A comparison of computer-generated and naturally occurring foraging patterns in route-network-constrained spider...

Area of Science:

• Primate behavior
• Computational modeling
• Ecology

Background:

• Attributing goal-directed behavior in primates requires ruling out simpler explanations.
• Computer simulations serve as null models to test complex cognitive foraging hypotheses.
• Comparing simulated non-goal-directed behavior with wild primate observations provides quantifiable data.

Purpose of the Study:

• To compare two computer-simulated null models of foraging behavior against observed spider monkey (Ateles belzebuth) foraging patterns.
• To determine if observed spider monkey foraging and ranging behavior can be explained by non-goal-directed travel models.

Main Methods:

• Two computer-simulated null models were developed: a modified random-walk model and a route-constrained model based on observed spider monkey travel.
• Simulated monkeys searched for feeding trees across varying densities (10 to 1,000).
• Travel distance, directness, and accuracy of direction were compared between models and with observed wild spider monkey data.

Main Results:

• Route-model simulations showed shorter travel distances and more direct paths to feeding trees compared to random-walk models.
• Route-model simulations also demonstrated higher accuracy in the initial direction towards discovered feeding trees.
• Observed wild spider monkeys significantly outperformed both simulated models across all measured variables (travel distance, directness, and accuracy).

Conclusions:

• The null hypothesis that observed spider monkey foraging and ranging behavior can be explained by non-goal-directed travel is rejected.
• Observed spider monkey foraging behavior is more efficient and accurate than predicted by non-goal-directed models, suggesting cognitive control.
• This study provides quantifiable evidence supporting goal-directed foraging strategies in wild spider monkeys.