temperature is sometimes lower than 13°C during the period of high T e x c e s s and it is estimated that the extra heat provided by the shelter contributes only 3°C above the threshold. In 1985, a total of 94 extra DD were provided to the larva by the shelter during the larval stage of C. salicella. C. salicella requires 286 DD to 107 complete its larval development, from hatching in May to pupation in October, when only 192 DD are available from air temperature alone. Without the effect of the shelter, the larvae would not develop beyond the 5th instar before frost occurred. The 94 DD provided by the shelter enables the larvae to be pupae and survive frost. Therefore the shelter is of critical value to C. salicella for survival in the climate under study. VI. CONCLUSION The shelters made by C. salicella larvae provide the insect wi th a microclimate warmer than the ambient air on clear days. Poikilotherms such as C. salicella larvae rely main ly on temperature for their development. Because warmer temperatures within the optimal range for development increase the rates of development, the warmer microclimate of the shelters allows the insects to complete their l a rva l stage earlier than i f they were subjected only to ambient temperatures. This is part icular ly important for the larvae of C. salicella, which pupate relat ively late in the fal l , even with the help of the higher shelter temperatures. The effect of the warmer microclimate of the shelter may be part icular ly important for the insect on clear and cold spring days when the low ambient air temperature prohibits development. The shelter, through radiative heating, would then provide the larvae with temperatures high enough to allow development to proceed. Solar radiation is one of the more important variables affecting the difference between shelter and air temperatures. Several factors that might modify the interception and absorption of incoming solar radiation by the leaf shelter need to be studied i f we are to better understand not only the microclimate in which C. salicella lives, but also the " tact ics " used by this species to survive in an uncertain environment. Some of these factors include the 108 109 orientation of the shelter to the sun, its cardinal position on the bush, the amount of shading it receives, and its colour, shape, and size. First- to fourth-instar larvae developed faster under the widely fluctuating regime than under the moderately fluctuating or the constant regimes. Later instars, however, required longer to reach the pupal stage in the widely fluctuating regime than in any other, probably because the high temperature of the thermophase interfered with their pupation processes. This slower development, on the other hand, allowed the larvae to produce heavier pupae. Overall, larvae in the widely fluctuating and in the constant regimes did not take a significantly different amount of time to develop from 1st instar to pupa. It seems that the long thermophase of the moderately fluctuating regime retarded the larval development of C. salicella. The developmental threshold of C. salicella was unknown at the time this study was undertaken. Since an accumulation of extra thermal units is the most likely explanation for faster development in the widely fluctuating temperature regime, this threshold is probably above 12 °C. The effects of the amplitude of temperature fluctuation on development, if any, cannot be distinguished from the effects of extra degree-days present in the fluctuating regimes. Although this experiment