Bumblebees,humble pollinators or assiduous invaders? A population comparison of foraging performance in <Emphasis Type="Italic">Bombus terrestris</Emphasis>

Worldwide trade in non-native bumblebees remains largely unrestricted despite well-documented cases where introductions of non-native bees have gone dramatically wrong. Within Europe, indiscriminate importation of non-native populations of bumblebees (Bombus terrestris) for the pollination of glasshouse crops continues on a massive scale. However, no risk assessment has been conducted for these introductions, perhaps because B. terrestris is considered a native species, so shipping populations from one region to another has been implicitly assumed to present no risk. This view is clearly unjustified because Bombus terrestris populations differ significantly in their genetic makeup as demonstrated by strong differences in coat colour and behavioural traits. Therefore, for the first time we compare an important competitive trait, namely foraging performance, between commercially available B. terrestris populations in contrasting environments. We test whether commercially reared populations differ in their nectar foraging performance and whether this is influenced by both their source environment and the one they are introduced into. We do this by means of a reciprocal transplant experiment. Strong, consistent inter-population differences in performance occurred irrespective of test location: Canary Island bees (B. t. canariensis) were superior to Sardinian bees (B. t. sassaricus), which were generally superior to mainland European bees (B. t. terrestris). These inter-population differences in performance were largely explained by inter-population variation in forager size, with larger bees being superior foragers. However, even when body size was accounted for, “native” bees were not superior to transplanted non-native bees in all but one case. We conclude that non-native populations, especially those with large foragers, can be highly competitive foragers. This could lead to their establishment and displacement of native bees. Therefore, we recommend that unregulated movements of non-native B. terrestris populations within Europe should not be carried out without a full risk assessment.     

Does intraspecific size variation in bumblebees allow colonies to efficiently exploit different flowers?

Abstract.  1. It has long been known that foraging bumblebee workers vary greatly in size, within species, and within single nests. This phenomenon has not been adequately explained. Workers of their relatives within the Apidae exhibit much less size variation.
2. For the bumblebee Bombus terrestris size, as measured by thorax width, was found to correspond closely with tongue length, so that larger bees are equipped to feed from deeper flowers.
3. The mean size of worker bees attracted to flowers was found to differ between plant species, and larger bees with longer tongues tended to visit deeper flowers.
4. Finally, handling time depended on the match between corolla depth and tongue length: large bees were slower than small bees when handling shallow flowers, but quicker than small bees when handling deep flowers.
5. Size variation within bumblebees may be adaptive, since it enables the colony as a whole to efficiently exploit a range of different flowers. Possible explanations for the marked differences in size variation exhibited by bumblebees compared with Apis species and stingless bees (Meliponinae) are discussed.     

Endoparasitic flies,pollen-collection by bumblebees and a potential host-parasite conflict

Summary Conopid flies (Conopidae, Diptera) are common larval parasites of bumblebees. The larva develops inside the abdomen of workers, queens and males. Development is completed within 10–12 days after oviposition when the host is killed and the parasite pupates in situ. Development results in parasitised bees becoming unable to carry large loads of nectar, as the conopid larvae reside where the honey crop is normally located. Furthermore, an addition to the bee's unloaden body mass is likely (average larval weight reached at pupation by the common parasite species Sicus ferrugineus: ±SD 36.3±12.3 mg, n=59; by Physocephala rufipes: 55.8±16.9 mg, n=108). We here asked whether the propensity of workers of the bumblebee Bombus pascuorum to collect nectar rather than pollen is related to the presence of conopid larvae. For samples of bees (n=2254 workers) collected over 3 years of field studies in northwestern Switzerland, there was no difference in the frequency of bees caught as pollen collectors among parasitised (38.1% of cases, n=210) as compared to non-parastised bees (43.9%, n=360) ( ²=1.83, n.s.). However, compared to the non-parasitised bees (n=360), those hosts containing a third (last) instar larva (n=9) were less likely to collect pollen than expected by chance ²=6.91, P=0.003. Similarly, hosts with short survival time between capture and being killed by the developing larva (which hence must have harboured a late instar parasite at time of capture) were less likely to collect pollen (8%, n=25) than those found not parasitised (37.6%, n=891 ²=9.16, P<0.001). Late instar larvae grow so big that they fill the entire abdomen. Although there was also a tendency for presumably older bees to collect less pollen, this is unlikely to explain the observations. We also discuss whether these changes in foraging behaviour of bumblebees may reflect a host-parasite conflict over the type of resource to be collected.     

