Behavioral Ecology Vol. 15 No. 3: 419-425
Behavioral Ecology vol. 15 no. 3 © International Society for Behavioral Ecology 2004; all rights reserved
Knowing your habitat: linking patch-encounter rate and patch exploitation in parasitoids
Zoological Institute, Department of Animal Ecology, Christian-Albrechts-University, Am Botanischen Garten 1–9, D-24098 Kiel, Germany
Address correspondence to A. Thiel. E-mail: athiel{at}zoologie.uni-kiel.de.
Received 21 August 2002; revised 5 March 2003; accepted 26 June 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
According to optimal foraging theory, animals should decide whether or not to leave a resource patch by comparing the current profitability of the patch with the expected profitability of searching elsewhere in the habitat. Although there is abundant evidence in the literature that foragers in general are well able to estimate the value of a single resource patch, their decision making has rarely been investigated with respect to habitat quality. This is especially true for invertebrates. We have conducted experiments to test whether parasitic wasps adjust patch residence time and exploitation in relation to the abundance of patches within the environment. We used the braconid Asobara tabida, a parasitoid of Drosophila larvae, as our model species. Our experiments show that these wasps reduce both the residence time and the degree of patch exploitation when patches become abundant in their environment, as predicted by optimal foraging models. Based upon a detailed analysis of wasp foraging behavior, we discuss proximate mechanisms that might lead to the observed response. We suggest that parasitoids use a mechanism of sensitization and desensitization to chemicals associated with hosts and patches, in order to respond adaptively to the abundance of patches within their environment.
Key words: Asobara tabida, behavioral plasticity, habitat quality, marginal value theorem, optimal foraging, patch-time allocation, travel time.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Most habitats are spatially structured such that resources occur in discrete and depletable patches. Consequently, foraging animals frequently have to make decisions with regard to staying in or leaving their current patch. If foragers stay too long, they forgo the chance of faring better elsewhere, and if they abandon a patch too early, they waste time in traveling between patches. A classical model dealing with the problem of optimal patch-time allocation is the marginal value theorem (MVT) proposed by Charnov (1976)
. It shows that at least two parameters determine the quality of a habitat and thus the optimal patch-leaving decision: the average distance between patches and the average patch quality. However, this model implicitly assumes that the forager possesses perfect knowledge about the environment (Godfray, 1994; Stephens and Krebs, 1986). This assumption is highly unlikely, especially as most habitats are subject to change, and foragers are forced to deal with variability (Stephens, 1993). A possible way of estimating the actual habitat quality for a non-omniscient forager is by the use of information gained while the forager searches. McNamara and Houston (1985, 1987) have revealed that animals that make use of some prior expectation, which is updated through foraging experience, will be able to obtain an estimate for habitat profitability very quickly. A behavioral response to habitat quality has been shown in birds foraging in environments with variable distances between patches (Cuthill et al., 1994; Kacelnik and Todd, 1992). It is widely accepted to date that cognitive processes play an important role in making foraging decisions, even though evidence for this mostly comes from studies on birds (Shettleworth, 1998).
Nevertheless, another group of animals that are known for their sophisticated cognitive abilities, especially when we consider their small brain sizes, are insects (Papaj and Lewis, 1993). Parasitic wasps, for example, have proven to be ideal study objects with respect to many aspects of optimal foraging theory (Godfray and Shimada, 1999). They search for hosts, mainly larvae of other insects, as the only food source for their offspring. Therefore, a female's foraging success is directly linked to the number of offspring that she can produce, and a strong selection pressure on optimizing foraging strategies can be expected (Godfray, 1994). Learning in its broadest sense frequently occurs in these animals (Godfray and Waage, 1988; van Alphen et al., 2003). The number of conspecifics a female parasitoid experiences before visiting a host patch (Hoffmeister et al., 2000; Visser et al., 1992), the host species encountered previously (van Alphen, 1982; Vos et al., 1998), and the quality of the patch itself (Driessen and Bernstein, 1999; Haccou et al., 1991; Hemerik et al., 1993; Vet et al., 1995; Wajnberg et al., 2000) all influence the degree to which she will exploit a patch with hosts. However, in most experiments performed on patch-time allocation, the study animals were subjected only to a single patch visit, whereas under natural conditions, parasitoids have to respond to information gained on two different spatial scales, that is, at the level of the individual patch and at the habitat level. For foraging parasitoids, evidence for the ability to respond to different patch densities within their habitat and thus to different patch encounter rates in a flexible manner comes only from a single empirical study. Cronin and Strong (1999) report patch-distance–dependent rates of parasitism in the field. However, no direct observations of patch visits were performed in their study, and patch exploitation levels cannot be attributed to the behavior of individual parasitoids. Hence, in order to fit purely theoretical, a priori optimization models to the natural circumstances of foraging parasitoids, experiments are needed under controlled laboratory conditions (Godfray, 1994; van Alphen and Jervis, 1996). To close this gap in our knowledge, we have combined the manipulation of patch encounter rates with direct observations of individual wasps. Through a detailed analysis of their foraging behavior, we show that patch encounter rates influence patch residence time and the degree to which patches are exploited by the searching parasitoid.