did not establish the effect of the amplitude of temperature fluctuation on development, it nevertheless provided useful information regarding the developmental threshold of C. salicella and confirmed the field observation that larvae subjected to a greater number of degree-days develop faster. Furthermore, 110 without the extra degree-days provided by the shelters, larvae of C. salicella would not be able to pupate before the first lethal frost in the fa l l . The leaf-tying behaviour thus contributes substantially to l a rva l development of C. salicella and is crit ical for its surv iva l in southern B.C. L I T E R A T U R E CITED Andrewartha, H.G., and L.C. Birch. 1954. The distribution and abundance of  animals. Univ. of Chicago Press, Chicago, 111. 782 pp. Barbosa, P., M.G. Waldvogel, and N.L. Breisch. 1983. Temperature modification by bags of the bagworm Thyridopteryx ephemeraeformi (Lepidoptera: Psychidae). Can. Ent. 115(7): 855-858. Beck, S.D. 1982. Thermoperiodic induction of larval diapause in the European cornborer, Ostrinia nubilalis. J. Insect Physiol. 28: 273-277. . 1983. Insect Thermoperiodism. Annu. Rev. Ent. 28: 91-108. . 1986. Effects of photoperiod and thermoperiod on growth of Agrotis ips.ilon (Lepidoptera: Noctuidae). Ann. ent. Soc. Am. 79: 821-828. Birk, Y., I. Harpaz, I. Ishaaya, and A. Bondi. 1962. Studies on the proteolytic activities of the beetles Tenebrio and Tribolium. J. Insect Physiol. 8: 417-429. Bradshaw, W.E. 1980. Thermoperiodism and the thermal environment of the Pitcher-plant mosquito, Wyeomyia smithii. Oecologia 46: 13-17. Bursell, E. 1974a. Environmental Aspects - Humidity, p. 43-84 in The physiology  of Insecta. M. Rockstein (ed.), Academic Press, New York and London. 568 pp. . 1974b. Environmental Aspects - Temperature, p. 1-41 in The physiology of Insecta. M. Rockstein (ed.), Academic Press, New York and London. 568 pp. Butler, G.D. Jr., and J.D. Lopez. 1980. Trichogramma pretiosum: development in two hosts in relation to constant and fluctuating temperatures. Ann. ent. Soc. Am. 73: 671-673. Casey, T.M. 1981. Behavioural mechanisms of thermoregulation, p. 80-114 in Insect thermoregulation. B. Heinrich (ed.), Wiley and Sons, N.Y. 328 pp. Champlain, R.A., and G.D. Butler. 1967. Temperature effects on development of the egg and nymphal stages of Lygus hesperus (Hemiptera: Miridae). Ann. ent. Soc. Am. 60: 519-521. Chapman, P.J., and S.E. Lienk. 1971. Tortricid fauna of apple in New York. Special Publ., New York State Agr. Expt. Sta., Cornell Univ., Geneva, N.Y. 122 pp. I l l 112 Chapman, R.F. 1982. The insects, structure and function. 3rd ed. Harvard University Press, Cambridge, Mass. 919 pp. Chauvin, G., G. Vannier, and A. Gueguen. 1979. Larval case and water balance in Tinea pellionella. Insect Physiol. 25: 615-619. Church, N.S. 1960. Heat loss and the body temperature of flying insects. II. Heat conduction within the body and its loss by radiation and convection. J. exp. Biol. 3 7 : 186-212. Cook, W.C. 1927. Some effects of alternating temperatures on the growth and metabolism of cutworm larvae. J. econ. Ent. 2 0 : 769-782. Davis, D.R. 1964. Bagworm moths of the western hemisphere (Lepidoptera: Psychidae). U.S. Nat'l. Mus. Bull. 2 4 4 : 1-233. Digby, P.S.B. 1955. Factors affecting the temperature excess of insects in sunshine. J. exp. Biol. 3 2 : 279-298. Dobkin, D.S. 1985. Heterogeneity of tropical floral microclimates and the response of hummingbird flower mites. Ecology 66(2): 536-543. Dorsett, D.A. 1962. Preparation for flight by hawk-moths. J. exp. Biol. 3 9 : 579-588. Douwes, P. 1976. Activity in Heodes virgaureae (Lepidoptera: Lycaenidae) in relation to air temperature, solar radiation and time of day. Oecologia 2 2 : 287-298. Dreisig, H. 1980. Daily thermoregulation and water loss in the tiger beetle Cicindela hybrida. Oecologia 4 4 : 376-389. Ferro, D.N., R.B. Chapman, and D.R. Penman. 