A model of pollinator‐mediated gene flow between plant populations with numerical solutions for bumblebees pollinating oilseed rape

We present a model that predicts the level of gene flow mediated by animal pollinators from a source population to a sink population. The model requires specification of three elements: (1) the paternity that originates from a single flower, the paternity shadow; (2) the mean number of flowers that pollinators visit during stays in the sink population, the residence; (3) the proportion of pollinators arriving at the sink that carry pollen from the source population. Provided that pollinators visit enough flowers in the sink to exhaust the paternity shadows from the source, the general results are that gene flow is inversely proportional to the mean pollinator residence in the sink population, and is proportional to the fraction of pollinators arriving with pollen from the source. These results are used to propose explanations for two of the widely observed patterns in gene flow among plant populations. Numerical solutions to the model are derived using experimentally determined values of elements (1) and (2) that represent bumblebees, Bombus spp., visiting agricultural fields of oilseed rape, Brassica napus L. In B. napus, the paternity shadow attenuates rapidly over approximately 20 recipient flowers. Mean bumblebee residences in the fields studied varied between 490 and 720 flowers. In the absence of a direct measurement of element (3), we calculated the maximum level of bumblebee‐mediated gene flow by assuming that all bees arrived at the sink saturated with pollen from extrinsic sources. In this case, the model predicts that bumblebee‐mediated gene flow accounted for between 0.1% and 0.5% of the progeny in the agricultural fields studied. A likelihood analysis of our observations is unable to reveal convincingly the proportion of bees arriving at the sink via a source population, but the literature suggests that bumblebees have high site fidelity, which implies that bee‐mediated gene flow may be substantially less than our estimated maximum. We consider the role of various factors, including wind pollination, in accounting for the differences between the model's predictions and the generally higher levels of gene flow observed in previous studies of oilseed rape.     

Consequences of the arms race between <Emphasis Type="Italic">Maculinea teleius</Emphasis> social parasite and <Emphasis Type="Italic">Myrmica</Emphasis> host ants for myrmecophilous butterfly conservation

The arms race between Maculinea butterflies and Myrmica host ants leads to local host-parasite adaptations. In our study, we assessed whether sympatric and allopatric Myrmica scabrinodis populations exhibit behavioural differences towards Maculinea teleius larvae during the adoption-period when butterfly larvae need to be taken inside the Myrmica nest. The second aim was to assess the butterfly survival rate inside ant colonies from different populations. We used one sympatric host population and three allopatric populations: one infested by M. teleius and two uninfested populations. We found that ants from the sympatric population showed a higher number of positive behaviours toward M. teleius larvae during adoption than ants from the allopatric populations. There were no differences in the number of inspection or negative behaviour events. The survival of butterfly larvae was highest inside sympatric host colonies and differed from the survival of M. teleius reared by ants from the allopatric, uninfested populations. No difference was found for the survival rate of M. teleius raised by infested, allopatric host colonies compared to sympatric host populations. Our results suggest the lack of behavioural counter-adaptations of local hosts of M. teleius that more easily adopt and rear butterfly caterpillars compared to naive M. scabrinodis colonies. Our results may also have implications for Maculinea butterfly conservation, especially for reintroduction programmes. We suggest that the existence of behavioural host defences should be checked for the source host population, as well as for the Myrmica population from the reintroduction site. It may also be reasonable to introduce several Myrmica host colonies from the source butterfly host population.     