The study system
To design multipatch experiments properly in the laboratory, a knowledge of naturally occurring host distributions is essential (van Alphen and Vet, 1986). We have chosen for our project a parasitic wasp that has been studied thoroughly in many respects: the parasitoid Asobara tabida Nees (Hymenoptera: Braconidae) is a larval endoparasitoid of Drosophila species (Carton et al., 1986). It is a solitary parasitoid, that is, only a single parasitoid larva can develop in a host, and supernumerary larvae are killed by competition (van Alphen and Nell, 1982). Ovaries contain at least 100 eggs available for oviposition immediately after a female hatches, and additional eggs mature throughout the lifetime of a female (Ellers, 1997). In their natural habitat, A. tabida females are very likely to die without having been able to deposit all their eggs, or even without finding any patch at all, and are thus considered to be limited in their reproductive success because of the time they must invest in searching rather than because of their egg complement (Ellers et al., 1998). This meets one of the basic assumptions of the MVT (Sevenster et al., 1998). A. tabida females in Northern Europe most probably encounter one of the following conditions after hatching from the host pupa (see Janssen et al., 1988): in early summer, hosts occur mainly on tree-sap fluxes, which are generally scarce; during late summer and early autumn, hosts can be found in fermenting fruits, which are abundant at that time and may occur in tight clusters under the fruit-bearing trees. We have designed our experimental set-up to resemble these natural situations.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because of the small body size of many parasitic wasps, it is almost impossible to follow individuals in the field while they are traveling between patches. On the other hand, in small-scale multipatch experiments, in which individuals can be followed, travel times tend to be quite short and show little variation (see Vos et al., 1998
). This type of experiment, by its very nature, does not provide long travel times to serve as controls, so conclusive results are difficult to obtain. We have found a solution to this problem by holding wasps captive for exactly defined periods of time between patch visits instead of letting them move freely between patches. Although this allows for a precise manipulation of "travel times" between patch visits, it does not allow wasps to travel actively in an energy-consuming way. However, as the success of A. tabida is considered to be time limited, such physiological aspects should not influence the qualitative outcome of our experiments (see Mangel, 1993).
Pre-experimental treatment
A. tabida was reared on a native host species, Drosophila subobscura Collin, in a climate-room at 18°C and 65% humidity and under a photoperiod of 16-h light/8-h dark. The rearing of the flies for the production of host larvae is described in Hoffmeister and Rohlfs (2001). Parasitoid females were isolated after hatching and kept with two males in small glass vials. They had constant access to honey as food. On the day before an experiment, males were removed, and female wasps were allowed to search a pre-experimental patch with 16 host larvae for 1 h to gain experience in handling hosts.
Experimental set-up
Three series of experiments, each consisting of four patch visits interspersed by travel times, were carried out. In series A, 3-day-old females had to wait a short time interval of 5 min as the travel time before they were introduced to the next patch: this seems to be a realistic travel time with regard to A. tabida searching under trees, where fermenting fruits are only a few centimeters apart from each other. In series B, females encountered the next patch only after 24 h, which should resemble a habitat in which only sap fluxes are available as host patches. These wasps were also 3 days old during their first patch visit but, of course, became 1 day older with each travel time. In series C, 7-day-old wasps that were otherwise treated like those from series A served as a control for the effects of aging.
