Search

The ideal-free distribution and central-place foraging are important ecological models that can explain the distribution of foraging organisms in their environment. However, this model ignores distance-based foraging costs from a central place (hive, nest), whereas central-place foraging ignores competition. Different foraging currencies and cooperation between foragers also create different optimal distributions of foragers, but are limited to a simple two-patch model. We present a hybrid model of the ideal-free distribution that uses realistic competitive effects although accounting for distance-based foraging, and test it using honey bees (Apis mellifera L.) foraging in canola fields (Brassica napus L.). Our simulations show that foragers maximizing efficiency (energy profits ÷losses) prioritize distance to their aggregation more than those maximizing net-rate (energy profits ÷time), and that social foragers move to more distant patches to maximize group benefits, meaning that social foragers do not approach an ideal-free distribution. Simulated efficiency-maximizers had a hump-shaped relationship of trip times with distance, spending shorter amounts of time in both nearby and far-away patches. Canola fields were far more attractive to simulated foragers than semi-natural areas, suggesting limited foraging on semi-natural lands during the bloom period of canola. Finally, we found that the observed distribution of honey bees in canola fields most closely resembled the optimal distribution of solitary efficiency-maximizers. Our model has both theoretical and practical uses, as it allows us to model central-place forager distributions in complex landscapes as well as providing information on appropriate hive stocking rates for agricultural pollination.

Insect pollination of flowers should change the within-season allocation of resources in plants. But the nature of this life-history response, particularly regarding allocation to roots, photosynthetic structures, and flowers, is empirically unresolved. This study uses a greenhouse experiment to investigate the effect of insect pollination on the reproductive output of 23 varieties of a globally important crop—canola (Brassica napus). Overall, insect pollination modified the functional characteristics (flower timing & effort, plant size & shape, seed packaging, root biomass) of canola, increasing seed production and quality, and pollinator dependence. Reproductive output and pollinator dependence were defined by strong trait trade-offs, which ranged from more pollinator-dependent plants favouring early reproductive effort, to less pollinator-dependent plants favouring a prolonged phenology with smaller plant size and lower seed quality. Seed production decreased with pollinator dependence in the absence of pollinators. The agricultural preference for hybrid varieties will increase seed production compared to open-pollinated varieties, but, even so, pollinators typically enhance seed production of both types. Our study elucidates how insect pollination alters the character and function of a globally important crop, supporting optimization of yield via intensification of insect pollination, and highlights the beneficial effects of insect pollination early in the season.

Wing wear reflects the accumulation of irreversible damage to an insect’s wings over its lifetime and this damage should influence flight performance. In the case of bumble bees, flight seems robust to variation in wing-area asymmetry and air pressure, but not to loss of wing area. However, how the pattern of wing wear affects flight performance remains unstudied. In nature, wing wear typically occurs in a ragged and haphazard pattern along the wing’s trailing margin, a shape strikingly different from the straight cut applied in past studies. In this study, we test if shape of wing wear (implemented as four distinct treatments plus a control) affects maximum load-lifting capabilities and wingbeat frequency of worker common eastern bumble bees (Bombus impatiens Cresson, 1863). We found that shape of wing wear of 171 mg bees had no detectable effect on maximum load-lifting capability (detectable effect size = 18 mg) or on wingbeat frequency (detectable effect size = 15 Hz), but that loss of wing area reduced load-lifting capability and increased wingbeat frequency. The importance of wing area in explaining the load-lifting ability of bumble bees is reinforced in this study. But, paradoxically, shape of wing wear did not detectably affect lift generation, which is determined by unsteady aerodynamic forces in these lift-reliant insects.

Increasing fragmentation of forests worldwide by timber and industrial development makes it important to understand the edge effects of common anthropogenic disturbances on forest fauna. We collected ground-active spiders along transects across the edge of logging clearcuts, gravel roads and gas pipelines in the boreal forest of Alberta, sampling on the disturbance (10 m from forest edge), and 10, 45, and 200 m into the forest. We asked whether the three disturbances were associated with edge effects on spider communities, and whether the extent of their associated edge effects were equivalent. The spider community at the edges of clearcuts was distinct from interior and on-disturbance communities 10 m into the forest from the clearcut edge, showing an edge effect of between 10 and 45 m from clearcut edges, while no edge effects were apparent at road and pipeline edges. Edge effects therefore differ at linear and non-linear openings in the boreal forest, which suggests that small linear openings may be associated with minimal edge effects compared to large polygonal forest openings. This result has important consequences for forest management, where clearcuts and other non-linear openings are likely to cause edge effects on spider communities that are between 10 and 45 m in their extent. The small size of clearcuts as practiced in the public forests of Canada, and their dense and broad application across the landscape, makes this edge effect of broad spatial significance in protecting biodiversity in managed landscapes.

