Print PageEcological niche models (ENM) use environmental and species presence data to model the potential distribution of that species. We are using the cactus Echinocereus reichenbachii as a model system to compare three ENM to determine which model best predicts its distribution. E. reichenbachii is a small, columnar cactus listed as an S2 plant in Oklahoma, meaning that it is imperiled because of extreme rarity or other factors that makes it vulnerable to extinction throughout its native range. Knowledge of its potential range and what factors determine suitable habitat are important to conservation efforts for this plant. The objectives of this research are to verify locations given in extant herbarium records for the species, obtain exact coordinates for sites where the plant is present, and use environmental layers and GIS technology to model habitat suitability. We are comparing Genetic Algorithm for Rule Set Production (GARP), Ecological Niche Factor Analysis (ENFA), and Maximum Entropy Modeling (MAXENT). The efficacy of these models will be compared via field validation. In addition, we will use the models’ output to determine which environmental factors are most important in determining habitat suitability.
Results/Conclusions
To date, 22 new sites of occurrence in Oklahoma, New Mexico, and Texas have been georeferenced and environmental data including minimum and maximum temperatures, precipitation, elevation, and soil characteristics have been obtained. In addition, 94 additional observation records have been obtained from herbarium records in the Global Biodiversity Information Facility (GBIF) and the Oklahoma Vascular Plants Database (OVPD) and georeferenced. Using aerial photography, we have determined that 28% of the GBIF points could be classified as unsuitable habitat because of conversion to agricultural or urban/suburban uses. An additional 36% of the GBIF points were classified as suitable but threatened because of encroaching agriculture, or other human uses and only 34% were classified as potentially suitable habitat. Of the 71 OVPD points obtained, only 39 could be assigned coordinates with some accuracy. Of those, 13% were unsuitable agricultural land. One fourth (10 of 39) were classified as suitable but threatened, primarily due to encroaching agricultural lands but also due to urban expansion. Over half of the OVPD data points (62%) were classified as potentially suitable habitat. Of those, many occurred within protected areas. Species presence and environmental data are currently being processed for use in the ENM models. Data collected to date suggest that land use change poses a significant threat to this imperiled species.
