Discrete choice modelling of natal dispersal: ‘Choosing’ where to breed from a finite set of available areas

Abstract

Summary1. Classic natal dispersal studies focus mainly on distance travelled. Although distances capture some of the main selective pressures related to dispersal, this approach cannot easily incorporate the properties of the actual destination vs. the available alternatives. Recently, movement ecology studies have addressed questions on movement decisions in relation to availability of resources and/or availability of suitable habitats through the use of discrete choice models (DCMs), a widely used type of models within econometrics, which explains individual choices as a function of the properties of a finite number of alternatives. 2. In this contribution, we show how the dispersal discrete choice model (DDCM) can be used for analysing natal dispersal data in patchy environments given that the natal and the breeding area of the disperser are observed. We test this method using a case study on Great Tits (Parus major) in an archipelago of small woodlots. 3. Our results show that DDCMs are able to capture the results of classic distance-based approaches and simultaneously allow testing hypotheses on how departure and settlement are affected by variables that characterize the disperser, the natal patch and the breeding area, as well as their interactions. 4. DDCMs can be applied to any other species and system that uses some form of discrete breeding location or a certain degree of discretization can be applied.

Ecological traps for large‐scale invasive species control: Predicting settling rules by recolonising American mink post‐culling

“…We used discrete choice models similar to Vardakis et al. () for considering dispersal as a choice process whereby individuals’ decisions are ruled by the characteristics of the chosen patch and of alternative choices. Using this relatively novel methodological approach, and utilising dispersal data from mink recolonising a large area repeatedly depleted of conspecifics by culling implemented by volunteers, we provided evidence of both density dependence and habitat quality affecting mink settlement decisions.…”

“…Dispersal movement data have recently been analysed using multinomial discrete choice models (multinomial regression models), where the probability of settling in a particular patch is a function of patch‐ and individual‐level covariates (Vardakis et al., ). We used multinomial probit models to describe individuals’ choice from a finite set of mutually exclusive alternatives.…”

“…Finally, following Vardakis, Goos, Adriaensen, and Matthysen (2015), we included a "home advantage" variable in the models. This allows the probability of staying in the natal patch to be independent of the distance-to-travel effect.…”

“…where the probability of settling in a particular patch is a function of patch-and individual-level covariates (Vardakis et al, 2015). We used multinomial probit models to describe individuals' choice from a finite set of mutually exclusive alternatives.…”

Choice models with mixtures: An application to a cocktail experiment

“…The literature about discrete choice models is very extensive (Hensher and Johnson, 2018;Louviere and Woodworth, 1983;McFadden, 1986;Ben-Akiva and Boccara, 1995;McFadden and Train, 2000;Louviere et al, 2000;Hensher et al, 2005) and applications can be found in marketing (Kamakura and Russell, 1989), environmental economics (Bennett and Blamey, 2001), transport (Kupfer et al, 2016), health care (Lancsar and Louviere, 2008;Luyten et al, 2015), ecology (Vardakis et al, 2015) and many other disciplines. However, in the analysis of choice data, products that are mixtures of ingredients have been largely overlooked.…”