The rate of floral nectar production, an important trait for supporting populations of pollinators, is known to be affected by local environment. This study examines the effect, 9 years after an anthropogenic disturbance—variable retention logging,—on the rate of daily nectar production of four common herbs ( Chamerion angustifolium , Delphinium glaucum , Mertensia paniculata and Vicia americana ) in the boreal forest of Alberta, Canada. We measured daily nectar production rates 9 years after experimental logging in three treatments: clear cuts (all merchantable trees removed), 50% of trees retained, and unlogged controls. Per-flower nectar production rate was unaffected by logging for C. angusifolium, D. glaucum , and M. paniculata , and lower in unlogged controls than in the 50% logging treatment for V. americana . Overall, we found a surprising robustness of nectar production of these herbaceous plants to the physical and environmental changes following logging. While other research has shown boreal forest logging to affect the densities of understory herbs, the per-flower nectar production in boreal forest understory herbs appears to be relatively insensitive to this disturbance.

SUMMARYFlying is an ecologically important behaviour in many insects, but it often results in permanent wing damage. Although wing wear in insects is often used as a method to determine insect age, and is associated with an increased risk of mortality, the causes of wing wear are unresolved. In this paper, we examine whether wing use while foraging explains wing wear in bumble bees (Bombus spp.). Wing wear may result from three distinct flight characteristics during foraging: time spent in flight, flight frequency and frequency of wing collisions with vegetation. To test these hypotheses for causes of wing wear, we recorded digital video of individually marked bumble bees foraging in nature on 12 different plant species that result in variation in these flight characteristics, and recaptured these individuals to photograph their wings over time. Bumble bees with a higher frequency of wing collisions showed an increased loss of wing area, which became more severe over time. Neither time in flight nor flight frequency was uniquely and significantly associated with wing wear. Therefore, the collision frequency hypothesis best explained wing wear in bumble bees. We conclude that wing use during foraging in bumble bees results in wing wear. Wing wear reflects behaviour, not simply age. Because wing wear has elsewhere been shown to increase mortality, this study provides an important mechanism linking foraging behaviour with lifespan.

Damage to structures that enable mobility can potentially influence foraging behavior. Bumble bees vary in extent of individual wing wear, a trait predicted to affect mechanical performance during foraging. This study asks 1) do bumble bees distribute themselves across different floral densities in accordance with their concurrent wing wear? and 2) does wing use in foraging bumble bees depend on concurrent wing wear? We observed individually identifiable bumble bees foraging in 1-m-super-2 patches of high and low floral density of 3 plant types (Geranium viscosissimum, Melilotus alba/officinalis, and Monarda fistulosa). Bee preference for flower density was determined by an interaction between wing area and bee body size: small worn-winged bees selected patches of higher density, whereas large worn-winged bees selected patches of lower density. Bees with more worn wings spent less time in flight and flew less often; however, only foragers collecting pollen appeared to reduce wing-use behavior (i.e., wing collision frequency) that leads to loss of wing area. This study demonstrates that individuals respond to wing damage in different ways depending on their body size and foraging task. Wing wear clearly influences foraging behavior of worker bumble bee. Copyright 2011, Oxford University Press.

Boreal forests are currently facing intensified logging pressures, yet we know little about impacts of logging on the pollination community. This study extends research done immediately before and after logging, to consider its longer term effects. We censused bumble bee and flower communities 8–9 years after experimental variable retention logging. Harvest treatments left 0%, 10–20%, 50–75%, or 100% of the original trees after logging. Total bumble bee and floral abundances were increased by all forms of logging; however the floral community in retention cuts (i.e., 10–20% and 50–75% of trees remaining) was more similar to undisturbed controls than was the community in clearcuts. Bumble bees in low-retention treatments (i.e. 10–20%) were in an ideal free distribution (IFD) with their floral resources, which allows for equal per-flower rate of visitation, regardless of local flower density. In clearcuts (i.e. 0%) and high-retention areas (i.e. 50–75%) bees were at an IFD with respect to the density of their floral resources, but undermatched (in 50–75% treatments) or were resource-independent (in 0% treatments) with respect to the energetic value of their floral resources. Bees in unlogged forest controls (i.e., 100%) deviated from an IFD by undermatching. Overall, low-retention logging (10–20%) was an improvement over clearcutting, because there was less change to the floral community, while still allowing for the achievement of an ideal free distribution. However, unlogged forest appeared to be negatively affected (in the lack of an IFD) by the presence of adjacent logged forests: numerical responses of bees to flowers were altered, with implications for bee foraging success and plant pollination service. Forest reserves should therefore incorporate buffer zones when adjacent to logged forest, to preserve the inter-relationships in their bee-influenced pollination community.