Haplotype structure,adaptive history and associations with exploratory behaviour of the DRD4 gene region in four great tit (Parus major) populations

The assessment of genetic architecture and selection history in genes for behavioural traits is fundamental to our understanding of how these traits evolve. The dopamine receptor D4 (DRD4) gene is a prime candidate for explaining genetic variation in novelty seeking behaviour, a commonly assayed personality trait in animals. Previously, we showed that a single nucleotide polymorphism in exon 3 of this gene is associated with exploratory behaviour in at least one of four Western European great tit (Parus major) populations. These heterogeneous association results were explained by potential variable linkage disequilibrium (LD) patterns between this marker and the causal variant or by other genetic or environmental differences among the populations. Different adaptive histories are further hypothesized to have contributed to these population differences. Here, we genotyped 98 polymorphisms of the complete DRD4 gene including the flanking regions for 595 individuals of the four populations. We show that the LD structure, specifically around the original exon 3 SNP is conserved across the four populations and does not explain the heterogeneous association results. Study‐wide significant associations with exploratory behaviour were detected in more than one haplotype block around exon 2, 3 and 4 in two of the four tested populations with different allele effect models. This indicates genetic heterogeneity in the association between multiple DRD4 polymorphisms and exploratory behaviour across populations. The association signals were in or close to regions with signatures of positive selection. We therefore hypothesize that variation in exploratory and other dopamine‐related behaviour evolves locally by occasional adaptive shifts in the frequency of underlying genetic variants.     

The genetic conditions in heterozygous and homozygous populations ofDrosophila I. The fate of alien chromosomes

In experimental populations ofDrosophila melanogaster lethal chromosomes with dominant markers and inversions were introduced and the frequency changes of the markers studied during a period of several generations. The base populations of the various experiments differed from each other with respect to their degree of heterozygosity. Monochromosomal populations were isogenic for a quasinormal + chromosome, dichromosomal populations contained the genetic material of two different + chromosomes, trichromosomal of three, tetrachromosomal of four, hexachromosomal of six and polychromosomal populations of many normal chromosomes. Marker chromosomes with the dominant genesLCy, Cy, Pm orD respectively were added to the populations with an initial frequency of 16,6 per cent. The fate of the dominant markers was different in different populations. In some the marker chromosome reached equilibrium frequencies, in others they were eliminated with variable speed. As a rule the lethal marker chromosomes were accepted by monochromosomal populations; however, they were eliminated from populations with a higher degree of heterozygosity. Since in all populations one genotype, namely the homozygote for the marker chromosome, was lethal, the adaptive values c of the +/LCy, +/Cy, +/Pm or +/D heterozygotes could easily be calculated from the experimental data. This c value can be used as a measure for the combining ability of the marker chromosomes. It could be shown that c depends on the degree of heterozygosity of a population or in other words that the average degree of heterozygosity of the marker free individuals determines the selection processes. An equation can be arrived at which fits the experimental results very well if superiority of heterozygous +/+ individuals over +/+ homozygotes is assumed. From that it was concluded that heterosis is the determining variable in our experiments. An attempt was undertaken in order to decide if in our case the observed heterosis was due to dominance or to overdominance. It was postulated that in di-, tri-, tetra- or hexachromosomal populations the adaptive values of the marker free normals should progressively increase if recessive detrimental genes are the cause of heterosis but not if heterozygosity on many loci leads to overdominance. The a values of the +/+ individuals were ealeulated from the frequency changes of the marker chromosomes for each subsequent two-generation period. Unfortunately only two different dichromosomal populations were available. These showed increasing adaptive values for the normals. The tri-, tetra-, and hexachromosomal populations, however, gave different results, some with increasing, some with fluctuating adaptive values. From that it was concluded that heterosis can be due in one case to dominance and in the other to overdominance. In either case, the recessive genetic load may be rather important as a determinating factor in the dynamics of populations.Dedicated to Prof. Th. Dobzhansky on the occasion of his seventieth birthday in deepest gratitude.     