1979. Observations on insect pest management. Environ. Entomol. 8: 1000-1003. Gamboa, G.J. 1976. Effects of temperature on the surface activity of the desert leaf-cutter ant Acromyrmex versicolor versicolor (Pergande). Am. Midi. Nat. 9 5 : 485-491. Geiger, R. 1950. The climate near the ground. Harvard Univ. Press, Cambridge, Mass. 482 pp. Gillespie, D.R. 1981. Introduced and native leafrollers (Lepidoptera: Tortricidae) on  berry crops in the Lower Fraser Valley, B.C. Ph.D Thesis, Simon Fraser University, Burnaby, B.C., Canada. 179 pp. 113 Gillespie, D.R., G.G.S. King, V.A. Salas-Reyes, and K.N. Slessor. 1984. Determination of the sex-pheromone components of Chemophila salicella (Lepidoptera: Oecophoridae), a pest of blueberry in British Columbia. Can. Ent. 116: 1397-1402. Gonzalez, S.E. 1987. Dythemis cannacrioides Calvert, a libellulid with unusual ovipositing behaviour (Anisoptera). Odonatologica 16(2): 175-182. Grossmueller, D.W., and R.C. Lederhouse. 1985. Oviposition site selection: an aid to rapid growth and development in the tiger swallowtail butterfly, Papilio glaucus. Oecologia 66: 68-73. Haarlov, N., and B.B. Petersen. 1952. Measurements of temperature in bark and wood of Sitka spruce (Picea sitchensis) with special reference to temperature in galleries of Hylesinus (Dendroctonus) nicans. Forstl. Forsoksvoesen Den. 21: 1-19. Heath, J.E., and P.A. Adams. 1965. Temperature regulation in the sphinx moth during flight. Nature 205: 309-310. Heinrich, B. 1972a. Temperature regulation in the bumblebee Bombus vagans: a field study. Science 175: 185-187. . 1972b. Energetics of temperature regulation and foraging in a bumblebee, Bombus terricola (terricola) Kirby. J. comp. Physiol. Tl: 49-64. . 1972c. Patterns of endothermy in bumblebee queens, drones, and workers. J. comp. Physiol. 77: 65-79. . 1972d. Physiology of brood incubation in the bumblebee queen, Bombus vosnesenski. Nature 239: 223-225. . 1974. Thermoregulation in endothermic insects. Science 185: 747-756. Henson, W.R. 1958a. Some ecological implications of the leaf-rolling habit in Compolechnia niveopulvella Chamb. Can. J. Zool. 36: 809-818. . 1958b. The effects of radiation on the habitat temperatures of some poplar-inhabiting insects. Can. J. Zool. 36: 463-478. Henson, W.R., and R.F. Shepherd. 1952. The effects of radiation on the habitat temperatures of the lodgepole needle miner, Recurvaria milleri Busck (Gelechiidae: Lepidoptera). Can. J. Zool. 30: 144-153. Hodges, R.W. 1974. The moths of America north of Mexico. Fascicle 6.2.  Gelechioidea Oecophoridae. E.W. Classe Ltd. and R.B.D. Publ. Inc., London. 142 pp. 114 Hodson, A . C , and M.A. Al Rawy. 1958. Temperature in relation to developmental thresholds in insects. Proc. 10th Int. Congr. Ent. 2: 61-65. Howell, T.A., D.W. Meek, and J.L. Hatfield. 1983. Relationship of photosynthetically active radiation to shortwave radiation in the San Joaquin Valley. Agric. Meteorol. 