Foraging animals must often be sensitive to the dual risks of predation and energy shortfall, and should devalue risk of predation as risk of starvation increases. This study examines whether the probability of colony-level energy shortfall influences the predation risks taken by individual foraging bumble bees (Bombus occidentalis) in confined colonies as they collected sugar water from artificial flowers. Before each trial, colony energy stores were manipulated by enhancing or removing the contents of the colony's honey pots. Feeding bees were significantly more likely to flee from an artificial predator model (a black styrofoam ball moved overhead) when their colony's energy stores were enhanced, relative to when their colony's energy stores were depleted. Despite their defensive stings and aposematic coloration, bumble bees are more likely to accept “risky” foraging situations when the energy needs of their colony are increased.

Habitat selection theory predicts that consumers should distribute themselves across alternative resources in a fitnessmaximizing manner. One such distribution is the ideal free distribution, where consumer density matches local resource density. Alternatively, if wing wear is an important cost of wing use for bees, there should be relatively more bee consumers in high density patches, resulting in “overmatching”. If the energetic costs of flight influence patch choice, there also should be relatively more consumers in high density patches (where flight costs are lower), resulting in “overmatching”. To examine these hypotheses, I observed the visitation of bumble bees (genus Bombus) across experimental manipulations of plant density in patches of bergamot (Monarda fistulosa) in southwest Alberta, Canada over 2 summers. Bees flew more often and for a greater proportion of time, and hit their wings more frequently, in low-density patches. In 2004, bees “undermatched” resources (relatively few bees...

This study examines how, over the short term, logging affects the density of bumble bees (Apidae: Bombus), the understory plants commonly visited by bumble bees, and the numerical relationship between bumble bees and flowers. In the summers before and after winter logging, bumble bees and plants were surveyed in 50 deciduous stands (each of 8–10 ha) in the boreal forest of northern Alberta, Canada. Logging was replicated at three different intensities: 0, 10–20, and 50–75% of trees remaining. There were generally more bumble bees, species of bumble bee-visited plants, and flowers in moderately (50–75%) logged sites, but this pattern depended on the time of year. Before logging, bumble bees matched resources according to an ideal free distribution (IFD). Logging affected the distribution of bumble bees across floral resources: the slope of the regression relating bumble bee and flower proportions was less than one for clearcut and control treatments (i.e., undermatching), with too many bumble bees in the flower-poor compartments and too few in the flower-rich ones. Deviations from an IFD were negative in control sites, such that fewer bumble bees occurred here than warranted by flower numbers. Controlling for flower density, bumble bee density was significantly greater in clearcuts than in the other treatments. By disproportionately visiting plants in clearcuts (relative to flower density), and by undermatching, bumble bees in clearcuts should experience higher levels of competition. Conversely, the fewer (and undermatching) bumble bees in control sites (relative to flower abundances there) may cause these plants to obtain diminished pollination service. The proximity of clearcut logging to pristine areas may therefore negatively impact plants and bumble bees in the pristine areas, at least in the season immediately following logging.

Aim We test the hypothesis that tarsus length in all shorebirds breeding in the Canadian arctic shows an evolutionary response to average metabolic stress encountered across the breeding range, such that birds nesting in metabolically stressful environments have relatively shorter legs. Longer-legged birds living in colder environments will experience greater metabolic costs because their torsos are elevated farther away from the ground's wind-dampening boundary layer. Methods We use weather data (temperature, wind speed, global solar radiation) from 27 arctic weather stations measured over 37 years, and a previously published model of heat transfer, to characterize the metabolic harshness over the breeding season of the ranges of each of the 17 shorebirds of the family Charadriidae nesting in the Canadian arctic. Results After controlling for the lengths of two other body extremities (wing and bill), there was a significant negative relationship between tarsus length and mean metabolic harshness. This result was obtained whether species were treated as independent data points, or in a comparative analysis using standardized independent contrasts. Main conclusions We support a unique extension of Allen's rule: body-supporting appendages of homeotherms may be shorter in colder environments so as to take advantage of a boundary layer effect, thereby reducing metabolic costs.