Effects of adaptive predatory and anti-predator behaviour in a two-prey—one-predator system

Summary Two prey populations that share a common predator can interact indirectly by causing changes in the predator's foraging behaviour. Previous work suggests that adaptive choice of prey by the predator usually has two related consequences: (i) the predation rate on a particular prey species increases with the relative and/or absolute abundance of that prey; and (ii) increases in either prey population produce a short-term increase in the fitness of the other prey (short-term indirect mutualism between prey). This paper investigates how these two consequences are changed if the prey exhibit adaptive anti-predator behaviour. In this case, the predation rate on a particular prey often decreases as the prey's density increases. The predator then usually exhibits negative switching between prey. However, the presence of adaptive antipredator behaviour does not change the short-term mutualism between prey. In this case, as a prey becomes less common, it achieves a larger growth rate by reducing its anti-predator effort. These results imply that observations of the relationship between prey density and predation rate cannot be used to infer the nature of the behavioural indirect effect between prey that share a predator.     

Bumblebees Do Not Respond to Variance in Nectar Concentration

Worker bumblebees (Bombus fervidus) were given repeated binary choices between two colors of artificial flowers with the same associated mean nectar concentration (X? = 20%), but with different variances in nectar concentration. Flowers of one color, yellow or blue, rewarded a bee with 1 μl of 20% sucrose solution (low-variance flower type) on each visit (p = 1) and flowers of the other color rewarded a bee on each visit with 1 μl of either 10% or 30% sucrose (p = 0.5; high-variance flower type). Of the 10 bees tested, nine showed no preference for either the high- or low-variance flowers (indifferent or risk-insensitive). This result is similar to honeybee responses to variation in nectar concentration, despite differences in foraging ecology between bumblebees and honeybees. Flower-choice behaviour in the presence of variance in nectar concentration is a response to the expected concentration of the alternative flower types. Possible mechanisms of risk-sensitive foraging behaviour in bees are discussed.     

Appetitive conditioning technique reveals behavioural limits to passage performance in fishes

Numerous environmental features have the potential to act as barriers to fishes in the field. Passage of these barriers depends on two main properties: willingness (behavioural) and physiological capacity. The physiological swimming capacity of fishes has been well studied. However, because most barrier research has focused on the population level effects of man – made barriers, little attention has been given to the behavioural aspects of fish passage. We used appetitive conditioning to create a simple laboratory based protocol to estimate behavioural limits to fish passage. We tested our protocol using barrier height (or passage space) as a model. Our protocol successfully identified behavioural limits in two species, juvenile rainbow trout, Oncorhynchus mykiss, and fathead minnows, Pimephales promelas. Our most interesting observation was that even a partial, submerged barrier had the potential to block fish movements. Our results suggest a potential use for conditioning techniques in assessing factors that constitute behaviour barriers to fishes. This information may help us to better predict fish movements in the field and may help us to design barriers that are more efficient at passing fishes.     

Local variability in population structure and density of the protogynous reef herbivoreSparisoma viride