28: 157-175. Huffaker, C.B. 1944. The temperature relations of the immature stages of the malarial mosquito, Anopheles quadrineaculatus Say, with a comparison of the developmental power of constant and variable temperatures in insect metabolism. Ann. ent. Soc. Am. 37(1): 1-27. Inoue, H. 1986. Habitat use by the refuging predator Agriosphodrus dohrni Signoret. I. Nymphal microhabitat suitability and density dependent microhabitat selection by ovipositing females. Res. Popul. Ecol. 28(2): 321-332. Johnson, G.G. 1940. Development, hatching and mortality of the eggs of Cimex lectularius L. (Hemiptera) in relation to climate, with observations on the effect of preconditioning to temperature. Parasitology 32: 127-173. Kershaw, K.A. 1985. Physiological ecology of lichens. Cambridge University Press, Cambridge. 293 pp. Kingsolver, J.G. 1985a. Thermal ecology of Pieris butterflies (Lepidoptera: Pieridae): a new mechanism of behavioural thermoregulation. Oecologia 66: 540-545. . 1985b. Thermoregulatory significance of wing melanization in Pieris butterflies (Lepidoptera: Pieridae): physics, posture and pattern. Oecologia (Berlin) 66: 546-553. Krenn, K.D. 1933. Bestrahlungsverhaltn. stehender u. liegender Stamme. Wien. Allg. F. u. Jagdztg. 51: 50-51, 53-54. (Cited in Geiger 1950). Lamb, R.J., and G.H. Gerber. 1985. Effects of temperature on development, growth and survival of larvae and pupae of a north-temperate chrysomelid beetle. Oecologia 67: 8-18. Lewis, T. 1962. The effects of temperature and relative humidity on mortality in Limothrips cerealium Halidaj' (Thysanoptera) overwintering in bark. Ann. appl. Biol. 50: 313-326. Lin, S., A.C. Hodson, and A.G. Richards. 1954. An analysis of threshold temperatures for the development of Oncopeltus and Triboilum eggs. Physiol. Zool. 27(4): 287-311. 115 Marino, P.C. 1986. Activity patterns and microhabitat selection in a desert Tenebrionid beetle (Coleoptera: Tenebrionidae). Ann. ent. Soc. Am. 79: 468-471. May, M.L. 1976a. Warming rates as a function of body size in periodic endotherms. J. comp. Physiol. I l l : 55-70. . 1976b. Thermoregulation and adaptation to temperature in dragonflies (Odonata: Anisoptera). Ecol. Monogr. 46(1): 1-32. . 1979. Insect thermoregulation. Ann. Rev. Ent. 24: 313-349. McNab, B.K. 1970. Body weight and the energetics of temperature regulation. J. exp. Biol. 53: 329-348. Messenger, P.S. 1964. The influence of rhythmically fluctuating temperatures on the development and reproduction of the spotted alfalfa aphid, Therioaphis maculata. J. econ. Ent. 57(1): 71-76. . 1969. Bioclimatic studies of the aphid parasite Praon exsoletum. 2. Thermal limits to development and effects of temperature on rate of development and occurrence of diapause. Ann. ent. Soc. Am. 62: 1026-1031. Messenger, P.S., and N.E. Flitters. 1958. Effect of constant temperatures on the egg stage of three species of fruit flies in Hawaii. Ann. ent. Soc. Am. 51(2): 756-765. . 1959. Effect of variable temperature environments on egg development of three species of fruit flies. Ann. ent. Soc. Am. 52: 191-204. Meyrick, E. 1927. A revised handbook of British Lepidoptera. Wakins Doncaster, London. 672 pp. Neumann, D., and F. Heimbach. 1975. Das Wachstum des Kohlwei flings bei konstanten und tagesperiodisch wechselnden Temperaturen [in German, English summary]. Oecologia 20: 135-141. Parry, D.A. 1951. Factors determining the temperature of terrestrial arthropods in sunlight. J. exp. Biol. 28: 445-462. Peairs, L.M. 1927. Some phases of the relation of temperature to the development of insects. W. Va. Univ. Agr. expt. Sta. Bull. 208: 1-62. Peterson, A., and G.J. Hauessler. 1928. Determination of the spring-brood emergence of oriental peach-moth and codling moth by various methods. J. Agric. Res. 37: 399-417. 116 Peterson, D.G., and L.S. Wolfe. 