Summary 1. Pollinators typically visit more flowers on plants with larger floral displays, which should present such attractive plants with significant pollen transport losses. 2. Many-flowered plants with hermaphrodite flowers might reduce the costs of attractiveness by encouraging fewer sequential flower visits by pollinators. One mechanism that accomplishes this is to produce variable nectar rewards, which will cause risk-averse foragers to leave the plant after visiting fewer flowers. 3. We test the prediction that within-plant variability in nectar production rate increases with the relative number of open flowers on a plant. A field survey of nine herbaceous angiosperm species native to Alberta, Canada revealed a significant positive correlation between nectar variability (measured as standard deviation) and the size of the floral display within species. This relationship existed over and above the null expectation of a positive correlation between mean and SD. 4. Our results suggest that multiflowered plants might maximize the male fitness returns associated with a plant’s attraction status (determined by relative display size), by taking advantage of risk-averse foraging by their pollinators.

Most hermaphroditic, many-flowered plants should suffer reduced fitness from within-plant selfing (geitonogamy). Large inflorescences are most attractive to pollinators, but also promote many flower visits during a single plant visit, which may increase selfing and decrease pollen export. A plant might avoid the negative consequences of attractiveness through modification of the floral display to promote fewer flower visits, while retaining attractiveness. This report shows that increasing only the variance in nectar volume per flower results in fewer flower visits per inflorescence by wild hummingbirds (Selasphorus rufus) and captive bumble bees (Bombus flavifrons) foraging on artificial inflorescences. Inflorescences were either constant (all flowers contained the same nectar volume) or variable (half the flowers were empty, the other half contained twice as much nectar as in the constant flowers). Both types of inflorescence were simultaneously available to foragers. Risk-averse foraging behaviour was expressed as a patch departure preference: birds and bees visited fewer flowers on variable inflorescences, and this preference was expressed when resource variability could be determined only by concurrent sampling. When variance treatments were clearly labelled with colour and offered to hummingbirds, the departure effect was maintained; however, when preference was measured by inflorescence choice, birds did not consistently prefer to visit constant inflorescences. The reduced visitation lengths on variable inflorescences by both birds and bees documented in this study imply that variance in nectar production rates within inflorescences may represent an adaptive trait to avoid the costs of geitonogamy.

Interaction webs, or networks, define how the members of two or more trophic levels interact. However, the traits that mediate network structure have not been widely investigated. Generally, the mechanism that determines plant-pollinator partnerships is thought to involve the matching of a suite of species traits (such as abundance, phenology, morphology) between trophic levels. These traits are often unknown or hard to measure, but may reflect phylogenetic history. We asked whether morphological traits or phylogenetic history were more important in mediating network structure in mutualistic plant-pollinator interaction networks from Western Canada. At the plant species level, sexual system, growth form, and flower symmetry were the most important traits. For example, species with radially symmetrical flowers had more connections within their modules (a subset of species that interact more among one another than outside of the module) than species with bilaterally symmetrical flowers. At the pollinator species level, social species had more connections within and among modules. In addition, larger pollinators tended to be more specialized. As traits mediate interactions and have a phylogenetic signal, we found that phylogenetically close species tend to interact with a similar set of species. At the network level, patterns were weak, but we found increasing functional trait and phylogenetic diversity of plants associated with increased weighted nestedness. These results provide evidence that both specific traits and phylogenetic history can contribute to the nature of mutualistic interactions within networks, but they explain less variation between networks.

Co-flowering plant species commonly share flower visitors, and thus have the potential to influence each other's pollination. In this study we analysed 750 quantitative plant–pollinator networks from 28 studies representing diverse biomes worldwide. We show that the potential for one plant species to influence another indirectly via shared pollinators was greater for plants whose resources were more abundant (higher floral unit number and nectar sugar content) and more accessible. The potential indirect influence was also stronger between phylogenetically closer plant species and was independent of plant geographic origin (native vs. non-native). The positive effect of nectar sugar content and phylogenetic proximity was much more accentuated for bees than for other groups. Consequently, the impact of these factors depends on the pollination mode of plants, e.g. bee or fly pollinated. Our findings may help predict which plant species have the greatest importance in the functioning of plant–pollination networks

The costs and benefits associated with feeding in the vicinity of a predator should vary within and among species, depending on an individual's vulnerability to a predator. In this paper, we investigate how willingness to risk exposure to a predator in order to gain access to food is expressed within groups of fish. We provided groups of six individually marked brook sticklebacks (Culea inconstans) and fathead minnows (Pimephales promelas) with the opportunity to feed in safety, or in the presence of a single predator (walleye, Stizostedion vitreum). Sticklebacks and fathead minnows are commonly found in the same locations at the same time, but sticklebacks also possess armour and spines that provide defence against some predators. Despite this morphological variation, patterns of individual behaviour were consistent between species. While feeding close to the predator, there was a positive relation between the size of the individual and the proportion of food consumed. No relation existed on the side farthest from the predator. These data suggest that fish may risk exposure to predators to exploit size-related differences in vulnerability to gain a competitive advantage for access to food.