The experiments were carried out at 20 ± 2°C under daylight. The experimental arena consisted of a 9-cm Petri dish containing a thin agar-layer with the patch in the center. The patch consisted of a viscous yeast suspension of 2 cm in diameter, containing 0.125 mg bakers' yeast. Sixteen first instar larvae of D. subobscura were added 24 ± 1 h before the patch was offered to a female. In addition to the yeast patch, two small droplets of yeast were added close to the wall on opposite sides of the Petri dish. A patch visit was either terminated when the wasp had left the experimental patch and started searching on one of the small yeast spots or when she walked up to the lid of the Petri dish. The latter happened in fewer than 5% of all patch visits. All host larvae were dissected after an experiment to check for superparasitism, that is, multiple ovipositions into a single host. All females were deep-frozen immediately after they had left the fourth experimental patch and were subsequently dissected to determine their remaining egg load. This was necessary because egg load is known to influence patch-leaving decisions in parasitoids (Rosenheim and Rosen, 1991) and is thus likely to be a confounding variable in our experimental results.
Data collection and statistical analysis
Wasps were observed continuously during an experiment. We recorded the occurrence of ovipositions, which were easy to recognize (see van Alphen and Drijver, 1982), together with the total duration of each of the four patch visits, by using the event-recording program The Observer 3.0. (Noldus, Wageningen). For series A, eight replicates were obtained. For series B and C, we used 10 replicates in the analysis.
For analyzing the impact of a female's egg load on her leaving decision, we had to calculate initial egg loads for every patch visit from the terminal egg loads that we obtained through dissections of the females. Because A. tabida females are able to mature new eggs throughout their life, especially when they have oviposition experience (Ellers, 1997), we can calculate initial egg loads directly only in the treatments A and C with short travel times of 5 min, for which egg maturation rate between patch visits should be close to zero. Here, we added the number of eggs laid at each patch visit to the terminal egg load. The calculation differs for the long travel time treatment in which patch visits were 24 h apart: by adding the number of eggs that the wasps laid on average in series B during their four patch visits to the terminal egg load of those wasps, we obtained an average initial egg load that was much higher than the average initial egg load for females from series A. Because wasps were designated at random to the different series, their initial egg loads (LI1) should have been similar. Taking the number of eggs of wasps from series B that exceeded the initial egg load of wasps from series A and dividing them by three long travel times of 24 h, we obtained the number of eggs that wasps from series B had matured on average per day (EM). Therefore, we added the number of eggs laid (EL) to the terminal egg load (LT) to obtain the initial egg load for the last (fourth) patch visit (LI4), thus LI4 = LT + EL4; for the third to first patch, we have to account for egg maturation, and thus, LIx = LIx+1 + ELx – EM, with x = rank-number of patch visit.
For each experimental series, generalized linear models (GLMs; Nelder and Weddenburn, 1972) were fitted to the data by using the procedure GENMOD (SAS, 1999). Into this procedure, an additional generalized estimating equation (GEE) was included to deal correctly with the time series data arising from the four measurements that we took from each individual female (SAS, 1999). Within each series, we investigated the influence of two variables, the rank-number (first to fourth) of a patch visit (Patch) and the egg load of a female (Eggs) on the following dependent variables: (1) patch residence time (Time), (2) number of ovipositions on a patch (Ovi), (3) time from the last parasitization until the patch was abandoned (giving-up time, GUT), and (4) number of larvae parasitized during the first 500 s of a patch visit (Search) as a measure of the female's searching efficiency. In a comparison among series (treatments) A through C, we tested for effects of the three variables Patch, Treatment, and Patch x Treatment (the interaction term between the former two) in their effect on the dependent variables Time, GUT, Ovi, Search, and Eggs.
For the variables Time and GUT, GLM with a gamma distribution and power(–1) link function provided the best fit to our data (program SAS 8.2: the scale parameter was estimated by maximum likelihood). For the variables Ovi, Search, and Eggs, GLM with a Poisson distribution and log link function were used (the scale parameter was estimated by the square root of "deviance/degrees of freedom").
