A Mathematical Modelling Approach to Functionally Defining Forest Edge and Its Utility for Primate Behavioural Edge Effects

With the majority of primates living close to forest edges, it is imperative to understand how edges affect primate distribution and behaviour. Definitions of edge, however, vary across studies, with many defining edge a priori and few explicitly measuring the depth of edge influence (DEI). In this study, we aimed to functionally define edge for mantled howler monkey ( Alouatta palliata ) distribution and howling behaviour in a fragmented rainforest. We surveyed howler monkey locations (walking 11 line transects a total of 253 times in May-August 2015 and 2016) and howling behaviour (584 observation hours across May-August 2017 and 2018) throughout the forest fragment and fit our data to several candidate models to estimate DEI. We found howler monkeys farther from anthropogenic edge and closer to river edge than expected by a random distribution of locations. We estimated DEIs for howl bout rate (bouts/hour), howl bout length, and howling rate per bout (howls/minute per bout) that were different to those used in previous work at this site, suggesting that a priori definitions of edge used in prior work were likely too small to detect edge effects and the methods used provide a more accurate picture of how edges impact these monkeys. Our results provide an example of how using different edge definitions can alter research findings, showing that mathematical methods of determining DEI are warranted. This mathematical approach to examining behavioural edge effects can be applied more broadly across primates to help us better understand how primates respond to edges.