Sexual spores (i.e., basidiospores) of the spruce cone rust (Chrysomyxa pirolata; Uredinales) produced on leaves of a small perennial herb, the common pink wintergreen (Pyrola asarifolia; Pyrolaceae), must travel from the forest floor to the forest canopy, where they infect the female cones of white spruce (Picea glauca; Pinaceae). Asexual spores (i.e., urediniospores) must move among pyrola plants on the forest floor. Spore transmission is complicated by relatively calm wind conditions within the forest, and the short critical period during which the cones of white spruce can be infected. Here, we document a correlation between the orientation of leaves in the pyrola and presence of rust pustules on the leaf’s ventral surface. We experimentally address the consequences to spore dispersal of this change in leaf position and presence of pustules. Transmission of rust spores can increase with higher leaf angles in two ways: (i) by raising the pustules further above the ground surface, thereby increa...

Foraging organisms (like bumble bees) move between resource points (like flowers) whose natural distributions vary enormously: from hyperdispersed to random to clumped. These differences in habitat structure may significantly influence the fitness of both plant and pollinator. To examine the effect of habitat structure on pollinator movement and fitness, we observed captive worker bumble bees collecting nectar from artificial flowers containing equal volumes of reward and arranged in two spatial configurations: a hexagonal array with constant distances between flowers ("constant"), and an "exploded hexagonal" array, with variable distances between flowers ("variable"). The mean nearest-neighbour distance was the same in both arrays, as was the general hexagonal appearance. The experiment therefore compares how resource dispersion, independent of nearest-neighbour distance, influences bee behaviour. Bees in the variable array showed decreased directionality, higher revisitation frequencies, and greater inter-flower flight distances than shown in the constant array. As a consequence, bees in the variable array had a 19% lower gross rate of nectar collection. Our results suggest that wild-foraging bees should prefer regularly spaced flowers (when all else, including mean nearest-neighbour distance, is equal), and that plants can decrease self-pollination by regular spacing between flowers, inflorescences, or individuals.

It is often desirable to estimate the metabolic costs incurred by homeothermic organisms of differing morphology living in different real or hypothetical environmental conditions. To address this problem, we describe a method, based on previously published empirical allometric and heat-transfer equations, that allows a rough estimate to be made of the daily maintenance metabolic costs (i.e., basal and thermoregulatory costs) incurred by a bird in a simple cold two-dimensional environment. The model uses widely available weather variables (temperature, wind speed, and global solar radiation), morphological variables (body mass, height of body's centre of gravity, diameter of torso), and a habitat variable (height of vegetation). We apply the model to weather data from the Canadian Arctic to predict daily metabolic costs for two calidridine sandpiper species (Calidris canutus and C. minutilla) during the summer. The model is extremely sensitive to error in the slope and intercept of the allometric equation predicting conductance from body mass, but is generally robust to other model parameters. Using ambient temperature (Ta) in place of operative temperature (Te) has only a minor (3.5%) effect on predicted metabolic costs, so, given that Teis difficult to estimate, we recommend this substitution (at least for arctic latitudes, where solar radiation is of reduced importance). The model predicts metabolic rates similar to those obtained from an equation based on a heated taxidermic mount for C. canutus, thereby providing some measure of validation. The model can easily be modified to predict metabolic costs for other groups of birds or mammals.

Some insects construct cocoons in which pupae complete their metamorphosis to adult form. Despite a variety of proposed benefits, the functional significance of the cocoon remains unclear. Here we experimentally examine the functional significance of the cocoon in two species of moth (Gynaephora rossii Curtis and G. groenlandica Wocke, Lepidoptera: Lymantriidae) inhabiting a cold and thermally stressful arctic environment. G. rossii larvae spin a single-layered dark cocoon, and G. groenlandica spin a double-layered pale cocoon, consisting of an inner dark layer and an outer translucent pale layer. Comparisons of developmental times of pupae in cocoons and naked pupae whose cocoons had been experimentally removed revealed that the pale cocoons of G. groenlandica accelerated pupal development substantially. The warming effect of sun was seemingly the principal influence accounting for the enhanced development. In contrast, the presence of a dark cocoon of G. rossii did not accelerate development detectably. To generate hypotheses for why these two closely related species might differ in cocoon type and function, we examined ecological correlates associated with each cocoon type. The experimental results for G. groenlandica provide clear evidence that cocoons enhance a correlate of fitness in insects.