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The values for the parameters measured during a patch visit (Time, Ovi, GUT and Search) (Table 1) were analyzed in two dimensions: (1) within a series of patch visits for detecting changes related to the rank-number of a patch visit and (2) among the experimental series to test for differences between the treatments. In addition, we analyzed whether the egg loads of the females (Eggs) (Table 2) could explain the treatment effects. The dissection of larvae revealed that multiple ovipositions into a single host did not occur in our experiments. The variable Ovi therefore directly reflected the degree of patch exploitation.
|
|
Changes within a series of patch visits
Wasps from series A showed a clear reduction in patch residence time with increasing number of short travel times experienced, that is, with increasing rank-number of a patch visit (GLM with GEE for Time, n = 8 wasps x 4 visits: pA =.02). The same was true for the older females of the age-control (series C, n = 10 x 4: pC =.01) that also experienced short time intervals between patch visits. In contrast, patch residence times did not change in wasps from series B that experienced long travel times (n = 10 x 4: pB =.53). The degree of patch exploitation was reduced across patch visits in the experiments with young wasps and short time intervals (GLM with GEE for Ovi: pA =.01), showed an insignificant negative trend in the age-control series (pC =.28), and remained unchanged by wasps that experienced long time intervals between patch visits (pB = 1.00). The giving-up times decreased slightly but insignificantly in series A, significantly in the age-controls, and were stable in series B (GLM with GEE for GUT: pA =.14, pC =.009, pB =.90). Searching efficiencies did not change significantly across patch visits in any of the treatments (GLM with GEE for Search: pA =.76, pB =.21, pC = 1.00).
For analyzing the impact of a female's egg load on her leaving decision, we calculated initial egg loads as described above. Females in series B must have matured an average of 15.3 eggs during each 24 h period of travel time. This just replaced the 15.2 eggs that they laid on average during a patch visit, and therefore, egg loads did not differ among patch visits in series B (Table 2). In contrast, egg loads decreased continuously in series A and C (Table 2), with the aged wasps from series C having a higher initial egg load than wasps from series A (GLM, n = 18: p = 0.05). Thus, the pattern of egg loads across experimental series resembled to some extent the pattern found in patch residence time and degree of patch exploitation. We therefore tested a female's egg load (Eggs) as an explanatory variable for the observed patch residence times (Time). However, we did not find a significant effect in any of the treatments, even though we found a trend in series A and C (pA =.06, pB =.38, pC =.17).
Differences among experimental series
We have shown above that the rank-number of a patch visit (Patch) has a strong influence on many aspects of a female's searching behavior in the series A and C but not in series B. To see whether these differences between series are significant, a full model, including the parameters Patch, Treatment, and Patch x Treatment, was fitted to the complete data-set. The interaction term Patch x Treatment is the most interesting variable to analyze, because it directly shows whether the behavioral response of a female parasitoid across the four patch visits differs between the three experimental series.
For patch residence times, all three parameters have a significant effect (GLM with GEE, n = 28 x 4: pP =.005, pT =.014, pPxT =.017). A closer look at series A and C, that is, the series with short travel times, revealed that the older wasps stayed significantly longer on all four patches than did the younger ones (pT =.017). The number of eggs laid during a patch visit and thus the degree of patch exploitation was strongly influenced by the rank-number of a patch visit (pP =.007), whereas there was no significant effect by the treatment per se (pT =.26). However, the degree of exploitation across patch visits differed significantly between the treatments (pPxT =.04). Giving-up times differed only among treatments (pT =.03), but not in response to the rank-number of a patch visit (pP =.07, pPxT =.18). The searching efficiencies of the parasitoids did not change with the rank-number of a patch visit or as a function of the experimental treatment (pP =.57, pT =.41, pPxT =.47).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
When parasitoids experience different patch encounter rates, foraging theory predicts that they adjust their patch residence time and the degree of patch exploitation in response (Stephens and Krebs, 1986
). We subjected A. tabida females to a series of four patch visits that were either interspersed by short or long travel-time intervals. One day before the experiment, all wasps had searched a pre-experimental patch and had thus experienced one long travel time. We predicted that the short travel-time regime indicates an increase in patch availability to the parasitoids and should therefore lead to a reduction in patch residence times and patch exploitation. The long travel-time regime, in contrast, should indicate unchanged conditions and should not lead to a change in a parasitoid's foraging behavior. Qualitatively, these predictions were fully met by our findings, but the quantitative fit of our data with predictions drawn from Charnov's MVT (Charnov, 1976) is only weak. Our results also show that the observed changes in behavior are most likely owing to the parasitoid's previous experience.
