Effects of habitat suitability and minimum patch size thresholds on the assessment of landscape connectivity for jaguars in the Sierra Gorda, Mexico

被引:18
作者
Ramirez-Reyes, Carlos [1 ]
Bateman, Brooke L. [1 ]
Radeloff, Volker C. [1 ]
机构
[1] Univ Wisconsin Madison, Dept Forest & Wildlife Ecol, SILVIS Lab, 1630 Linden Dr, Madison, WI 53706 USA
关键词
Maxent; Conefor Sensinode; Species distribution modeling; Panthera onca; SPECIES DISTRIBUTION MODELS; PANTHERA-ONCA; SELECTING THRESHOLDS; POTENTIAL HABITAT; CAMERA-TRAPS; SAMPLE-SIZE; LAND-USE; CONSERVATION; PERFORMANCE; IDENTIFICATION;
D O I
10.1016/j.biocon.2016.10.020
中图分类号
X176 [生物多样性保护];
学科分类号
090705 ;
摘要
Maintaining habitat and its connectivity is a major conservation goal, especially for large carnivores. Assessments of habitat connectivity are typically based on the output of habitat suitability models to first map potential habitat, and then identify where corridors exist. This requires separating habitat from non-habitat, thus one must choose specific thresholds for both habitat suitability and the minimum patch size that can be occupied. The selection of these thresholds is often arbitrary, and the effects of threshold choice on assessments of connectivity are largely unknown. We sought to quantify how habitat-suitability and patch-size thresholds influence connectivity assessments for jaguars (Panthera onca) in the Sierra Gorda Biosphere Reserve in central Mexico. We modeled potential habitat for jaguars using the species distribution modeling algorithm Maxent, and assessed potential habitat connectivity with the landscape connectivity software Conefor Sensinode. We repeated these analyses for 45 combinations of habitat suitability based thresholds and minimum patch sizes. Our results indicated that the thresholds influenced connectivity assessments greatly, and different combinations of the two thresholds yielded vastly different map configurations of suitable habitat for jaguars. We developed an approach to identify the pair of thresholds that best matched the jaguar occurrence points based on the connectivity scores. Among the combinations that we tested, a threshold of 03 for habitat suitability and 2 km(2) for minimum patch size produced the best fit (area under the curve = 0.9). Surprisingly, we found low suitable habitat for jaguars in most of the core areas of the reserve according to our best potential habitat model, but highly suitable areas in the buffer zones and just outside of the reserve. We conclude that the best and most connected potential areas for jaguar habitat are in the central eastern part of the Sierra Gorda. More broadly, landscape connectivity analyses appears to be highly sensitive to the thresholds used to identify suitable habitat, and we recommend conducting sensitivity analyses as introduced here to identify the optimal combination of thresholds. Published by Elsevier Ltd.
引用
收藏
页码:296 / 305
页数:10
相关论文
共 79 条
  • [1] [Anonymous], 2009, SPATIAL CONSERVATION
  • [2] [Anonymous], 2015, LANDSCAPE ECOLOGY TH
  • [3] Nice weather for bettongs: using weather events, not climate means, in species distribution models
    Bateman, Brooke L.
    VanDerWal, Jeremy
    Johnson, Christopher N.
    [J]. ECOGRAPHY, 2012, 35 (04) : 306 - 314
  • [4] The effects of small sample size and sample bias on threshold selection and accuracy assessment of species distribution models
    Bean, William T.
    Stafford, Robert
    Brashares, Justin S.
    [J]. ECOGRAPHY, 2012, 35 (03) : 250 - 258
  • [5] Spatial filtering to reduce sampling bias can improve the performance of ecological niche models
    Boria, Robert A.
    Olson, Link E.
    Goodman, Steven M.
    Anderson, Robert P.
    [J]. ECOLOGICAL MODELLING, 2014, 275 : 73 - 77
  • [6] Species detection vs. habitat suitability: Are we biasing habitat suitability models with remotely sensed data?
    Bradley, Bethany A.
    Olsson, Aaryn D.
    Wang, Ophelia
    Dickson, Brett G.
    Pelech, Lori
    Sesnie, Steven E.
    Zachmann, Luke J.
    [J]. ECOLOGICAL MODELLING, 2012, 244 : 57 - 64
  • [7] SDMtoolbox: a python']python-based GIS toolkit for landscape genetic, biogeographic and species distribution model analyses
    Brown, Jason L.
    [J]. METHODS IN ECOLOGY AND EVOLUTION, 2014, 5 (07): : 694 - 700
  • [8] Landscape connectivity: A conservation application of graph theory
    Bunn, AG
    Urban, DL
    Keitt, TH
    [J]. JOURNAL OF ENVIRONMENTAL MANAGEMENT, 2000, 59 (04) : 265 - 278
  • [9] Calabrese JM, 2004, FRONT ECOL ENVIRON, V2, P529, DOI 10.1890/1540-9295(2004)002[0529:ACGTCM]2.0.CO
  • [10] 2