Search

Innovative foraging behaviour has been observed in many species, but little is known about how novel behaviour emerges or why individuals differ in their propensity to innovate. Here, we investigate these questions by presenting 36 wild-caught adult male Carib grackles (Quiscalus lugubris) with a novel problem-solving task. Twenty birds solved the task ("innovators") while 16 did not ("non-innovators"). We compared innovators to non-innovators and explored variation in latency to innovate to determine the characteristics of an innovative bird. Innovativeness was not predicted by any morphological trait, but innovators had higher exploration scores and lower object neophobia scores than non-innovators. Within the innovators, latency to innovate was positively correlated with learning speed. Video analysis also revealed a marked difference in the way individuals interacted with the novel apparatus: when innovators contacted the correct part of the apparatus, they continued to do so until they solved the problem. Non-innovators often contacted the correct part of the apparatus, but did not persist in doing so. The importance of obstacle movement cues was confirmed by an experiment where they were manipulated., (Copyright © 2011 Elsevier B.V. All rights reserved.)

Animals' ability to adjust their behaviour when environmental conditions change can increase their likelihood of survival. Although such behavioural flexibility is regularly observed in the field, it has proven difficult to systematically quantify and predict inter-individual differences in free-living animals. We presented 24 Zenaida doves (Zenaida aurita) on 12 territories with two learning tests in their natural habitat in Barbados. The dove pairs showed high site fidelity and territoriality, allowing us to test individuals repeatedly while accounting for the effects of territorial chases and pair bonds on our learning measures. We used a foraging apparatus that enabled Zenaida doves to access seed, yet excluded other species, and measured doves' performance on colour discrimination and reversal learning tests. We found that (1) doves on all 12 territories passed the two tests; (2) mates within a pair were consistently solvers or scroungers; (3) sex, body condition and territorial chases did not consistently affect learning rates; (4) tameness was a significant negative predictor of learning to feed from the foraging apparatus and (5) scrounging within pairs seemed to facilitate learning. Our study presents a method to quantify intraspecific differences in behavioural flexibility in the field and relate these to individuals' physical and social traits., (Copyright (c) 2010 Elsevier B.V. All rights reserved.)

Dunking, the softening of dry food in water to speed up consumption time, is normally a very rare behaviour in wild Carib grackles (Quiscalus lugubris) of Barbados. Its frequency can be experimentally increased when large numbers of dry items are repeatedly placed near a standing source of water in conditions that minimize intraspecific competition and risk of theft. To reconcile the normally low frequency of the behaviour in the wild with the high rates obtained in previous experiments, we tested three conditions where dunking varied between 0 and 70%. Dunking was very rare when it had been made unnecessary by pre-soaking the food, water was far from the dry items offered and only one food item was given, focusing all competitive interactions and theft attempts on a single individual. In contrast, dunking rate was high when food was not pre-soaked, water was close to dry food and more than one item (and hence target for competition and theft) was given. These experiments confirm that dunking rates, like other proto-tool-like food-processing techniques, depend on the costs and benefits of the situation where they are used.

The use of tool or tool-like food processing behaviours can render animals vulnerable to theft (kleptoparasitism) because (1) large, nutritious items are usually involved, (2) value is added to the food due to long and/or complex handling, and (3) physical control of items is often temporarily lost during handling. In Barbados, Carib grackles (Quiscalus lugubris) immersing items in water before consumption (a behaviour known as food dunking) lose a larger proportion of items to conspecific food thieves than grackles that do not dunk. In this paper, we first show that dunking in Carib grackles functions as a proto-tool food-processing technique that speeds up ingestion. We then examine five potential predictors of kleptoparasitism: only conspecific density and loss of physical control on food were found to influence the probability that birds would be attacked and successfully robbed of food by conspecifics. Grackles could reduce the probability of kleptoparasitism by holding items in the bill while dunking and engaging in head-up displays. These behaviours were used flexibly depending on variation in the risk of kleptoparasitism. We suggest that costs like the ones incurred from theft might limit the profitability and frequency of tool and proto-tool food processing behaviours, creating a context where counter-strategies might be selected.

Parent-offspring transmission is usually thought to be the major route for cultural learning. We tested this assumption in ringdoves (Streptopelia risoria), a flock feeding Columbid that is easily raised in captivity. An aviary study first determined the foraging associations of juveniles placed with their sibling, their parents and a pair of non-kin adults. Juveniles foraged more often with kin than with non-kin and joined food discoveries in proportion to these foraging associations; aggression was relatively rare and came more often from unrelated adults than from parents. Two cage experiments showed that parents were not copied more often than unrelated adults when the two tutor types provided different, but equally productive, solutions to a feeding problem. Neither the ingestion of unfamiliar food (two seed types the juveniles had never encountered) nor the learning of a new food searching skill (opening a box containing seed) showed a differential effect of father versus non-kin tutors. Animals that scramble compete for food may thus acquire social information from whatever knowledgeable individuals are present, whether these are kin, unrelated conspecifics or heterospecifics.