Synopsis We compare the (relative) abundance of life phases [juveniles (JU), initial phase (IP) and terminal phase (TP) fish], social categories (territorial and group adults), and fish following alternative mating styles, in three local populations of the protogynous reef herbivore,Sparisoma viride, on the fringing reef of Bonaire (Netherlands Antilles). In order to determine the adaptive significance of variations in social organization, they are related to the density of conspecifics and other herbivores and to the availability of food, shelter and mating sites. The most striking difference is the high abundance of JU and group fish at one location (Playa Frans) and the total absence of group fish at another (Red Slave). These differences are coherent with a gradient in population density, total herbivore density, scarid grazing pressure, and reproductive output, all of which are highest at Playa Frans and lowest at Red Slave. Exposure to waves and currents shows an inverse trend. The differences in the relative abundance of territorial fish can be explained by the concept of economic defendability, which is reduced at higher population density. In a life history context, small TP group males represent bachelors that sacrifice current reproduction for better future prospects. As predicted by life history theory, early sex change is promoted at sites where the future rewards are higher (higher spawning rates of large TP males) and where the costs incurred during the bachelor phase are reduced (more spawning opportunities for group TP males). At Red Slave an alternative male mating style (streaking) appears to be promoted by the lack of a refuge for group TP males and by a dense gorgonian canopy, allowing IP males to reside inside territories. We conclude that most observed differences in population structure can be considered adaptive in an ecological and in a life history context. Population density is a major factor in both contexts. Analysis of the variability in adult density in relation to JU density and the availability of food and shelter indicates that theS. viride populations at Bonaire are not totally controlled by stochastic processes. Considering the small spatial scale and the high dispersal of the planktonic embryos and larvae, the observed variability in behavioural and life history traits ofS. viride points to a high degree of phenotypic plasticity.     

Are there genetically controlled habitat-specific differences in spatial aggregation of drosophilids?

Variation in spatial patterns of competing organisms is of fundamental importance for community and population processes. Yet the mechanisms controlling subjects like the degree of spatial aggregation in competing insect larvae across fragmented resources have rarely been addressed. In the present study, we tested for systematic differences in the spatial distribution patterns of Drosophila subobscura in natural fly communities, and found significant differences in two habitats that differ in the availability of breeding substrate types (decaying fruits). Assuming that the spatial egg-laying behaviour of drosophilids is under genetic control, and that different breeding substrates mediate different density-dependent larval fitness consequences, we tested whether adaptive genotypic variability is involved in the local variation of egg distribution patterns. We extracted isofemale lines from both habitats and analysed the spatial distribution of eggs achieved by single female flies under controlled laboratory conditions. This is a reasonable first test, because spatial patterns at the fly population level can be attributed to individual egg-laying behaviour. The degree of individual egg aggregation significantly depended on fly line identity, which indicates the existence of behavioural variants in natural populations. Based on habitat-dependent differences in the degree of spatial aggregation, we discuss to what extent our findings may reflect a behavioural adaptation to local breeding conditions.     

Different demography of friends and strangers: an experiment on the impact of kinship and familiarity in Clethrionomys glareolus

Summary We examined demographic effects of familiarity and relatedness in the bank vole Clethrionomys glareolus (Schreber) in four 0.5-ha enclosures in Central Finland. In two enclosures were mature voles which had overwintered together and some of their mature off-spring (hereafter referred to as Friends), and in the other two individuals of the same species captured from different localities near the study area (Strangers). The experiment lasted from June to September. The populations of Friends reached densities twice as high as those of Strangers with a significantly higher rate of recruitment and survival of the young. This may have been due to mutual familiarity decreasing antagonism towards the juveniles. The conflicting results obtained from studies of Clethrionomys and Microtus are discussed. We believe that these genera represent behavioural adaptations to different habitats and ways of life. Most behavioural population regulation hypotheses are based on studies of Microtus. We conclude that these results should be applied with great caution to other rodent genera.     

Genome-wide analysis of signatures of selection in populations of African honey bees (Apis mellifera) using new web-based tools

Background

With the development of inexpensive, high-throughput sequencing technologies, it has become feasible to examine questions related to population genetics and molecular evolution of non-model species in their ecological contexts on a genome-wide scale. Here, we employed a newly developed suite of integrated, web-based programs to examine population dynamics and signatures of selection across the genome using several well-established tests, including F_ST, pN/pS, and McDonald-Kreitman. We applied these techniques to study populations of honey bees (Apis mellifera) in East Africa. In Kenya, there are several described A. mellifera subspecies, which are thought to be localized to distinct ecological regions.

