Feeding, Growth, and the Thermal Environment of Cabbage White Caterpillars,Pieris rapaeL

Feeding, Growth, and the Thermal Environment of Cabbage White Caterpillars, Pieris rapae L.Author(s): Joel G. KingsolverSource: Physiological and Biochemical Zoology, Vol. 73, No. 5 (September/October 2000), pp.621-628Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/10.1086/317758 .

Accessed: 16/10/2013 23:51

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access toPhysiological and Biochemical Zoology.

http://www.jstor.org

This content downloaded from 128.197.26.12 on Wed, 16 Oct 2013 23:51:09 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=ucpress

http://www.jstor.org/stable/10.1086/317758?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


621

Feeding, Growth, and the Thermal Environment of Cabbage White

Caterpillars, Pieris rapae L.

Joel G. Kingsolver*

Department of Zoology, Box 351800, University ofWashington, Seattle, Washington 98195

Accepted 6/19/00

ABSTRACT

Laboratory studies of temperature effects on short-term feedingand growth rates were combined with field data on thermalenvironments to explore the consequences of temperature var-iation for growth of caterpillars of the cabbage white butterfly,Pieris rapae. Mean short-term (24-h) consumption and growthrates of fourth-instar P. rapae feeding on collard leaves increasedcontinuously with increasing temperatures between 107 and357C, peaked at 357C, and declined rapidly with temperaturesabove 357C. Physical models can mimic temperatures of realfifth-instar caterpillars under collard leaves within 17–27C insunny summer conditions in Seattle, Washington. Continuousrecordings of operative temperatures of model caterpillars in acollard garden suggest that, at the timescale of the duration ofthe fifth instar (5–8 d in the field), P. rapae caterpillars fre-quently experience temperatures spanning a 257C range, theyspend most of their time at temperatures well below those thatmaximize growth, and they encounter substantial variation inthe frequency distribution of operative temperatures betweentime periods. Combining these data on growth rate as a func-tion of temperature and the distribution of operative temper-atures in the field, I illustrate how growth rates at higher tem-peratures can make disproportionate contributions to theoverall mean growth rates even when higher temperatures arerelatively infrequent. Fluctuating thermal conditions may gen-erate variable patterns of selection on reaction norms forgrowth rate in the field.

Introduction

The feeding and growth rates of caterpillars and other ecto-thermic animals depend strongly on their body temperatures.

*Present address: Department of Biology, CB3280, University of North Carolina,

Chapel Hill, North Carolina 27599; e-mail: [email protected].

Physiological and Biochemical Zoology 73(5):621–628. 2000. q 2000 by TheUniversity of Chicago. All rights reserved. 1522-2152/2000/7305-99139$03.00

In the face of variable thermal conditions in the field, somegregarious or unpalatable caterpillar species actively regulatetheir body temperatures by behavioral postures or movements(Casey 1993). For example, in the tent caterpillar Malacosomaamericanum, the tent provides a variety of available microhab-itats during the day that allow caterpillars to actively regulatetheir body temperatures quite precisely (Joos et al. 1988). Othercaterpillars bask to elevate body temperatures during cool con-ditions, sometimes in aggregations that reduce convective heatloss and increase body temperatures (Casey 1976; Rawlins andLederhouse 1981; Kevan et al. 1982; Porter 1982; Kukal et al.1988; Stamp and Bowers 1990). These behaviors can reducethe range or increase the average temperatures experienced byforaging caterpillars and thereby increase overall feeding andgrowth rates (Casey 1993).

However, most caterpillars are palatable and cryptic and donot actively thermoregulate; they are “thermo-conformers”(Casey 1993). In such species, the thermal conditions in whichfeeding and growth occur are largely determined by weatherand the microclimate of the plants on which they live. Forexample, Manduca sexta caterpillars are cryptically colored,avoid solar radiation, and feed on the undersides of the leavesof their host plants. These caterpillars may experience diurnalfluctuations in body temperatures of more than 307C (Casey1976). Many studies have explored how the ranges of temper-atures experienced in the field may relate to the thermal sen-sitivity of growth and feeding (Casey 1993), but quantitativepatterns of temporal and spatial variation in body temperatureexperienced within the life span of individual caterpillars arelargely unknown. Such information is needed for any quan-titative understanding of how temperature variation relates tovariation in growth rates in caterpillars and how environmentalvariation may generate selection on the thermal sensitivity ofgrowth and feeding in the field (Kingsolver and Huey 1998).

