Primate cognition test battery in parrots

Abstract

Systematic, broad phylogenetic comparisons of diverse cognitive abilities are essential to understand cognitive evolution. Few studies have examined multiple skills comparatively, using identical tasks across species. Previous research centered on primates, but recent evidence suggests that complex cognition may have evolved in distantly related taxa. We administered the tasks of the primate cognition test battery (PCTB) to 4 parrot species for a first direct comparison with primates. The parrots did not perform significantly worse than the previously tested primates in all but one of the test scales, but remained at chance levels throughout. Chimpanzees outperformed them in the physical but not the social domain. No differences between the domains nor across the parrot species were detected. It remains questionable whether the chance level performance reflects the parrots’ cognitive capacity or results from task constraints, which would limit the suitability of PCTB for phylogenetic comparisons. Possible implications for the field are discussed.

The lemur baseline: how lemurs compare to monkeys and apes in the Primate Cognition Test Battery

“…These results support earlier studies indicating that lemurs do not differ from haplorhine primates in numerosities and simple arithmetic operations (Jones & Brannon, 2012;Merritt et al, 2011;Santos, Barnes & Mahajan, 2005a). Since a comparable numerical understanding as tested in the PCTB has also been reported for various taxa outside the primate order, including fish and insects (e.g., Agrillo et al, 2012;Chittka & Geiger, 1995;Pahl, Si & Zhang, 2013; but see Krasheninnikova et al, 2019), a basal numerical understanding may be present in many animals.…”

“…These results, which are based on a small sample size, reject the notion of a direct correlation between brain size and cognitive abilities assessed in the PCTB and, may question assumptions of domain-general cognitive skills in primates. Overall, our results strengthen the view that when comparing cognitive abilities among species, it is of vital importance to include a diverse set of tests from both cognitive domains that are applicable to a diverse range of species and taxa (Auersperg et al, 2011;Auersperg, Gajdon & von Bayern, 2013;Burkart, Schubiger & Van Schaik, 2016;Maclean et al, 2012;Schmitt, Pankau & Fischer, 2012) and to carefully consider the internal and external validity of the specific tests (Krasheninnikova et al, 2019;Schubiger, Fichtel & Burkart, 2020).…”

“…For example, performance in a memory task increased in common marmosets (Callithrix jacchus) and common squirrel monkeys (Saimiri sciureus) from a two-choice task to a nine-choice task, in which the probability of success was lowered from 50% to 11%, making a wrong choice more costly, appeared to favour an appropriate learning strategy over a random choice strategy (Schubiger, Kissling & Burkart, 2016). The application of a random choice strategy may also explain why four parrot species that were tested with the PCTB, may have failed to solve the tasks, besides morphological differences in performing the tasks (Krasheninnikova et al, 2019).…”

“…Hence, comparative studies of cognitive abilities, ideally using identical tests, across the primate order and beyond are required. However, comparisons of performance in cognitive experiments across species may fail due to variation in the experimental set-up and specific methods (Van Horik & Emery, 2011;Krasheninnikova et al, 2019;MacLean et al, 2009;Schubiger, Fichtel & Burkart, 2020).…”

Causal Understanding of the Stone Dropping Task in Two Species of Macaw

“…We tested two species of macaws: Ara glaucogularis and Ara ambiguus. These macaws had previously completed a range of causal understanding tasks as part of a larger cognitive test battery (Krasheninnikova et al, 2019) and also completed a modified trap-tube task (O'Neill et al, 2018). Both of these experiments were object choice tasks with two or three options to choose from and it turned out the subjects developed some rules of thumb (heuristics), such as side biases or random choice, that allowed subjects to attain a satisfactory amount of rewards without any 'cognitive effort'.…”