Qualitative fit with the theoretical predictions and confounding variables
In accordance with our predictions, patch times and patch exploitation levels were reduced in the high patch-encounter rate treatments (series A and C, Table 1) but remained unchanged in wasps that experienced a low patch-encounter rate (series B, Table 1). Because wasps of series A and B were equally young at the beginning of an experiment but differed in age at the end, the increasing age of the wasps in series B could be a confounding variable. However, the older control animals (series C) confirmed that our results were the effect of interpatch time intervals rather than wasp age, because the age-control wasps also showed the expected reduction in patch residence time in response to short travel times. Interestingly, these wasps did not show a reduction in the number of ovipositions and exploited all four patches close to the maximum of 16 larvae. However, as patch times in the age-control treatment were much higher right from the beginning than they were in series A, even the reduced residence times allowed the detection of almost all of the hosts. Moreover, because the time needed for parasitizing the hosts was similar in young and older wasps (i.e., their searching efficiencies were not different), the reduced patch exploitation in the older wasps was mirrored only in significantly reduced giving-up times. These findings may be attributed to the possibility that the age-control wasps estimated their habitat to be extremely poor, because they had more days without hosts before the experiment started, compared with females from the other series. In addition, they had a reduced future life expectancy on the day of experimentation. This has previously been shown to increase patch exploitation in other parasitoid species (Fletcher et al., 1994; Roitberg et al., 1993).
The searching efficiencies of the parasitoid females did not change across patch visits in any of the treatments (Table 1). Therefore, the observed reduction in patch residence times in series A and C is not attributable to the possibility that females became more skilled with the increasing number of patches visited within an experiment. In addition, we can exclude that changes in the diurnal activity pattern of the wasps affected patch residence times. This finding is also supported by data from Fleury et al. (2000), who have described a stable activity level for A. tabida during the time the experiments took place.
The egg load of a parasitoid female in series A and C is reduced from patch visit to patch visit (Table 2), and optimal foraging theory predicts that females with high egg loads should exploit patches better than females with low egg loads (Stephens and Krebs, 1986). However, by comparing p values from the analysis with Patch as an explanatory variable with that with Eggs, we find that the rank-number of the patch visit is a much better explanatory variable for the behavioral pattern that we have observed than is a female's egg load.
If females become partly egg-limited during the course of an experiment, they should become more choosy and reject low-quality hosts, which could lead to a similar pattern as that observed in the experiments. However, in our set-up, all unparasitized hosts should have been of equal quality, and all parasitized host were rejected by the females from the beginning. We also know from van Alphen (1982) and our own experience that healthy larvae of D. subobscura are almost never rejected. It is therefore highly unlikely that the rejection of low-quality hosts accounts for the pattern observed in our study. Moreover, egg-limited parasitoids should not have any incentive to reduce their patch residence times (Godfray, 1994).
Consequently, we conclude that the observed changes in patch-leaving behavior are a direct response of the parasitoids to the patch encounter rate and thus the habitat quality that they experience.