This study examines the thermal ecology of growth and feed-ing in a thermo-conforming caterpillar, the cabbage white Pierisrapae L. (Lepidoptera: Pieridae). First, the effects of tempera-ture on short-term consumption and growth rates of cabbagewhite caterpillars feeding on collard leaves in the lab were de-termined. Second, temporal and spatial variation in operativetemperatures of caterpillars was quantified. By combining thesedata on thermal dependence of growth rate and the frequencydistribution of temperatures in the field, I explore the relativeimportance of growth at different temperatures for total growthduring the fifth instar. My results suggest that the frequency of



622 J. G. Kingsolver

higher temperatures is the most important determinant of var-iation in total growth in field conditions.

Material and Methods

The Study System

The small cabbage white, Pieris rapae L., is native to Europebut is now found throughout North America and other con-tinents. Its larval host plants include a wide variety of wild anddomesticated species of Brassica and other genera in the mus-tard family. As its common name implies, the cabbage whiteis an agricultural pest on cabbage, broccoli, and other domes-ticated forms of Brassica oleracea L. Pieris rapae is one of themost abundant butterflies in the Puget Sound region of westernWashington (U.S.A.), where it may have four to five generationsper year, with adult flight seasons extending from April intoSeptember.

Cabbage white adult females lay their eggs singly on leaves,laying preferentially on host plants in open, sunny areas (Jones1977a; Ohsaki 1979). The caterpillars are cryptic green andtypically rest near the leaf axis on the undersides of leaves,generally moving to the leaf margins to feed (Jones 1977b;Mauricio and Bowers 1990); there is no evidence that thesecaterpillars actively regulate their body temperatures (M. F.Wehling and J. G. Kingsolver, unpublished data). Cabbage whitecaterpillars rarely leave a single plant unless it dies, senesces,or is defoliated (Jones 1977b). Lab studies using constant tem-perature conditions show that larval development rate increasescontinuously with increasing temperatures between 157 and307C (Jones and Ives 1979; Gilbert 1984b, 1988; Jones et al.1987b), with similar effects of temperature in each of the fivelarval instars (Jones and Ives 1979; Chen and Su 1982). Thecaterpillars feed and grow at similar rates in both light anddark at a given temperature (Neumann and Heimbach 1975).In the field, cabbage white caterpillars are often strongly affectedby natural enemies, including insect parasitoids and social wasppredators (Dempster 1967; Van Driesche 1988; Biever 1992).As a result, there may be strong selection for rapid growth anddevelopment rates to decrease the time caterpillars are exposedto enemies. For example, field experiments with cabbage whitecaterpillars feeding on cultivated B. oleracea showed that slowergrowth was correlated with higher mortality due to the para-sitoid Cotesia glomerata (Benrey and Denno 1997).

Lab Studies: Consumption and Growth Rates

In the studies described here, wild-caught, mated female P.rapae butterflies were collected from organic farms and urbangardens in the Seattle area. Females were fed twice daily with25% honey water, placed in individual cages, and allowed tooviposit on young collard leaves (B. oleracea var. collards); allcollards were grown in the greenhouse using a single varietydeveloped for use in the Pacific Northwest (Champion collards

C0350, Territorial Seed Co.). Females typically mate a singletime early in adult life, and thus eggs from a single femalerepresent a group of full sibs (Rutowski and Gilchrist 1986).After hatching, caterpillars were reared in petri dishes on young,greenhouse-grown collard leaves; the leaves were sprayed withantibiotic 2 d before feeding to reduce bacterial infections inthe caterpillars (Riddiford 1967). The caterpillars were reareduntil molting into the fourth instar (or fifth instar, dependingon the experiment) in a walk-in environmental chamber whosediurnal temperature cycle varied between 107 and 307C with a16L : 8D photocycle, to mimic the diurnal temporal patternand range of temperatures typically experienced by caterpillarsin the field in midsummer (see “Results”).

Standard gravimetric methods were used to measure short-term (24-h) wet-weight consumption and growth rates at aseries of temperatures between 107 and 457C (Waldbauer 1968;Harrison and Fewell 1995; Kingsolver and Woods 1997). I be-lieve that wet weights more realistically reflect the short-termgrowth and feeding responses of caterpillars. Within 24 h ofmolting into the fourth instar, each test caterpillar was accli-mated in a petri dish with young collard leaves at the testtemperature for 1 h. The caterpillar was then weighed (50.1mg), put in a petri dish with a damp sponge and with a freshgreenhouse-grown collard leaf whose mass has been deter-mined, and placed in an environmental chamber. After the 24-h test period, the caterpillar and remaining leaf were reweighed.

