Nathan Lopez-Brody was awarded the 2016 New MexicoView Geospatial Scholarship. Nathan is working on a project entitled “Spatio-temporal dynamics of woody plants and bighorn sheep in the San Andres Mountains, New Mexico, U.S.A.” Nathan is a graduate student in the Department of Geography at NMSU. (See more below)
Reza Goljani was awarded the 2016 New MexicoView Geospatial Scholarship. Reza is working on a project entitled “Predicting spatial factors associated with cattle depredations by the Mexican wolf (Canis lupus baileyi) in Arizona and New Mexico.” Reza is a graduate student in the Department of Fish, Wildlife, and Conservation Ecology at NMSU. (See more below)
2016 New MexicoView Geospatial Scholarship Recipients
Spatio-temporal dynamics of woody plants and bighorn sheep in the San Andres Mountains, New Mexico, U.S.A.
Located in south-central New Mexico, the San Andres National Wildlife Refuge provides habitat for desert bighorn sheep (Ovis canadensis nelsoni). Refuge managers believe that the habitat of these translocated sheep has been negatively impacted by local expansions of piñon (Pinus edulis) and juniper (Juniperus monosperma) trees throughout the 20th century. Understanding the relationships between piñon-juniper population dynamics and bighorn sheep habitat is important for making management decisions. In my thesis, I aim to improve this understanding by a) mapping woody plant cover dynamics in the Refuge; and, b) modeling bighorn sheep habitat using woody plant cover and other explanatory variables. To map changes in woody plant cover, I apply multiple endmember spectral mixture analysis to Landsat TM and OLI imagery from 1985 and 2016, respectively. To model bighorn sheep habitat in 1985 and 2016, I use logistic regression with bighorn sheep radio collar locations, the woody plant cover data, and various other explanatory data layers (e.g., slope and ruggedness). I anticipate that piñon-juniper cover will have increased in the Refuge between 1985 and 2016. I also anticipate that this expansion in woody plant cover will reduce the quantity of bighorn sheep habitat in the San Andres Mountains. Remotely sensed data compose the backbone of this study. The Landsat record makes this type of retrospective longitudinal study possible.
Predicting spatial factors associated with cattle depredations by the Mexican wolf (Canis lupus baileyi) in Arizona and New Mexico
My thesis is titled “Predicting spatial factors associated with cattle depredations by the Mexican wolf (Canis lupus baileyi) in Arizona and New Mexico”. The overarching goal of this study is to develop models that explain spatial factors associated with Mexican wolf depredations on livestock. We compare various presence-only (e.g. Maximum Entropy [Maxent], Artificial Neural Network [ANN], Support Vector Machines [SVM]), and presence-absence (Generalized Linear Models [GLM], Generalized Additive Models [GAM], and Zero Inflated Models [ZIM]) spatial modeling methods to find the best approach to identify landscape features that increase risk of wolf depredation, and produce a risk map that depicts the spatial distribution of predicted risk of cattle depredation by Mexican wolves in the study area (i.e., Mexican wolf recovery area which is all of Arizona and New Mexico south of Interstate-40). Our spatial predictors include biotic communities, vegetation structure, water resources, developed areas, and topographic variables. Moreover, we are using Maxent and GLM to develop spatial models that will represent the abundance of natural prey (elk, mule deer, white-tailed deer) and livestock to use as predictors in the risk models. To prepare predictor variables, build models and evaluate models, we employ a variety of geospatial software and methods including Arc GIS, ERDAS, R, python, open modeler, Maxent, ENFA, and ENMTools. The results of this study will provide guidance for future spatial depredation risk modeling and help Mexican wolf managers to detect and modify Mexican wolf-livestock conflict hotspots before the conflicts occur.