Typically, tests of risk-sensitive foraging involve observing a subject's choices of alternative prey types differing in some combination of mean and variance of expected foraging gain. Here, we consider the problem of risk-sensitive foraging when there is a single prey type. We observed worker bumble bees ( Bombus occidentalis ) foraging in an array of artificial 2-flower inflorescences. After visiting the bottom flower in an inflorescence and obtaining a reward of some size, the bee decides whether to visit the top flower or to move to a new inflorescence (a patch departure). Here, risk-sensitive behavior is expressed as the forager's choice of patch departure threshold (PDT) of reward obtained in the bottom flower. We measured the PDTs of bees whose colony energy stores (and therefore energy requirements) had been manipulated (Enhanced or Depleted). Simulations led us to predict that shortfall-minimizing bees should decrease their PDTs when their colony energy reserves were depleted, relative to when the reserves were enhanced. Bees did not use a strict patch departure threshold, but instead the probability of departure varied with nectar volume in the bottom flower. Colony energy stores did affect patch departure behavior, but this effect was confounded by the order in which manipulation of colony reserves was applied. Further, simulations of observed bee patch departure decisions did not produce behavior expected if the decisions were based on shortfall-minimization. We conclude that a bee's decision of when to leave an inflorescence is not predicted by a static shortfall-minimizing model. Our results also implicate an important interaction between learning and foraging requirements. We review risk-sensitivity in bees, and discuss why risk-sensitive foraging may be adaptive for bumble bees.

This study relates foraging behaviour and division of labour to a colony's energy state in wild-foraging bumblebee colonies whose energy stores had been experimentally manipulated. It examines: (1) colony-level effects of energy manipulation, to see if shifts in a colony's energy requirements produce shifts in the behaviour of its workers; and (2) the behaviour of individuals who changed tasks following manipulation and those that did not, to see if task specialization, based on learning, can explain differences between these groups. Individually marked bumblebees were observed at three colonies of each of three species ( Bombus flavifrons, B. melanopygus , and B. occidentalis ). Overall, colonies responded to changes in their energy stores, both by adjusting the relative sizes of the foraging and within-nest work-force, and by changing the behaviour of individual foragers. In response to energy manipulations, colonies allocated effort where demand was relatively greatest (foraging versus within-nest tasks), collected more of the resource that was in greater immediate demand (nectar versus pollen), and changed the rate at which each forager collected food (within-nest bout durations). Task switching behaviour was generally consistent with that expected by learning-constrained task specialization: (1) less efficient foragers switched to withinnest tasks when energy stores were enhanced; (2) workers who began foraging after energy depletion were less efficient than were continuing foragers and (3) workers who switched from pollen to nectar collection following energy depletion were less efficient foragers both before and after the manipulation. However, workers who switched from nectar to pollen collection were no less efficient at their old or new tasks than non-switching bees. In general, worker bumblebees show a flexible division of labour in the face of short-term changes in their colony's energy state.

Risk-sensitive foraging decisions (i.e., those based on both reward mean and variance) have often been observed in laboratory contexts, but there are few observations of risk-sensitive foraging in the wild. Indeed, wild foragers need not show risk sensitivity, since they can adopt alternative tactics for minimizing their probabilities of energy shortfall. I tested for risk-sensitive foraging in a natural context in coastal southwest British Columbia, Canada. Seablush (Plectritis congesta) and dwarf huckleberry ( Vaccinium caespitosum) are co-occurring plants that offer foraging bumble bees equivalent expected rates of net energy intake, but that differ in variability. This equalization of mean profitabilities of floral rewards results from the actions of nectar-collecting foragers. Risk-sensitive diet choice can therefore be based solely on the variances of the foraging alternatives. I manipulated the energy requirements of bumble bee colonies (Bombus melanopygus, B. mixtus, B. sitkensis) by draining or enhancing their honey pots. I then censused foragers from these colonies on the two flower species. If bees were risk sensitive in their diet choice, they should have increased their relative use of the more variable flower type (dwarf huckleberry) when their colonies were nectar depleted compared to when their colonies were nectar enhanced. This was indeed the case. Wild-foraging bumble bees thus appear to be sensitive to both the mean and the variance of the energy rewards offered by alternative flower species, and this sensitivity is affected by colony energy requirements. These results suggest that wild foragers may change their risk sensitivity.

Bumble bees rely on stored nectar to maintain high colony temperatures. This study examines some of the costs associated with exhausting stored nectar for a day in confined colonies of Bombus occidentalis Greene and B. melanopygus Nylander. Workers from energy-deprived colonies ceased incubating and allowed brood temperatures to drop to ambient levels. Workers from energy-rich colonies responded to a simulated vertebrate predator by actively moving about and buzzing loudly (apparently searching for the source of disturbance), but those from energy-depleted colonies mostly remained stationary on the comb. Workers from energy-rich colonies responded to an intruding social parasite (a Psithyrus insularis Smith queen) by chasing and attacking it, whereas those from energy-depleted colonies adopted stationary threat postures. In both cases, the more successful defensive strategy of simultaneous attack by several workers was not employed by nectar-depleted colonies. Cooling of final-instar larvae and pupae added to their development times (disproportionately so in the case of pupae), but low temperature per se did not affect their probability of survival. Hence, the costs of short-term energy shortfall include increased susceptibility to predators and parasites and a lengthened period of development. These costs probably relate to energy stores in a nonlinear manner, providing the foundation upon which risk-sensitive foraging decisions can be based.