The results of two studies of consumption and growth ratesare presented here. First, (wet-weight) mass-specific consump-tion rates were measured for sets of newly molted fourth-instarcaterpillars over a 24-h test period at test temperatures between107 and 407C at 57C intervals. Fifteen caterpillars were measuredat each test temperature; each individual caterpillar was mea-sured at only a single temperature. Leaf wilting was assessedby measuring mass loss of five control leaves (without cater-pillars) at each test temperature, and these values were used tocorrect the estimates of consumption rate for mass loss due towilting. Second, mass-specific growth rates were measured forsets of newly molted fourth-instar caterpillars over a 24-h testperiod at test temperatures between 107 and 407C at 57C in-tervals. Fifteen to 20 caterpillars were measured at each testtemperature; each individual caterpillar was measured at onlya single temperature. This study was repeated for fifth instarcaterpillars.

Field Studies: Caterpillar Operative Temperatures

Physical models of caterpillars were used to model operativetemperatures of caterpillars in the field (Bakken et al. 1985;Grant and Dunham 1988; Grant and Porter 1992). The modelswere cylindrical, the approximate length and diameter of anaverage fifth-instar cabbage white caterpillar, and made froma moldable epoxy whose thermal conductivity is similar to thatof water. The models were painted with acrylic paint (leaf



Caterpillar Growth and the Thermal Environment 623

Figure 1. Mean (51 SD) mass-specific consumption rates (g/g/h) over24 h of fourth-instar Pieris rapae caterpillars as a function oftemperature.

Figure 2. Mean (51 SD) mass-specific growth rates (g/g/h) over 24h of fourth- (solid line) and fifth-instar (dashed line) Pieris rapae cat-erpillars as functions of temperature. An asterisk indicates that mor-tality occurred in some caterpillars at this temperature; only caterpillarsthat fed and were alive at the end of the test period were included inthe growth rate estimates.

green), and a 0.5-mm-diameter copper-constantan thermo-couple was imbedded in the center of each model along itslong axis.

To determine whether the models adequately mimic the bodytemperatures of real cabbage white caterpillars in the field, Imeasured temperatures of real and model caterpillars on collardplants in the field during one sunny summer day (0900–1700hours PDT [Pacific daylight time], September 9, 1997) in Se-attle, Washington. Each model was placed in contact with theunderside of a collard leaf near the leaf axis and held in placeby wrapping the model’s thermocouple wire to the leaf petiole.One model was placed on a leaf on each of six mature collardplants outside the University of Washington greenhouse. Model(operative) temperatures (Top) were monitored continuously at1-min intervals throughout the day with a data logger (Camp-bell 21XL). Between 0900 and 0930 hours PDT, a fifth-instarcaterpillar (50–100 mg mass) was placed on the underside ofan adjacent leaf on each of the six plants. Every 30–45 minbetween 0950 and 1710 hours PDT, the body temperatures oftwo or three randomly chosen caterpillars were measured bygrabbing, pressing, and wrapping each caterpillar around a fine(0.076-mm-diameter) copper-constantan thermocouple; bodytemperature was measured within 5 s of grabbing with a ther-mocouple thermometer (Wescor TH-65; Stone and Willmer1989). A new caterpillar was then placed on the plant.

Temporal and spatial variation in caterpillar operative tem-peratures in the field was evaluated in a collard garden (Cham-pion collards #C0350, Territorial Seed Co.) at the Center forUrban Horticulture at the University of Washington, Seattle.The garden consisted of 10 rows, 1 m apart and 15 m in length,with the rows oriented north-south. Two-week-old seedlingsthat had been germinated in the greenhouse were planted inthe garden on June 9, 1997; after thinning, the plants were

spaced approximately 1 m apart. Starting in late July, 10 modelcaterpillars were attached to the shady undersides of leaves inthe garden. A single model was placed on any one plant; themodels were distributed among three to four plants in threedifferent rows, including both outer (older) and inner leaves.The mean temperature of each model was output at 5-minintervals to a data logger (Campbell 21XL). Model temperatureswere monitored continuously for three different 6–8-d periodsin 1997: July 25–30, August 12–19, and August 21–26.

Because the duration of the fifth instar in the field underthese conditions is approximately 5–9 d (J. G. Kingsolver, un-published data), these measurements characterize the variationin thermal conditions experienced by cabbage white caterpillarsover the course of the fifth instar. I used these recordings ofmodel temperatures measured at 5-min intervals to estimatethe frequency distribution of the (spatially averaged) operativetemperatures, f(Top), for each of the three time periods, binningdata at 17C intervals.

