Abstract: 1. The recent recovery of the wolf in southern Europe has not yet removed the risk
of local extinction. Wolf populations are fragmented and often comprise fewer than
500 individuals. In North America, northern and eastern Europe, wolves feed maiiily
on wild herbivores. In southern Europe, this canid has apparently adapted to feed
also on fruit, rubbish, livestock, small and medium-size mammals.
2. The main conservation problem lies with predation o n domestic ~ingulates,w liich
leads to extensive killing of wolves. The reintroduction of wild large herbivores has
been advocated as a means of reducing attacks on livestock, but predatiori on the
latter may remain high if domestic ungulates are locally abundant.
3. Our synthesis of 15 studies, published in the last 15 years, on food habits of the
wolf in southern Europe, has shown that ungulates have been the main diet component
overall. A significant inverse correlation was found between the occurrence (%) of
wild and domestic ungulates in the diet. The presence of relatively few wild ungulate
species was necessary to reduce predation on livestock.
4. Selection of wild and domestic ungulate prey was influenced mainly by their local
abundance, but also by their accessibility. Feeding dependence on rubbish was local
and rare. In Italy, the consumption of riibbish/fruit and that of ungulates was significantly
negatively correlated. Diet breadth increased as the presence of large prey
in tlie diet decreased.
5. The simultaneous reintroduction of severa1 wild ungulate species is likely to reduce
predation on livestock and may prove to be one of the most effective conservation
measures.

Abstract: We review socially biased learning about food and problem solving in monkeys, relying especially on studies with tufted capuchin monkeys (Cebus apella) and callitrichid monkeys. Capuchin monkeys most effectively learn to solve a new problem when they can act jointly with an experienced partner in a socially tolerant setting and when the problem can be solved by direct action on an object or substrate, but they do not learn by imitation. Capuchin monkeys are motivated to eat foods, whether familiar or novel, when they are with others that are eating, regardless of what the others are eating. Thus, social bias in learning about foods is indirect and mediated by facilitation of feeding. In most respects, social biases in learning are similar in capuchins and callitrichids, except that callitrichids provide more specific behavioral cues to others about the availability and palatability of foods. Callitrichids generally are more tolerant toward group members and coordinate their activity in space and time more closely than capuchins do. These characteristics support stronger social biases in learning in callitrichids than in capuchins in some situations. On the other hand, callitrichids' more limited range of manipulative behaviors, greater neophobia, and greater sensitivity to the risk of predation restricts what these monkeys learn in comparison with capuchins. We suggest that socially biased learning is always the collective outcome of interacting physical, social, and individual factors, and that differences across populations and species in social bias in learning reflect variations in all these dimensions. Progress in understanding socially biased learning in nonhuman species will be aided by the development of appropriately detailed models of the richly interconnected processes affecting learning.

Abstract: Two chimpanzees and a rhesus macaque rapidly learned the ordinal relations between 5 colors of containers (plastic eggs) when all containers of a given color contained a specific number of identical food items. All 3 animals also performed at high levels when comparing sets of containers with sets of visible food items. This indicates that the animals learned the approximate quantity of food items in containers of a given color. However, all animals failed in a summation task, in which a single container was compared with a set of 2 containers of a lesser individual quantity but a greater combined quantity. This difficulty was not overcome by sequential presentation of containers into opaque receptacles, but performance improved if the quantitative difference between sizes was very large.

Abstract: Domestic hens were trained to run a Y-maze and make an association between differently coloured doorways and access to food pellets or sand. The hens were tested for their choice of doorway when the goals were not visible from the choice point and when they were food or sand deprived. Hens made the choice appropriate to their deprivation state (correct choice) significantly more often for food than sand and were faster at choosing and entering the goal box when food deprived. In a follow up experiment, the goals were visible from the choice point. Again the hens chose correctly significantly more often when food than sand deprived and made the choice and entered the goal box faster when food deprived. Thus, failure to choose sand in the first experiment was not due to an inability to learn the association, but appears to result from a strong motivation to feed in the Y-maze, even when not food deprived, and a weak motivation to dustbathe or forage, even when sand deprived.