1958. The biology and control of black flies (Diptera: Simuliidae) in Canada. Proc. 10th Int. Congr. Entomol. 3: 551-564. Raine, J. 1965. Control of Dasystoma salicellum, a new pest of blueberries in British Columbia. Can. J. Plant Sci. 4 5 : 243-245. . 1966. Life history of Dasystoma salicellum Hbn. (Lepidoptera: Oecophoridae) a new pest of blueberries in British Columbia. Can. Ent. 98: 331-334. Rausher, M.D. 1979. Larval habitat suitability and oviposition preference in three related butterflies. Ecology 60(3): 503-511. Reichert, A. 1932. Rosenschadlinge (Rose Pests) Die kranke Pflanze, IX, nos. 5-4. 8: 75-78. Richards, A.G. 1959. Studies on temperature thresholds in insect development. Biol. Zentrabl. 78: 308-314. Richards, A.G., and S. Suanraksa. 1962. Energy expenditure during embryonic development under constant versus variable temperatures (Oncopeltus fasciatus (Dallas)). Entomol. Exp. Appl. 5 : 167-178. Rock, G.C. 1985. Thermal and thermoperiodic effects on larval and pupal development and survival in the tufted apple bud moth (Lepidoptera: Tortricidae). Environ. Entomol. 14: 637-640. Ruesink, W.G. 1976. Status of the systems approach to pest management. Annu. Rev. Ent. 21: 21-44. SAS Institute. 1985. SAS user's guide: statistics, version 5 edition. SAS Institute, Cary, N.C. 956 pp. Sharpe, P.J.H., and D.W. DeMichele. 1977. Reaction kinetics of poikilotherm development. J. theor. Biol. 64: 649-670. Shelly, T.E., and D. Ludwig. 1985. Thermoregulatory behaviour of the butterfly Calisto nubila (Satyridae) in a Puerto Rican forest. Oikos 44: 229-233. Sheppard, R.F., and G.R. Stairs. 1976. Factors affecting the survival of larval and pupal stages of the bagworm, Thyridopteryx ephemeraeformis (Lepidoptera: Psychidae). Can. Ent. 108(5): 469-473. Siddiqui, W.H., CA. Barlow, and P.A. Randolph. 1973. Effects of some constant and alternating temperatures on population growth of the pea aphid, Acyrtosiphon pisum (Homoptera: Aphididae). Can. Ent. 105: 145-156. 117 Shreeve, T.G. 1984. Habitat selection, mate location, and microclimatic constraints on the activity of the wood butterfly Pararge ageria. Oikos 42: 371-377. Sullivan, C.R., and W.G. Wellington. 1953. The light reactions of larvae of the tent caterpillar, Malacosoma disstria Hbn., M. americanum (Fab.) and M. pluviale (Dyar) (Lepidoptera: Lasiocampidae). Con. Ent. 85: 297-310. Uvarov, B.P. 1931. Insects and climate. Trans, ent. Soc. London 79: 1-232. Vielmetter, W. 1958. Physiologie des Verhaltens zur Sonnenstrahlung bei dem Tagfalter Argynnis paphia L. I. Untersuchungen im Freiland. J. Insect Physiol. 2: 13-37. Waloff, Z. 1963. Field studies on solitary and transient desert locusts in the Red Sea area. Anti-Locust Bull. 40: 1-93. Welbers, P. 1975. Der Einflu/3 von tagesperiodischen Wechseltemperaturen bei der Motte Pectinophora I. Entwicklungsdauer, Larvengewicht und Reproduktionsrate [in German, English, summary]. Oecologia 21: 31-42. Wellington, W.G. 1950. Effects of radiation on the temperatures of insectan habitats. Sci. Agr. 30: 209-234. . 1959. Individual differences in larvae and egg masses of the western tent caterpillar. Can. Dep. Agric. For. Biol Div. Bi-m. Prog. Rep. 15: 3-4. . 1965. Some maternal influences on progeny quality in the western tent caterpillar, Malacosoma pluviale (Dyar). Can. Ent. 97: 1-14. . 1977. Returning the insect to insect ecology: some consequences for pest management. Environ. Ent. 6(1): 1-8. Wigglesworth, V.B. 1972. The principles of insect physiology. 7th ed. Chapman and Hall Publ., London. 