Results

Consumption and Growth Rates

Mean mass-specific consumption rates for fifth-instar Pierisrapae caterpillars increased with increasing temperatures from107 to 357C, peaked at 357C, and declined rapidly above 357C(Fig. 1). Similarly, mean mass-specific growth rates for fourthand fifth instar caterpillars increased continuously for temper-atures from 107 to 357C and declined rapidly above 357C (Fig.2). There was substantial mortality at 407 and 457C in theseexperiments; for example, 42% of the fifth-instar caterpillarsand 20% of the fourth-instar caterpillars were dead at the endof 24 h at 407C. (Note that the growth and consumption rates




Figure 4. Operative temperatures of 10 model caterpillars as a functionof time in a collard garden in Seattle, Washington, August 21–26, 1997.

Figure 3. Relation between temperatures of model caterpillars and ofreal fifth-instar Pieris rapae caterpillars under collard leaves on Sep-tember 9, 1997, in Seattle, Washington.

reported are only for those caterpillars that fed and were aliveat the end of the test period.) Mean mass-specific growth rateswere similar at 107–207C and slightly greater at 257–357C forfifth-instar compared with fourth-instar caterpillars, but thesedifferences were not significant.

The Thermal Environment

Field measurements comparing temperatures of physical mod-els (operative temperatures) with those of real fifth-instar P.rapae caterpillars during a sunny summer day (Fig. 3) showeda strong correspondence between model and real caterpillartemperatures ( , pairs). Caterpillar and2r p 0.865 N p 26model temperatures were generally within 17C and alwayswithin 27C during these measurements. One would expect evengreater similarity between caterpillar and model temperaturesunder cloudy and nighttime conditions when solar radiativeheating is reduced. These results suggest that operative tem-peratures of physical models may provide a useful, standardizedway of assaying caterpillar temperatures for P. rapae in the field,at least in north temperate areas such as Seattle.

Recordings of operative temperatures of 10 physical modelsunder collard leaves in a collard garden in Seattle over a 6-dperiod in August 1997 suggest several interesting patterns (Fig.4). First, diurnal variation experienced by the models was ex-tensive, with a daily range of more than 207C on sunny dayswith clear nights. As a result, daily average temperatures providelittle information about the conditions experienced by the cat-erpillars. Second, there was substantial day-to-day variation inboth maximum and minimum operative temperatures. Third,

most models experienced similar temperatures throughout thenight and day except during sunny, midday conditions: duringthese conditions (when operative temperatures were highest)models on different leaves and plants sometimes differed by37C or more.

The 6-d time period illustrated in Figure 4 represents a time-scale similar to the duration of the fifth larval instar of P. rapaein the field under these conditions. One can use these data toestimate the frequency distribution of operative temperaturesexperienced in the field during this time period (Fig. 5). Forexample, for August 21–26, 1997 (the time period illustratedin Fig. 5), the mode of the frequency distribution of operativetemperatures was 157–167C (Fig. 6, solid line); the median op-erative temperature during this time period was only 18.17C,and only 14% of the period was at temperatures above 257C.By contrast, during the previous week (August 12–19, 1997),median operative temperature was somewhat higher (20.67C),but nearly 35% of the period was at temperatures above 257C.During the July 25–30, 1997, time period, there were two majormodes in the operative temperature distribution at 167 and297C; median operative temperature was 19.57C, and 30% ofthe period experienced temperatures above 257C. Thus, at thetimescale of the fifth larval instar of P. rapae, median operativetemperatures in the field varied rather little, but maximumtemperatures and the frequency of higher temperatures variedsubstantially between time periods.

Discussion

Temperature Effects on Feeding, Growth, and Development

As in other thermoconforming caterpillars, cabbage whites canfeed and grow effectively over a wide range of temperatures(Casey 1993). Short-term consumption and growth rates in-crease six- to eightfold as temperature increases from 107 to357C (indicating ), peak near 357C, and decline rapidlyQ ! 210




Figure 5. Frequency distributions of mean operative temperatures ofmodel caterpillars for three time periods in a collard garden in Seattle,Washington.

Figure 6. Weighted mass-specific growth rate [GT(T)] as a function oftemperature for fifth-instar Pieris rapae caterpillars during one weekfor three time periods in a collard garden in Seattle, Washington. Seetext.

at higher temperatures (Figs. 1, 2). Short-term consumptionand growth rates in thermoconforming Manduca sexta cater-pillars show a very similar pattern of thermal dependence, withQ10 values less than 2 for temperatures between 147 and 347C(Casey 1976; Reynolds and Nottingham 1985; Kingsolver andWoods 1997). These relatively low values reflect the fact thatthermoconforming herbivores like Pieris rapae must be able tosustain growth over the wide and fluctuating range of bodytemperatures they will routinely experience in the field (Casey1993). By contrast, caterpillars and grasshoppers that activelyregulate body temperature may have Q10 for feeding and growthrate exceeding 4, reflecting the capacity of these insects toachieve higher mean body temperatures via thermoregulation(Casey and Knapp 1987; Harrison and Fewell 1995).

