A Hypothetical Modelling and Experimental Design for Measuring Foraging Strategies of Animals

Area of Science:

• Behavioral Ecology
• Computational Neuroscience
• Animal Cognition

Background:

• Animal foraging behavior is complex, influenced by memory and environmental factors.
• Modeling foraging strategies aids in understanding decision-making processes in animals.
• Distinguishing between short-term and long-term memory's role in foraging is crucial.

Purpose of the Study:

• To develop a modeling approach for capturing animal foraging behaviors.
• To construct and compare optimal foraging paths based on LMRFT and SMRFT.
• To assess agent intelligence using a devised structural metric and optimal parameter learning.

Main Methods:

• Devised a modeling approach based on LMRFT and SMRFT.
• Constructed optimal foraging paths for different strategy levels.
• Calculated structural distance between modeled and real foraging paths using Euclidean distance and ranked matrices.
• Employed optimal parameter learning to identify foraging strategies.

Main Results:

• Agent 3 exhibited the highest intelligence under LMRFT (strategy level 5).
• Agents 1 and 2 showed equal intelligence under SMRFT (strategy level 5), surpassing Agent 3 (strategy level 2).
• Simulated foraging data revealed distinct intelligence rankings depending on the memory model used.

Conclusions:

• The LMRFT and SMRFT models provide distinct insights into animal foraging intelligence.
• Agent intelligence is context-dependent, varying with the memory strategy employed.
• This modeling approach offers a framework for quantifying and comparing animal cognitive abilities in foraging tasks.