827 pp. Williams, E.H. 1981. Thermal influences on oviposition in the montane butterfly Euphydryas gilletii. Oecologia 50: 342-346. Willmer, P.G. 1982. Microclimate and the environmental physiology of insects. Advan. in Insect Physiol. 16: 1-57. Willmer, P.G., and D.M. Unwin. 1981. Field analyses of insect heat budgets: reflectance, size and heating rates. Oecologia 50: 250-255. Yeargan, K.V. 1980. Effects of temperature on development rate of Telenomus podisi (Hymenoptera: Scelionidae). Ann. ent. Soc. Am. 73: 339-342. APPENDIX I Correlation analysis between T e x c e s s (difference between air and shelter temperatures) and levels of incoming solar radiation recorded for the leaf shelters under sunny conditions in a high-bush blueberry field in Richmond, B.C. in 1985. SOURCE OF VARIATION df SS F RADIATION 1 4773.99 401.12 *** 0.415 ERROR 563 6700.61 TOTAL 564 11474.59 ***, P<0.0001. 118 APPENDIX II Correlation analysis between T e x c e s s (difference between air and shelter temperatures) and levels of incoming solar radiation recorded for the leaf shelters under cloudy conditions in a high-bush blueberry field in Richmond, B.C. in 1985. SOURCE OF VARIATION df SS F RADIATION 1 230.52 60.43 *** 0.310 ERROR 129 492.13 TOTAL • 130 722.64 ***, P<0.0001. 119 A P P E N D I X III Analysis of variance of daily maximum T e x c e s s (difference between air and shelter temperatures) under different weather conditions (sunny, cloudy) between Aug. 3 and Aug. 17, 1985 in a high-bush blueberry field in Richmond, B.C. SOURCE OF VARIATION df SS F WEATHER CONDITIONS 1 241.87 16.20 ** ERROR 49 731.36 TOTAL 50 973.23 **, P<0.0005. 120 A P P E N D I X I V Analysis of variance of the length of development of male and female early instar Cheimophila salicella larvae reared in 3 temperature regimes. SOURCE OF VARIATION df SS F TEMPERATURE REGIME 261.93 16.39 *** EGG MASS WITHIN TEMPERATURE REGIME 40 319.63 1.67 * SEX 0.59 0.12 ns TEMPERATURE X SEX 24.23 2.53 ns ERROR 122 584.50 TOTAL 167 1191.79 ***, P<0.0001; *, P<0.05. 121 APPENDIX V Analysis of variance of the length of development of male and female middle instar Cheimophila salicella larvae reared in 3 temperature regimes. SOURCE OF VARIATION df SS F TEMPERATURE REGIME 2 2044.60 131.94 *** EGG MASS WITHIN TEMPERATURE REGIME 39 302.17 0.70 ns SEX 1 83.91 7.54 * TEMPERATURE X SEX 2 4.46 0.20 ns ERROR 110 1223.86 TOTAL 154 3607.42 ***, P<0.0001; *, P<0.05. 122 A P P E N D I X V I Analysis of variance of the length of development of male and female late instar Cheimophila salicella larvae reared in 3 temperature regimes. SOURCE OF VARIATION df SS F TEMPERATURE REGIME 2 3013.85 36.30 *** EGG MASS WITHIN TEMPERATURE REGIME 40 1660.72 1.18 ns SEX 1 1803.29 51.26 *** TEMPERATURE X SEX 2 162.19 2.31 ns ERROR 111 3904.78 TOTAL 156 11432.33 P<0.0001. 123 APPENDIX VII Analysis of variance of the length of development of male and female Cheimophila salicella larvae reared in 3 temperature regimes. SOURCE OF VARIATION df SS F TEMPERATURE REGIME 2 2246.16 20.96 *** EGG MASS WITHIN TEMPERATURE REGIME 41 2196.47 1.36 ns SEX 1 1174.72 29.92 *** TEMPERATURE X SEX 2 56.70 0.72 ns ERROR 123 4829.36 TOTAL 169 11743.51 ***, P<0.0001. 124 APPENDIX VIII Analysis of variance of the weight of 14 day-old male and female Cheimophila salicella pupae from larvae reared in 3 temperature regimes. SOURCE OF VARIATION df SS TEMPERATURE REGIME 1433.33 6.16 EGG MASS WITHIN TEMPERATURE REGIME 41 4767.31 1.65 SEX 4406.75 62.41 TEMPERATURE X SEX 149.17 1.06 ns ERROR 124 8756.13 TOTAL 170 21933.13 ***, P<0.0001; **, P<0.005; *, P<0.05. 125 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items