Gilbert, Jones, and others have carefully documented theeffects of temperature on larval development rates and pupalmass of P. rapae from a number of geographic populationsaround the world (Gilbert 1984a, 1984b, 1984c; Jones et al.1987a, 1987b; Gilbert and Coaker 1988; Gilbert and Raworth1996). These studies show that under constant temperatureconditions, developmental rate (measured as the inverse of theduration of the larval stage or instar) increases approximatelylinearly with temperatures between 16.57 and 30.57C and thatthe relation of temperature to development rate is similar foreach of the five larval instars. These studies also show thatdevelopment rate is maximal at 30.57C under constant tem-perature conditions, with substantial mortality at higher tem-peratures. By contrast, the present results indicate that short-term (24-h) growth rates are maximized at 357C—temperaturesthat under constant conditions cause high mortality rates (Chenand Su 1982); there was no mortality during our short-termmeasurements except at temperatures of 407C and above. Sim-ilar differences in the thermal sensitivity of growth and devel-

opment at different timescales are also seen in M. sexta cat-erpillars (Reynolds and Nottingham 1985; Kingsolver andWoods 1997). Note that these differences in growth rates atdifferent timescales cannot be accounted for simply by the non-linear effects of temperature (Worner 1992). These differencescould result from different thermal sensitivities of growth (bodymass increase) as opposed to molting (Stamp and Horwath1992), or from transient effects of temperature on growth rates.This issue of timescale is of interest because P. rapae routinelyexperience rapid temperature changes on the scale of minutesto hours under field conditions (Fig. 4; see below).

The Thermal Environment of Caterpillars

Physical models have been used widely to characterize the op-erative temperatures available to animals in variable thermalenvironments (Bakken et al. 1985; Grant and Porter 1992). Forexample, physical models of caterpillars (Joos et al. 1988) andlizards (Grant and Dunham 1988) have been used to explorethe consequences of behavioral positioning and microhabitatchoice for regulating body temperatures and for patterns ofgrowth. For relatively sedentary, cryptic, and thermoconform-ing animals such as P. rapae caterpillars, physical models maybe used to quantify the temporal and spatial patterns of tem-perature variation that occur within the life of a caterpillar.Comparisons of real and model fifth-instar P. rapae caterpillarssuggest that models can generally mimic temperatures of realcaterpillars on the undersides of collard leaves within 17–27Cduring sunny, summer conditions in the field (Fig. 3). Onewould expect even better agreement under cloudy, nighttime,and other conditions in which solar radiative heating is reduced.

Measurements of operative temperatures of caterpillar mod-




els in a collard garden in Seattle clearly illustrate the variabilityof thermal environments experienced by caterpillars over thetimescale of a single larval instar (Fig. 4). Diurnal variation inoperative temperatures frequently span 207C. One model re-corded operative temperatures ranging from below 107C toabove 357C during a single 14-h period, thus spanning theentire body temperature range allowing effective growth in P.rapae (Fig. 2). These data also may provide the first documentedevidence of sunshine in the Seattle area (anonymous reviewer,personal communication).

When data on operative caterpillar temperatures are viewedas frequency distributions, several interesting patterns emerge(Fig. 5). First, at the timescale of a larval instar, there is con-siderable variation among time periods in the relative frequencyof high temperatures. For example, the frequency of time pe-riods with operative temperature between 317 and 367C, atwhich growth rates are highest (Fig. 2), ranges from 1.2% to17.4% among the three time periods measured. Second, basedon the operative temperature measurements, much of a cat-erpillar’s time is spent at cooler temperatures below 157C, wheregrowth rates are quite low. Third, “average” temperature con-ditions vary rather little—for example, median operative tem-peratures ranged between 18.17 and 20.67C for the three timeperiods measured—and in fact operative temperatures near themedians are relatively infrequent. It is useful to note that, ifoperative temperatures vary sinusoidally over a daily cycle, thenthe frequency of operative temperatures must necessarily havea bimodal distribution in which intermediate temperatures areinfrequent. Clearly, average temperatures cannot usefully char-acterize the variation experienced by growing P. rapae cater-pillars in field conditions, even in the moderate conditions thatprevail in western Washington.