This study examines whether the foraging behavior of worker bumble bees (Bombus: Apidae) collecting nectar on inflorescences of seablush (Plectritis congesta: Valerianaceae) is affected by colony energetic requirements, which were experimentally manipulated either by adding sucrose solution to honey pots or by removing virtually all available nectar from the pots. The competing hypotheses tested were: (1) no change; energetic requirements do not affect behavior, since there is a single best way to collect food in a given environment; (2) energetic currency; the energetic currency maximized by foragers changes according to colony energetic condition, with nectar-depletion causing a shift from maximizing long-term productivity to maximizing immediate energetic gain, thereby de-emphasizing energetic costs; and (3) predation; foragers devalue risk of predation as risk of starvation increaes, with colony nectar-depletion causing foragers to be less predation riskaverse in order to increase immediate energetic gain. Relative to when their colony energy reserves were enhanced, foragers from nectar-depleted colonies selected smaller inflorescences, visited fewer flowers per inflorescence, probed flowers at a higher rate while on each inflorescence, and walked between inflorescences less often, thereby spending a greater proportion of their foraging trip in flight. These behaviors increased a bee's energetic costs while foraging, and should also have increased its immediate energetic gains, allowing rejection of the no change hypothesis. Predictions of the predation hypothesis were generally not supported, and our results best support the energetic currency hypothesis. Foraging currency of bumble bees therefore appears to be a function of colony energetic state.

Risk-sensitive behavior is the phenomenon of animals exhibiting preferences when offered choices whose outcomes differ in their degree of variance. It is now well demonstrated that many animals are sensitive to differences in the magnitude of variance in rewards. A considerable body of literature on risk sensitivity has been generated in recent years, encompassing theoretical and empirical work, from both functional and mechanistic approaches. Nevertheless, it is less than clear how animal preferences for variability are determined. It is therefore timely to look broadly at what has been accomplished, and to identify the most promising synthetic approaches for future work.

In a controlled laboratory experiment, we re-examined the question of bumble bee risk-sensitivity. Harder and Real's (1987) analysis of previous work on bumble bee risk aversion suggests that risk-sensitivity in these organisms is a result of their maximizing the net rate of energy return (calculated as the average of expected per flower rates). Whether bees are risk-sensitive foragers with respect to minimizing the probability of energetic shortfall is therefore still an open question. We examined how the foraging preferences of bumble bees for nectar reward variation were affected by colony energy reserves, which we manipulated by draining or adding sucrose solution to colony honey pots. Nine workers from four confined colonies of Bombus occidentalis foraged for sucrose solution in two patches of artificial flowers. These patches yielded the same expected rate of net energy intake, but floral volumes were variable in one patch and constant in the other. Our results show that bumble bees can be both risk-averse (preferring constant flowers) and risk-prone (preferring variable flowers), depending on the status of their colony energy reserves. Diet choice in bumble bees appears to be sensitive to the “target value” a colony-level energetic requirement.

As is obvious from the papers in this symposium, the field of risk sensitivity is a chimera. Risk sensitivity applies to a wide diversity of phenomena: from foraging, to reproduction, to habitat choice, to social behavior. Many more phenomena, such as mate choice, territoriality, and plant foraging, remain to be explored from a risk-sensitive perspective. What do these studies have in common? At the broadest level, they belong to a branch of evolutionary biology that deals with the response of organisms to stochastic environments. In particular, they address the case of organisms responding to stochasticity at a fine-grained level. The common denominator in all studies of risk sensitivity is within-generation decision-making where outcomes are not uniquely determined by mean expectations, but are also influenced by variance about the mean.