Social foraging can inhibit the learning and performance of food-finding behaviours. Confusion, overshadowing and frequency-dependent payoffs may all contribute to the inhibition, but standard experimental procedures make the separation of these effects difficult. In this study, we combine characteristics of cage and aviary experiments and present either a single naive pigeon or groups of three naive pigeons with a pre-trained producer opening an apparatus in an aviary. All naive birds scrounged on the 3456 openings they witnessed. In a post-test given in the absence of other birds, all single scroungers opened the apparatus, but only one of the group-scrounging pigeons did. Scrounger numbers appear to play an important role in the inhibition of food-finding behaviour, suggesting that confusion is a major component of learning in a social context.

Interspecific variation in learning and cognition is often accounted for by adaptive specialization, an ecological framework where variation between species in the environmental problems they face is thought to select for quantitatively and/or qualitatively different abilities. Adaptive specialization theory relies on the comparative method for testing its hypotheses and assumes a naturally selected basis for the predicted differences. This review examines social learning as a specialization to group-living and scramble feeding competition. It points out one important problem with current studies in the area, the lack of quantitative controls for confounding variables that may cause type 1 or 2 error in comparative tests. A linear regression technique is proposed to measure and remove interspecific differences on control tests for which there is no predicted adaptive specialization; as in other areas of comparative biology, the adaptive prediction is then made on the residual deviation from the regression of these confounding variables. Examples are given from research on opportunistic Columbids, the group-living feral pigeon Columbia livia, and the territorial Zenaida dove, Zenaida aurita.

One of the most widely cited cases of cultural transmission in animals is the opening of milk bottles by British birds. Bottle opening was first reported in Swaythling in 1921 and its spread from that date to 1947 was mapped by Fisher and Hinde (1949). Using data from Fisher and Hinde, this paper tests two quantitative models of cultural transmission: (1) the logistic model describing the cumulative number of bottle opening sites in the Belfast area and in the whole of the UK, and (2) the linear wave-of-advance model describing the progressive spread over time of bottle opening from its presumed single point of origin, Swaythling. For both the UK and Belfast, the logistic provides a poorer fit than alternative accelerating functions: the positive exponential and the reverse S-shaped hyperbolic sine respectively yield the best corrected fits to the Belfast and UK data. Neither functions have the terminal deceleration phase typical of the logistic, but both have an accelerating phase consistent with the cultural assumption of an auto-catalytic increase in the rate of spread over time. For both Belfast and the whole UK, the wave-of-advance model can clearly be rejected, even with the addition of a second source of innovation in County Durham. The results support the view that bottle opening originated from several independent sites, but spread through an accelerating process that could have included direct and/or indirect social influences.

Grooming behaviours in the veery (Catharus fuscescens ) and the American robin (Turdus migratorius ) cluster into four hierarchical groups based on function and anatomy: (1) non-preening, (2) upper body and wing preening, (3) lower body cleaning, and (4) tail preening in the veery; (1) non-preening, (2) body preening, (3) wing preening and (4) tail preening in the robin. The frequency distributions of the number of consecutive acts per cluster and the number of cluster visits per bout approximate geometric distributions in both species. These results provide evidence for the generality of the timing hierarchy model., (Copyright © 1986. Published by Elsevier B.V.)

Structural rules for grooming are examined in the budgerigar (Melopsittacus undulatus). Two types of bouts are found: short bouts with no predictable temporal pattern and long bouts with some degree of periodic occurrence. Transitions in long bouts are mostly based on anatomical proximity. Grooming movements cluster together according to function and body region. Evidence for hierarchical organization obtained by clustering is confirmed by the proportion of occurrence of movements in long and short bouts. Individual birds are similar enough on measures of occurrence, transition and hierarchica structure for pooling of subjects to be justified; this suggests that the organizational rules found here are to a large extent species-typical., (Copyright © 1982. Published by Elsevier B.V.)

Feral and laboratory flocks of rock doves (Columba livia ) show a pattern of grouped sequential exploitation when simultaneously presented with two dispersed, depleting patches of seed. This behavior contrasts with the ideal free distribution pattern shown when patches are small and concentrated. Grouped sequential exploitation consists of two phases: all pigeons first land together and feed at one patch, then leave one by one for the other patch. Departure times of individuals for the second patch are correlated with feeding rate at patch 1, which is in turn correlated with position in the dominance hierarchy. The decision to switch from patch 1 to patch 2 improves individual feeding rates in all cases, but is done slightly later than it should according to optimal foraging theory., (Copyright © 1985. Published by Elsevier B.V.)