Growth Rates in Natural Environments

How does extensive variation in the natural thermal environ-ment affect the patterns of growth for caterpillars in the field?Field growth rates will reflect both the distribution of temper-atures experienced in the field and the effects of temperatureon short-term growth rate. To explore this, it is useful to definean environmentally weighted growth rate function, GT(T), asthe product of the frequency distribution of operative tem-peratures, f(T) (Fig. 5), and the thermal performance curve formass-specific growth rate, g(T) (Fig. 2):

G (T) p f(T) # g(T). (1)T

The product is a weighted average growth rate thatf(T) # g(T)represents the relative contribution to overall growth rate madewhile at a particular temperature T. Equation (1) ignores theeffects of transient changes in temperature on growth rate anddoes not consider allometric changes in growth rate with size;as a result, it cannot be used to predict quantitatively patterns

of growth over days or the duration of one or more larvalinstars. Using data on mass-specific growth rates for fifth-instarP. rapae (Fig. 2) to estimate g(T) and data on operative cat-erpillar temperatures in the field (Fig. 5) to estimate f(T), Icomputed the weighted growth rate function, GT(T), for eachof the three time periods shown in Figure 5. We used unpub-lished data on short-term growth rates at 57C to extend thisfunction to temperatures below 107C (see Figs. 2, 5, and 6).For purposes of comparison among time periods, I defined

for each time period. The results (Fig. 6) revealS[f(T)] p 1two useful insights. First, growth at higher temperatures cancontribute substantially to the overall mean growth rate, evenwhen higher temperatures are relatively infrequent. For ex-ample, growth rates at temperatures ≥287C contributed14%–50% of the overall growth rate for the three periods, eventhough such temperatures represented only 9%–28% of thetotal time during these periods. This skewing of contributionstoward higher temperatures simply reflects the much greaterrates of growth at higher temperatures. Second, there is sub-stantial variation among time periods in the relative contri-butions of different temperatures toward overall growth, es-pecially at high and at low temperatures.

These results have several interesting implications for un-derstanding thermal sensitivity of growth and selection on re-action norms in variable environments. First, growth rates athigher temperatures have a disproportionate effect on overallgrowth during some time period even when higher tempera-tures are relatively infrequent. If overall growth rates contributeto fitness, for example, by decreasing larval mortality or gen-eration time, one would expect selection on growth rate athigher temperatures will typically be stronger than on growthrate at lower temperatures (Gilchrist 1995). This occurs for thesame reason that selection may strongly favor the ability tothermoregulate to achieve high body temperatures unless thereare large physiological or fitness costs of thermoregulation(Huey and Slatkin 1976; Heinrich 1977; Huey and Kingsolver1989; Samietz 1997). Second, the opportunities for growth atdifferent temperatures, and the contributions of growth rate atdifferent temperatures toward overall growth, may vary con-siderably from one week to the next in the life of a caterpillaror population (Fig. 6). As a result, selection on the reactionnorms for growth rate at different temperatures may vary inintensity and direction over quite short timescales in the field.How selection on continuous reaction norms for growth andother physiological processes relates to patterns of environ-mental variation in the field remains largely unexplored (Go-mulkiewicz and Kirkpatrick 1992; Via et al. 1995).

Although these studies involve a widespread “pest” speciesfeeding on a domesticated cultivar, I believe that the qualitativeresults should apply rather generally to thermoconforming cat-erpillars feeding on herbaceous plants in temperate regions.First, the range and pattern of environmental temperature var-iation is dominated by the diurnal temperature cycle and day-




to-day weather variation, which will be similar for differentherbaceous plants in a given region. One important differenceis that spatial variation within and between plants is minimizedin a cultivar garden: one would expect greater spatial variabilityin thermal conditions, for example, in more natural vegetationtypes, expanding the degree of thermal variation experiencedby caterpillars. Second, P. rapae thrives in a wide range ofweather and climatic conditions throughout its large geographicrange (Gilbert 1984a, 1984b, 1984c). As a result, its responseto temperature may be more generalized than for more geo-graphically and seasonally restricted species that are more spe-cialized for particular temperature conditions (Jones et al.1987b). In this sense, the results presented here may under-estimate the magnitude of thermal environmental variation(Fig. 5) and its consequences for overall patterns of growth inthe field (Fig. 6) for thermoconforming caterpillars in general.

Acknowledgments

I thank Kristina Williams, Yvette Maylett, Gwen Schlichta, andMartha Wehling for help with the lab and field experimentsand Doug Ewing for his green thumb with collards. Ray Hueyand the members of KingHuey lab groups provided helpfuldiscussion; two anonymous reviewers provided useful com-ments on a previous draft of the manuscript. Research wassupported by National Science Foundation grants IBN 9419850and IBN 9818431 to J.G.K.