Quantitative fit with the MVT and a forager's constraints
We have thus shown that the data from our experiments qualitatively fit with expectations from foraging theory. In addition, we can use our data to compute quantitative predictions (Charnov, 1976). By combining the data on host encounters for all patch visits from all series, we have estimated the average gain function (i.e., the mean number of larvae parasitized in a patch at any time), which is shown as a solid line in Figure 1. The optimal leaving points for our experimental environments have been obtained by using the graphical method described in Stephens and Krebs (1986), which builds upon the MVT. These leaving points are shown as black arrows in the Figure 1, a through c, for series A through C, respectively. In the figure, we have also included the degree of patch exploitation (i.e., the leaving time and the number of larvae parasitized up to that moment) as estimated by the GLM; the underlying data are given in Table 1. Because giving-up times are not included in the gain function but occur in real wasp behavior, the GLM leaving points do not lie directly on the estimated gain function but a little further to the right. From Figure 1, a through c, it can be readily seen that a female's leaving point comes closer to the predicted leaving point with every short travel time that this wasp had experienced, as predicted. However, even on patch 4, patches are still overexploited by the searching wasps. This is also true for the four patch visits in series B, in which the degree of patch exploitation remains constant. By using a literature survey, Nonacs (2001) has shown that, in quantitative tests of the MVT, remaining too long on a patch is a general phenomenon in animals. In our study, it may be attributable to the following reasons. First, it is possible that the wasps in series A and C did not visit enough patches during an experiment to reach the adequate leaving point for the given environment and that further reductions would have occurred when given additional experience. Second, it has previously been shown that the MVT is too constrained in its assumptions to allow exact predictions for real foragers (Godfray, 1994, but for a quantitative fit between theory and data, see Cowie, 1977). Third, the searching females do not know the exact patch quality at any given time but have to estimate it from experience. In our case, we used two-dimensional yeast patches in order to observe wasps continuously throughout an experiment. On natural patches, i.e., three-dimensional fruits, hosts are likely to temporarily hide deeper inside the fruit, in a spatial refuge (Hoffmeister and Rohlfs, 2001). Therefore, a fruit may contain more hosts than those that the wasp can locate at any moment on the surface. Host-encounter rates on our artificial yeast patches may thus indicate, to the searching female, a host density higher than is actually present. Hence, an overestimation of patch quality by the wasp is likely, a point that could also explain the relatively long giving-up times observed in all treatments. Fourth, other constraints than merely time optimization may play a role in parasitoid foraging decisions. If, for example, the mortality risk is higher while traveling between patches (see Völkl and Kraus, 1996), a higher degree of patch exploitation should result.
|
In conclusion, our results agree with general foraging theory, even though we have not found a quantitative fit with predictions drawn from the MVT alone.
Mechanisms by which parasitoids could measure patch encounter rate
Although our results clearly show that parasitoids respond to variation of habitat quality via experienced patch-encounter rates, they do not reveal the exact mechanism that underlies the observed response. First of all, it is possible that the parasitoids are able to measure time intervals directly as has been described for Trichogramma minutum (Schmidt and Smiths, 1987) for very short time intervals. Alternatively, a wasp's egg load might be involved in the mechanism. Even though we have shown above that egg load per se is not a sufficient explanatory variable for the behavior observed in our study, A. tabida females may use an "egg timer" by comparing their egg maturation rate with their oviposition rate to assess the availability of hosts in the environment.
Another mechanism can be derived from a model developed by Waage (1979). He has suggested that a parasitoid enters a host patch with a certain "level of responsiveness to the patch edge"; this level is set by the amount of host-produced chemicals in the patch. Waage assumes that the responsiveness declines with the time spent in the patch until finally the wasp cannot perceive the patch edge any more and leaves. Ovipositions into hosts also influence the level of responsiveness. Waage's model allows parasitoids, without any prior knowledge of patch quality, to respond adaptively to patches containing different host densities. Shettleworth (1998) has extended this single patch model by suggesting that desensitization while searching a patch and becoming more sensitive for patch odors when traveling between patches could be a possible mechanism for parasitoids to respond to different travel times within their environment (Figure 2).
|
We cannot decide which of the three mechanisms given above is indeed used by the parasitoids, and thus, additional experiments are under way. However, Waage's model fits well with the behavior of A. tabida females searching a single patch (van Alphen and Galis, 1983
), and the model extension proposed by Shettleworth follows clearly from studies on habituation processes (for examples, see Shettleworth, 1998). Thus, the mechanism of sensitization and desensitization to chemicals associated with patches may very well be one of the driving forces in parasitoid decision making, because it would allow parasitoids to use the same mechanism for estimating both patch and habitat quality.
|
ACKNOWLEDGEMENTS |
|---|
We thank Gerard Driessen, several members of our department, and two anonymous referees for valuable comments on earlier versions of the manuscript; Ina Berndt for help with the cultures; and Jacques van Alphen for providing the parasitoid strain. A.T. was supported by a graduate program of the federal state of Schleswig-Holstein.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Carton Y, Bouletreau M, van Alphen JJM, van Lenteren JC, 1986. The Drosophila parasitic wasps. In: The genetics and biology of Drosophila. (Ashburner M, Carson HL, Thompson JN, eds). Orlando, Florida: Academic Press; 347–394.