Results

We performed whole genome sequencing of 11 worker honey bees from apiaries distributed throughout Kenya and identified 3.6 million putative single-nucleotide polymorphisms. The dense coverage allowed us to apply several computational procedures to study population structure and the evolutionary relationships among the populations, and to detect signs of adaptive evolution across the genome. While there is considerable gene flow among the sampled populations, there are clear distinctions between populations from the northern desert region and those from the temperate, savannah region. We identified several genes showing population genetic patterns consistent with positive selection within African bee populations, and between these populations and European A. mellifera or Asian Apis florea.

Conclusions

These results lay the groundwork for future studies of adaptive ecological evolution in honey bees, and demonstrate the use of new, freely available web-based tools and workflows (http://usegalaxy.org/r/kenyanbee) that can be applied to any model system with genomic information.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1712-0) contains supplementary material, which is available to authorized users.     

The foraging movements of bumblebees on vertical “inflorescences”: An experimental analysis

Summary The bumblebees,Bombus edwardsii, move upward while visiting consecutive flowers on artificial inflorescences. This response is unrelated to the vertical patterning of rewards in the flowers of inflorescences. However, when rewards are greatest in the bottommost flowers the bees learn to start lower and leave before reaching the topmost (empty) flowers. Conversely, when rewards are greatest in the topmost flowers they tend to start in the middle of the inflorescence and depart from the top. When rewards are equal in all flowers bees start near the bottom and depart near the top of inflorescences. These behavioral patterns tend to maximize the number of visits to rewarding flowers while minimizing visits to non-rewarding flowers, thereby enhancing foraging returns.Supported by NSF grant DEB77-08430.     

Predation mediated population divergence in complex behaviour of nine‐spined stickleback (Pungitius pungitius)

The proximate and ultimate explanations for behavioural syndromes (correlated behaviours – a population trait) are poorly understood, and the evolution of behavioural types (configuration of behaviours – an individual trait) has been rarely studied. We investigated population divergence in behavioural syndromes and types using individually reared, completely predator‐ or conspecific‐naïve adult nine‐spined sticklebacks (Pungitius pungitius) from two marine and two predatory fish free, isolated pond populations. We found little evidence for the existence of behavioural syndromes, but population divergence in behavioural types was profound: individuals from ponds were quicker in feeding, bolder and more aggressive than individuals from marine environments. Our data reject the hypothesis that behavioural syndromes exist as a result of genetic correlations between behavioural traits, and support the contention that different behavioural types can be predominant in populations differing in predation pressure, most probably as a result of repeated independent evolution of separate behavioural traits.     

The effects of landscape on bumblebees to ensure crop pollination in the highland agricultural ecosystems in China

Honey bees and wild bees provide critical pollination services to agricultural ecosystems; however, the relative contributions of different bee taxa are not well understood. The natural habitats surrounding farmland support food and nesting resources for wild bees and therefore play an important role in the maintenance of crop pollination. In this study, we selected Cucurbita pepo L. (squash) as a model crop to investigate the relative importance of honey bees and bumblebees in pollinating the crop. Thirteen fields, which were surrounded by a gradient of natural habitat, were investigated on the Yunnan‐Guizhou Plateau in China. We measured the visit densities of honey bees and bumblebees, the number of pollen grains deposited in a single visit by the two bee taxa, as well as the overall pollen grains deposited on stigmas during a flowering day, and then used Bayesian inference to decouple the pollen grains deposited by either the honey bees or the bumblebees. Compared with honey bees, bumblebees deposited a higher number of pollen grains on stigmas in a single visit, but had a lower visit density than honey bees. Meanwhile, the bumblebee visit density increased along the proportion of natural habitat, while the honey bee visit density was not affected by the surrounding natural habitat. Data simulations using Bayesian inference showed that on a flowering day, the number of pollen grains deposited by bumblebees increased with the proportion of natural habitat in the surrounding landscape, but the number of pollen grains deposited by honey bees did not. Moreover, the total numbers of pollen grains deposited by honey bees or bumblebees alone were all below 2000 (the critical level to satisfy the pollination requirement of this crop). Pollen calculations demonstrated that the number of pollen grains deposited by the two bee taxa was greater than 2000 in fields surrounded by more than 13% natural habitat (grasslands and forests). The results revealed that bumblebees ensured C. pepo pollination in combination with honey bees in the highland agricultural ecosystems.     