Literature Cited

Bakken G.S., W.R. Santer, and D.J. Erskine. 1985. Operativeand standard operative temperature: tools for thermal en-ergetics studies. Am Zool 25:933–943.

Benrey B. and R.F. Denno. 1997. The slow-growth-high-mor-tality hypothesis: a test using the cabbage butterfly. Ecology78:987–999.

Biever K.D. 1992. Distribution and occurrence of Cotesia ru-becula (Hymenoptera: Braconidae), a parasite of Artogeiarapae in Washington and Oregon. J Econ Entomol 85:739–742.

Casey T.M. 1976. Activity patterns, body temperature and ther-mal ecology in two desert caterpillars (Lepidoptera: Sphin-gidae). Ecology 57:485–497.

———. 1993. Effects of temperature on foraging of caterpillars.Pp. 5–28 in N.E. Stamp and T.M. Casey, eds. Caterpillars:Ecological and Evolutionary Constraints on Foraging. Chap-man & Hall, New York.

Casey T.M. and R. Knapp. 1987. Caterpillar thermal adaptation:behavioral differences reflect thermal sensitivities. CompBiochem Physiol 86A:679–682.

Chen C.N. and W.Y. Su. 1982. Influence of temperature ondevelopment and leaf consumption of three caterpillars on

cauliflower Artogeia rapae crucivora, Trichoplusia ni, Spodop-tera litura. Chung Hua Chih Wu Pao Hu Husueh Hui PlantProt Bull 24:131–141.

Dempster J.P. 1967. The control of Pieris rapae with DDT. I.The natural mortality of the young stages of Pieris. J ApplEcol 4:485–500.

Gilbert N. 1984a. Control of fecundity in Pieris rapae. I. Theproblem. J Anim Ecol 53:581–588.

———. 1984b. Control of fecundity in Pieris rapae. II. Dif-ferential effects of temperature. J Anim Ecol 53:589–597.

———. 1984c. Control of fecundity in Pieris rapae. III. Syn-thesis. J Anim Ecol 53:599–609.

———. 1988. Control of fecundity in Pieris rapae. V. Com-parisons between populations. J Anim Ecol 57:395–410.

Gilbert N. and T.H. Coaker. 1988. Differential survival of Britishand Mediterranean strains of Pieris rapae (L.) (Lepidoptera:Pieridae) on different Brassica cultivars. Bull Entomol Res78:669–671.

Gilbert N. and D.A. Raworth. 1996. Insects and temperature—ageneral theory. Can Entomol 128:1–13.

Gilchrist G.W. 1995. Specialists and generalists in changing en-vironments. I. Fitness landscapes of thermal sensitivity. AmNat 146:252–270.

Gomulkiewicz R. and M. Kirkpatrick. 1992. Quantitative ge-netics and the evolution of reaction norms. Evolution 46:390–411.

Grant B.W. and A.E. Dunham. 1988. Thermally imposed timeconstraints on the activity of the desert lizard Scloporus mer-riami. Ecology 69:167–176.

Grant B.W. and W.P. Porter. 1992. Modeling global macrocli-matic constraints on ectotherm energy budgets. Am Zool 32:154–178.

Harrison J.F. and J.H. Fewell. 1995. Thermal effects on feedingbehavior and net energy intake in a grasshopper experiencinglarge diurnal fluctuations in body temperature. Physiol Zool68:453–473.

Heinrich B. 1977. Why have some animals evolved to regulatea high body temperature? Am Nat 111:623–640.

Huey R.B. and J.G. Kingsolver. 1989. Evolution of thermalsensitivity of ectotherm performance. Trends Ecol Evol 4:131–135.

Huey R.B. and M. Slatkin. 1976. Cost and benefits of lizardthermoregulation. Q Rev Biol 51:363–384.

Jones R.E. 1977a. Movement patterns and egg distribution incabbage butterflies. J Anim Ecol 46:195–212.

———. 1977b. Search behaviour: a study of three caterpillarspecies. Behaviour 60:237–259.

Jones R.E. and P.M. Ives. 1979. The adaptiveness of searchingand host selection behaviour in Pieris rapae (L.). Aust J Ecol4:75–86.

Jones R.E., V.G. Nealis, P.M. Ives, and E. Scheermeyer. 1987a.Seasonal and spatial variation in juvenile survival of the cab-




bage butterfly Pieris rapae: evidence for patchy density-de-pendence. J Anim Ecol 56:723–737.