Charnov EL, 1976. Optimal foraging, the marginal value theorem. Theor Popul Biol 9:129-136.[CrossRef][ISI][Medline]
Cowie RJ, 1977. Optimal foraging in great tits (Parus major). Nature 268:137-139.[CrossRef]
Cronin JT, Strong DR, 1999. Dispersal-dependent oviposition and the aggregation of parasitism. Am Nat 154:23-36.[CrossRef]
Cuthill IC, Haccou P, Kacelnik A, 1994. Starlings (Sturnus vulgaris) exploiting patches: response to long term changes in travel time. Behav Ecol 5:81-90.
Driessen G, Bernstein C, 1999. Patch departure mechanisms and optimal host exploitation in an insect parasitoid. J Anim Ecol 68:445-459.[CrossRef]
Ellers J, 1997. Life history evolution in Asobara tabida: plasticity in allocation of fat reserves to survival and reproduction. J Evol Biol 10:771-785.[CrossRef]
Ellers J, van Alphen JJM, Sevenster JG, 1998. A field study of size-fitness relationships in the parasitoid Asobara tabida. J Anim Ecol 67:318-324.[CrossRef]
Fletcher JP, Hughes J, Harvey IF, 1994. Life expectancy and egg load affect oviposition decisions of a solitary parasitoid. Proc R Soc Lond B 258:163-167.[Medline]
Fleury F, Allemand R, Vavre F, Fouillet P, Bouletreau M, 2000. Adaptive significance of a circadian clock: temporal segregation of activities reduces intrinsic competitive inferiority in Drosophila parasitoids. Proc R Soc Lond B 267:1005-1010.[Medline]
Godfray HCJ, 1994. Parasitoids: behavioral and evolutionary ecology. Princeton, New Jersey: Princeton University Press.
Godfray HCJ, Shimada M, 1999. Parasitoids as model organisms for ecologists. Res Pop Ecol 41:3-10.
Godfray HCJ, Waage JK, 1988. Learning in parasitic wasps. Nature 331:211.[CrossRef]
Haccou P, de Vlas SJ, van Alphen JJM, Visser ME, 1991. Information processing by foragers: effects of intra-patch experience on the leaving tendency of Leptopilina heterotoma. J Anim Ecol 60:93-106.[CrossRef]
Hemerik L, Driessen G, Haccou P, 1993. Effects of intra-patch experiences on patch time, search time and searching efficiency of the parasitoid Leptopilina clavipes. J Anim Ecol 62:33-44.[CrossRef]
Hoffmeister TS, Rohlfs M, 2001. Aggregative egg distributions might promote species coexistence: but why do they exist ? Evol Ecol Res 3:37-50.
Hoffmeister TS, Thiel A, Kock B, Babendreier D, Kuhlmann U, 2000. Pre-patch experience affects the egg distribution pattern in a polyembryonic parasitoid of moth egg batches. Ethology 106:145-157.[CrossRef]
Janssen A, Driessen G, de Haan M, Roodbol N, 1988. The impact of parasitoids on natural populations of temperate woodland Drosophila. Neth J Zool 38:61-73.
Kacelnik A, Todd IA, 1992. Psychological mechanisms and the marginal value theorem: effect of variability in travel time on patch exploitation. Anim Behav 43:313-322.[CrossRef]
Mangel M, 1993. Motivation, learning, and motivated learning. In: Insect learning: ecological and evolutionary perspectives (Papaj DR, Lewis AC, eds). New York: Chapman & Hall; 158–173.
McNamara JM, Houston AI, 1985. Optimal foraging and learning. J Theor Biol 117:231-249.[CrossRef]
McNamara JM, Houston AI, 1987. Memory and the efficient use of information. J Theor Biol 125:385-395.[CrossRef][ISI][Medline]
Nelder JA, Weddenburn RWM, 1972. Generalized linear models. J R Stat Soc A 135:370-384.
Nonacs P, 2001. State dependent behavior and the marginal value theorem. Behav Ecol 12:71-83.
Papaj DR, Lewis AC, 1993. Insect learning: ecological and evolutionary perspectives. New York: Chapman & Hall.
