Open AccessSaving time and money by using diurnal vehicle counts to monitor roe deer abundance
Author(s) -
Pellerin Maryline,
Bessière Aurélie,
Maillard Daniel,
Capron Gilles,
Gaillard JeanMichel,
Michallet Jacques,
Bonenfant Christophe
Publication year - 2017
Publication title -
wildlife biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.566
H-Index - 52
eISSN - 1903-220X
pISSN - 0909-6396
DOI - 10.2981/wlb.00274
Subject(s) - capreolus , abundance (ecology) , distance sampling , transect , roe deer , population , population density , index (typography) , statistics , sampling (signal processing) , ecology , physical geography , geography , biology , demography , mathematics , computer science , sociology , world wide web , filter (signal processing) , computer vision
Despite being a widespread and important game species in Europe, scientifically reliable, easy applicable and cost effective methods for monitoring abundance of roe deer Capreolus capreolus populations do not yet exist. The currently recommended kilometric index (AI‐p) captures temporal variation in the relative abundance of populations; however, because this index is carried out on foot, it is demanding in terms of sampling effort and difficult to apply at spatial scales of several hundred km 2 typical of deer management units. Here, we propose and test a modified version of the kilometric index by using a vehicle to carry out transects over large areas (AI‐v). To validate this abundance index, we compared variation in population abundance estimated with AI‐p and AI‐v with capture—mark—recapture (CMR) estimates of population density in a roe deer population, Chizé (France), monitored for 24 years (including eight years when both indices were collected). We found no detectable effect of conditions of observation (temperature and precipitation) on either AI‐p or AI‐v. AI‐p and AI‐v were both positively and linearly related (on a log scale) to CMR estimates of population density, after accounting for uncertainty of CMR estimates by using a bootstrap procedure. AI‐p was slightly better correlated to population density (r = 0.76) than AI‐v (r = 0.58). The positive correlation of AI‐p and AI‐v with CMR density estimates as well as the reduced costs of conducting surveys by car instead on foot (‐47%) suggest that diurnal vehicle counts of roe deer can provide a suitable abundance index to monitor temporal trends in roe deer populations at operational management scales. For reliable management of wildlife populations, diurnal vehicle counts of roe deer could be used in association with measures of animal performance and herbivore impacts on the habitat, within the framework of the indicators of ecological change.