Jones R.E., J. Rienks, L. Wilson, C. Lokkers, and T. Churchill.1987b. Temperature, development and survival in monoph-agous and polyphagous tropical pierid butterflies. Aust J Zool35:235–246.

Joos B., T.M. Casey, T.D. Fitzgerald, and W.A. Buttemer. 1988.Roles of the tent in behavioral thermoregulation of easterntent caterpillars. Ecology 69:2004–2011.

Kevan P.G., T.S. Jensen, and J.D. Shorthouse. 1982. Body tem-peratures and behavioral thermoregulation of high arcticwooly-bear caterpillars and pupae (Gynaephora rosii, Ly-mantriidae: Lepidoptera) and the importance of sunshine.Arct Alp Res 14:125–136.

Kingsolver J.G. and R.B. Huey. 1998. Selection and evolutionof morphological and physiological plasticity in thermallyvarying environments. Am Zool 38:545–560.

Kingsolver J.G. and H.A. Woods. 1997. Thermal sensitivity offeeding and digestion in Manduca caterpillars. Physiol Zool70:631–638.

Kukal O., B. Heinrich, and J.G. Duman. 1988. Behavioral ther-moregulation in the freeze-tolerant arctic caterpillar Gynae-phora groenlandica. J Exp Biol 138:181–193.

Mauricio R. and M.D. Bowers. 1990. Do caterpillars dispersetheir damage? larval foraging behaviour of two specialist her-bivores, Euphydryas phaeton (Nymphalidae) and Pieris rapae(Pieridae). Ecol Entomol 15:153–161.

Neumann D. and F. Heimbach. 1975. Das Wachstum des Kohl-weißlings bie konstanten und tagesperiodisch wechselndenTemperaturen. Oecologia 20:135–141.

Ohsaki N. 1979. Comparative population studies of three Pierisbutterflies, P. rapae, P. melete and P. napi, living in the samearea. I. Ecological requirements for habitat resources in theadults. Res Popul Ecol 20:278–296.

Porter K. 1982. Basking behaviour in larvae of the butterflyEuphydras aurinia. Oikos 38:308–312.

Rawlins J.E. and R.C. Lederhouse. 1981. Developmental influ-ences of thermal behavior on monarch caterpillars (Danaus

plexippus): an adaptation for migration (Lepidoptera: Nym-phalidae: Danainae). J Kans Entomol Soc 54:387–408.

Reynolds S.E. and S.F. Nottingham. 1985. Effects of temperatureon growth and efficiency of food utilization in fifth instarcaterpillars of the tobacco hornworm, Manduca sexta. J In-sect Physiol 31:129–134.

Riddiford L.M. 1967. Antibiotics in the laboratory-rearing ofcecropia silkworms. Science 157:1451–1452.

Rutowski R.L. and G.W. Gilchrist. 1986. Copulation in Coliaseurytheme (Lepidoptera: Pieridae): patterns and frequency. JZool (Lond) 209:115–124.

Samietz J. 1997. Behavioral thermoregulation quantified: im-proved reproductive value and population consequences ina grasshopper. Adv Ethol 32:56–61.

Stamp N.E. and M.D. Bowers. 1990. Variation in food qualityand temperature constrain foraging of gregarious caterpillars.Ecology 71:1031–1039.

Stamp N.E. and K.L. Horwath. 1992. Interactive effects of tem-perature and concentration of the flavonol rutin on growth,molt and food utilization of Manduca sexta caterpillars. En-tomol Exp Appl 64:135–150.

Stone G.N. and P.G. Willmer. 1989. Endothermy and temper-ature regulation in bees: a critique of “grab and stab” mea-surement of body temperature. J Exp Biol 143:211–223.

Van Driesche R.G. 1988. Survivorship patterns of larvae of Pierisrapae (L.) (Lepidoptera: Pieridae) in Massachusetts kale, withspecial reference to mortality due to Apanteles glomeratus(L.) (Hymenoptera: Braconidae). Bull Entomol Res 78:397–405.

Via S., R. Gomulkiewicz, G. de Jong, C. Schlichting, S. Scheiner,and P. van Tienderen. 1995. Adaptive phenotypic plasticity:consensus and controversy. Trends Ecol Evol 10:212–217.

Waldbauer G.P. 1968. The consumption and utilization of foodby insects. Adv Insect Physiol 5:229–288.

Worner S.P. 1992. Performance of phenological models undervariable temperature regimes: consequences of the Kaufmannor rate summation effect. Environ Entomol 21:689–699.



Documents

Feeding, Growth, and the Thermal Environment of Cabbage White Caterpillars,Pieris rapaeL