Roitberg BD, Sircom J, Roitberg CA, van Alphen JJM, Mangel M, 1993. Life expectancy and reproduction. Nature 364:108.[CrossRef][Medline]
Rosenheim JA, Rosen D, 1991. Foraging and oviposition decisions in the parasitoid Aphytis lingnanensis: distinguishing the influences of egg load and experience. J Anim Ecol 60:873-893.[CrossRef]
SAS Institute,, 1999. SAS OnlineDoc, version 8. Cary, North Carolina: SAS Institute.
Schmidt JM, Smiths JJB, 1987. Short interval time measurement by a parasitoid wasp. Science 237:903-905.
Sevenster JG, Ellers J, Driessen G, 1998. An evolutionary argument for time limitation. Evolution 52:1241-1244.[CrossRef]
Shettleworth SJ, 1998. Cognition, evolution and behavior. Oxford: Oxford University Press.
Stephens DW, 1993. Learning and behavioral ecology: incomplete information and environmental predictability. In: Insect learning: ecological and evolutionary perspectives (Papaj DR, Lewis AC, eds). New York: Chapman & Hall; 195–218.
Stephens DW, Krebs JR, 1986. Foraging theory. Princeton, New Jersey: Princeton University Press.
van Alphen JJM, 1982. Foraging behaviour of Asobara tabida, a larval parasitoid of Drosophila (PhD thesis). Leiden: Leiden University.
van Alphen JJM, Bernstein C, Driessen G, 2003. Information acquisition and time allocation in insect parasitoids. Trends Ecol Evol 18:81-87.[CrossRef]
van Alphen JJM, Drijver RAB, 1982. Host selection by Asobara tabida Nees (Braconidae; Alysiinae), a larval parasitoid of fruit inhabiting Drosophila species, I: host stage selection with Drosophila melanogaster as host species. Neth J Zool 32:215-231.
van Alphen JJM, Galis F, 1983. Patch time allocation and parasitization efficiency of Asobara tabida, a larval parasitoid of Drosophila. J Anim Ecol 52:937-952.[CrossRef]
van Alphen JJM, Jervis MA, 1996. Foraging behavior. In: Insect natural enemies: practical approaches to their study and evaluation (Kidd MJN, ed). New York: Chapman & Hall; 1–62.
van Alphen JJM, Nell HW, 1982. Superparasitism and host discrimination by Asobara tabida Nees (Braconidae; Alysiinae) a larval parasitoid of Drosophilidae. Neth J Zool 32:232-260.
van Alphen JJM, Vet LEM, 1986. An evolutionary approach to host finding and selection. In: Insect parasitoids (Waage JK, Greathead DJ, eds). London: Academic Press; 23–61.
Vet LEM, Lewis WJ, Cardé RT, 1995. Parasitoid foraging and learning. In: Chemical ecology of insects 2 (Cardé RT, Bell WJ, eds). New York: Chapman & Hall; 65–101.
Visser ME, van Alphen JJM, Nell HW, 1992. Adaptive superparasitism and patch time allocation in solitary parasitoids: the influence of pre-patch experience. Behav Ecol Sociobiol 31:163-171.
Völkl W, Kraus W, 1996. Foraging behaviour and resource utilization of the aphid parasitoid Pauesia unilachni: adaptation to host distribution and mortality risks. Entomol Exp Appl 79:101-109.[CrossRef]
Vos M, Hemerik L, Vet LEM, 1998. Patch exploitation by the parasitoids Cotesia rubecula and Cotesia glomerata in multi-patch environments with different host distributions. J Anim Ecol 67:774-783.[CrossRef]
Waage JK, 1979. Foraging for patchily-distributed hosts by the parasitoid, Nemeritis canescens. J Anim Ecol 48:353-371.[CrossRef]
Wajnberg E, Fauvergue X, Pons O, 2000. Patch leaving decision rules and the marginal value theorem: an experimental analysis and a simulation model. Behav Ecol 11:577-586.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Tentelier, E. Desouhant, and X. Fauvergue Habitat assessment by parasitoids: mechanisms for patch use behavior Behav. Ecol., July 1, 2006; 17(4): 515 - 521. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||