Much empirical work has examined whether foraging animals distribute themselves across patches of resources according to an ideal free distribution (IFD). Most studies consider situations with 2 patches, where choice of one patch is synonymous with rejection of the alternative. In this paper, we use computer simulations to determine the effect of the number of patches on achievement of an IFD. Identical foragers were allowed to move between patches so as to maximize their intake rate. To generate departures from an IFD, we assigned all foragers a perception limit of prey numbers such that they could not perceive differences in patch quality that exceed this limit. Foragers would choose randomly from the patches among which they could perceive no differences. We found that, for the same perception limit, departures from an IFD were greatest in the 2-patch situation. In addition, we are less likely to detect perception limit-generated deviations from an IFD (or even a random distribution) in greater than 2-patch situations unless we examine cases in which there are large among-patch differences in quality. Hence, lab and field studies that incorporate the greater realism achieved by using more than 2 patches must be cautious about falsely accepting the hypothesis of an IFD, particularly if the causes of deviations from an IFD are independent of the

Pollinators usually maintain directionality while moving through a flower patch, presumably to increase foraging success by minimizing revisitation of previously- emptied flowers. Two alternative directionality-generating mechanisms have been proposed: foragers may keep track of their arrival direction at a flower or cluster of flowers and prefer to depart in the same direction (the arrival hypothesis), or they may show little change in orientation while foraging at a flower or flower cluster and prefer to depart in the direction they are last facing (the last-faced hypothesis). Rotation of a flower or inflorescence while the forager is feeding on it will unequivocally distinguish the two hypotheses

1. The insect wing appears to degenerate with use, so that wing wear increases with increased flight activity. Wing degeneration may affect an insect's mortality. 2. The rate of mortality is known to increase with age in worker bumble bees. This study examines whether wing wear can account for mortality in foraging bumble bees. 3. Workers in eight wild-foraging colonies of bumble bee (Bombus melanopygus Nylander) were divided into two treatment groups: clipped (the outer margin of each forewing trimmed, reducing wing surface area by an average of 18%) and unclipped controls. The mortality and behaviour of foraging individuals was assessed with colony watches and night censuses. 4. Treatment did not detectably affect the proportion of foraging bees, the lengths of foraging or within-nest bouts of foragers, or pollen load sizes, but wing-clipping did cause bees with the greatest amounts of initial wing wear to stop foraging. 5. Mortality was positively related to natural wing wear among unclipped bees. In addition, relative to control foragers, foragers with clipped wing margins experienced significantly reduced life expectancies. 6. These results support the hypothesis that wing wear is a proximate factor responsible for an increase in mortality in older workers. The wing-wear hypothesis therefore provides one functional reason for foragers to adopt a currency that maximizes the ratio of net benefit to cost (i.e. 'efficiency').

Forty species of birds were observed during field studies at Sarcpa Lake, Melville Peninsula, N.W.T. during the summers of 1981 and 1982. Evidence of breeding was found for 22 species and the first definite breeding records for the Melville Peninsula were obtained for Glaucous Gull, Pectoral Sandpiper and Semipalmated Sandpiper and Semipalmated Sandpiper. A hybrid pair of gulls (Glaucous x Herring Gull) also nested and this appears to be the first definite North American record of such a hybrid nesting. Fourteen additional breeding species expected to be present, based on maps in Godfrey (1966), were not found. Average breeding bird density (35 pr/sq km) was comparable to that on Bylot Island, but considerably higher than that measured at other High Arctic sites. Neither average breeding bird densities nor phenologies changed appreciably from year to year despite a late spring melt in 1982. In both years birds began their breeding activities as soon as suitable nesting and feeding habitat became available. A biogeographic analysis based on the occurrence of breeding birds at 25 other sites across the Canadian Arctic indicates that the avifauna at Sarcpa Lake is more similar to those of High Arctic island sites than to those of mainland sites, but includes none of the species whose ranges are mainly within the Arctic Archipelago. Key words: arctic biogeography, birds, phenology, nesting density, Melville Peninsula, multivariate analysis

The factors affecting the evolution of avian leks are not clearly understood (Bradbury and Gibson 1983). Leks are communal displays where males congregate to display to females, which come to obtain matings (Emlen and Oring 1977). Since a few males on leks account for most of the matings, a lek is a type of polygyny. In another type of polygyny, resource defense polygyny, males obtain matings by defending resources that females require for reproduction (e.g., nest sites, feeding areas). Economic defensibility of resources is thought to be the major ecological factor favoring resource defense polygyny over lekking (Emlen and Oring 1977). Although males clearly do defend resources in many polygynous species, this does not demonstrate that leks result when resources or females are not defensible

We studied the nest attentiveness of six female White-rumped Sandpipers (Calidris fuscicollis), a species with uniparental care which incubates in the continuous daylight of the arctic summer. We used correlation and multiple regression analyses to determine how well day-to-day variation in recess time/day, number of trips/day, and average trip length/day could be explained by date, concurrent weather, behavior on the previous day, and weather on the previous day. Both previous and concurrent weather were important predictors of incubation behavior, while date and previous behavior were not. We therefore conclude that incubation behavior on a given day is not simply a function of current conditions. Behavior appears at least to integrate the effects of both present weather and weather on the previous day.