On the coevolution of social responsiveness and behavioural consistency

Recent research focuses on animal personalities, that is individual differences in behaviour that are consistent across contexts and over time. From an adaptive perspective, such limited behavioural plasticity is surprising, since a more flexible structure of behaviour should provide a selective advantage. Here, we argue that consistency can be advantageous because it makes individuals predictable. Predictability, however, can only be advantageous if at least some individuals in the population respond to individual differences. Consequently, the evolution of consistency and responsiveness are mutually dependent. We present a general analysis of this coevolutionary feedback for scenarios that can be represented as matrix games with two pure strategies (e.g. hawk-dove game, snowdrift game). We first show that responsive strategies are favoured whenever some individual differences are present in the population (e.g. due to mutation and drift). We then show that the presence of responsive individuals can trigger a coevolutionary process between responsiveness and consistency that gives rise to populations in which responsive individuals coexist with unresponsive individuals who show high levels of adaptive consistency in their behaviour. Next to providing an adaptive explanation for consistency, our results also link two key features associated with personalities, individual differences in responsiveness and behavioural consistency.     

Individual adaptations in stochastic environments

Summary Many natural populations undergo radical and unpredictable fluctuations, associated with stochastic environmental conditions. Under such circumstances, fitness of a genotype (or strategy) is defined as the geometric mean of the intergenerational genotypic population growth ratel(t). Unfortunately, this population-level criterion has proved difficult to apply at the level of individual organisms.After developing a formula for the variance ofl as the sum of developmental and environmental variance, we discuss several models of individual adaptations, involving clutch size, progeny size and number, and foraging behaviour under risk of predation, based on the geometric-mean fitness concept. We then show how the method of dynamic programming can be extended to deal with facultative behaviour in stochastic environments. Finally we discuss the concept of an evolutionarily stable strategy in a stochastic environment.Our analysis suggests several novel interpretations of field and laboratory observations. Under the geometric mean criterion behaviour may be determined primarily by the worst likely environment; behaviour may appear suboptimal if observed only under normal or average conditions. For example,except under extreme environmental conditions, avian clutches larger than those that are observed might result in increased fecundity, with little if any cost of reproduction in terms of parental survival; however, in unusually bad years such large clutches might be disastrous, in terms of parental survival. This consideration may help explain some recently reported experimental clutch-size manipulation results. Similarly, our analysis indicates that the known phenomenon of seasonal reduction in seed size may constitute a double bet-hedging strategy, determined by parental mortality risk and future seed survival probability. We also discuss circumstances in which phenotypic polymorphism is an adaptation to environmental uncertainty. Thus almost any individual life history or behavioural adaptation may be affected by environmental stochasticity.     

Behavioural evidence of colour vision in free flying stingless bees

Colour vision was first demonstrated with behavioural experiments in honeybees 100 years ago. Since that time a wealth of quality physiological data has shown a highly conserved set of trichromatic colour receptors in most bee species. Despite the subsequent wealth of behavioural research on honeybees and bumblebees, there currently is a relative dearth of data on stingless bees, which are the largest tribe of the eusocial bees comprising of more than 600 species. In our first experiment we tested Trigona cf. fuscipennis, a stingless bee species from Costa Rica in a field setting using the von Frisch method and show functional colour vision. In a second experiment with these bees, we use a simultaneous colour discrimination test designed for honeybees to enable a comparative analysis of relative colour discrimination. In a third experiment, we test in laboratory conditions Tetragonula carbonaria, an Australian stingless bee species using a similar simultaneous colour discrimination test. Both stingless bee species show relatively poorer colour discrimination compared to honeybees and bumblebees; and we discuss the value of being able to use these behavioural methods to efficiently extend our current knowledge of colour vision and discrimination in different bee species.