Landscape variables that influence the presence of brocket deer (Mazama sp.) in the Campeche State, México

Abstract

The brocket deer (Mazama temama and M. pandora) are two of the Mexican tropical deer with great importance for subsistence and sport hunters. Brocket deer are considered species with preference for pristine habitats conserved, however at the landscape level, factors such as hunting pressure could influence the presence and abundance of these species. Campeche, is the only Mexican state where it has been authorized the exploitation of the brocket deer for sport hunting. This hunting pressure along with the intense subsistence hunting that occurs in Campeche, could have important effects on the conservation of the species. Establishing what factors determine the presence of the species at the landscape level is essential to elaborate comprehensive management plans of these species to allow their long-term conservation. The aim of this study was to determine the habitat variables at the landscape level that influence the brocket deer presence in the central-western area of Campeche State in Mexico. The study area in the state of Campeche (19° 15’ N, -91º 20´ W) was divided into six study regions (Chekubul-Cristalina; Chenkan; Chicbul-Km 74; Pixtún-Felipe Carrillo Puerto, Miguel Colorado; Centenario-La Rigueña) following the main paved road of the area and according to the conservation level of the natural ecosystems, presence of flooded areas, and intensity of the human activities where we determine the presence of brocket deer. We searched for brocket deer tracks along 263 transects of 500 m each. The tracks were identified to genus only. The presence/absence of brocket deer in the transects was combined with variables of human and natural origin of the study area in a geographical information system. We use principal component analysis and a logistic regression to determine the relationship between variables and the presence/absence of the brocket deer. We were not able to distinguish by their tracks the two potentially distributed species of brocket deer in the area. Due to this constraint, we did not perform a distinction between both species and all the observed tracks were classified as brocket deer. We found that brocket deer were present only in three of the six studied regions: Miguel Colorado (21.3 % transect had brocket deer presence), Centenario-La Rigueña (20.4 %) and Chenkan (1.7 %). The principal component analysis indicated that the variables determining the presence of brocket deer were, the percentage of preserved forest and distance to pastures, positively associated, while livestock density was negative. The best model obtained by the logistic regression showed that the presence of brocket deer increased when the percentage of preserved forests around transect is increased, while decreasing with increasing livestock density. The presence of conserved forests favors the presence of brocket deer in the study area, while a higher density of cattle decreases the probability of detection. The few records of brocket deer in flooded zones in this study and the lack of detection of the species in previous studies, suggests that these flooded ecosystems may not be suitable for brocket deer, which would have a strong impact on their known distributional range. Finding that brocket deer are present mainly in preserved sites indicates the species could be subject to an intense population reduction in the study area due to the continued habitat loss. The results suggest that before granting authorizations for sport hunting in this region, it is necessary to make accurate estimates of the population distribution and abundance of the species in the region.