One of the fundamental concepts in conservation biology involves understanding where a species occurs, either historically or in present times. While this is an intuitively simple concept, species occurrence combines of variety of ecological processes, such as the environmental characteristics that create suitable habitat, the behavior of the animal, the movement ability of the animal, the geologic and evolutionary history that led to modern distributions, and, recently, how these factors interact with an increasingly human-dominated landscape. An important realization is that it is exceedingly difficult if not impossible to precisely determine where most species occur through typical field surveys. Indeed, field surveys almost always have large gaps in coverage because of some combination of limited resources, the cryptic or secretive nature of the species in question, or an inability to access certain areas for surveys. This is especially true for species whose range encompasses large geographic areas, including a mix of public and private properties.
Because of these challenges, creating models that predict and map suitable habitat within a species’ range or a specific study area has become a common conservation technique. Known by many terms (most commonly Species Distribution Models or Habitat Suitability Models), these models provide a useful tool in the conservation biologist’s toolbox. The basic framework of this approach relies on relating species observations to a variety of environmental characteristics (e.g., landcover type, soil type, fire frequency, etc.) and then using these relationships to make predictions across a broader area. The end result is a broad, landscape view of where a species may occur (note the distinction between occurrence and habitat suitability) based on the available habitat.
We recently published an article in the journal Ecological Applications where we created the first range-wide habitat suitability model for Eastern Indigo Snakes (Chandler et al. 2022). We compiled indigo snake records from our own datasets, museum collections, partner organizations, and observations made by the public. The final dataset included 1,215 contemporary (2000–2020) indigo snake observations that spanned from southern Georgia to the Florida Keys. We then acquired a number of GIS layers that quantified various aspects of the environment known to be important for indigo snake habitat (many of these have been used in similar projects for other imperiled herpetofauna in the southeast; Crawford et al. 2020). Think of these layers as square grids where each grid cell has a number that describes some part of the environment. For example, one environmental layer described whether each 30 by 30 m cell across the study area was composed of wetland habitat or not. With these data acquired, we set out to actually create our habitat model.
One of the reasons that this type of work had not been previously conducted for indigo snakes is because their ecology and habitat associations change across their range. In the northern portion of their range, indigo snakes depend on Gopher Tortoise burrows to provide refugia from potentially lethal winter temperatures. However, indigo snakes in southern Florida do not depend on tortoise burrows and use a wider variety of habitat types throughout the year. To account for this variation in habitat associations, we included an effect of winter temperature in our model. Without going into too much detail, this allowed the relationship between habitat suitability and the environmental layers to change depending on where the model was predicting (i.e., different habitat relationships in southern Georgia and northern Florida when compared to southern Florida). A second key aspect of our approach was that we assessed the environmental layers at multiple spatial scales. This is important, especially for a species like indigo snakes that have large home ranges, because habitat suitability may have more to do with the characteristics of the broader landscape than at any one point on the landscape (e.g., how often is the entire sandhill burned vs. how often is the spot where an indigo snake was observed burned).
Our results indicated that the approach of including winter temperature and accounting for multiple spatial scales produced a more accurate model. Unsurprisingly, we found positive associations between indigo snake habitat suitability and the amount of upland landcover, while there was a negative relationship between habitat suitability and urban landcover. In contrast to these relationships that were consistent across the species’ range, agriculture had a negative impact on habitat suitability in the northern portion of the range but had a positive effect on suitability in southern Peninsular Florida. Similarly, fire frequency had a positive effect at cooler winter temperature, but that effect was mostly absent in southern Florida. Finally, we used our top model to assess range-wide trends in indigo snake habitat suitability, using three suitability categories that ranged from low to high.
Overall, our results highlight that there are still large areas of suitable indigo snake habitat throughout the species distribution but that land protections are often lacking. The amount of habitat available in the Florida Panhandle, where the species is presumed to be extirpated, is striking but highlights the limitations in our ability to model the current distribution of tortoise populations, which are a limiting factor in this region. However, it suggests that reintroductions of both tortoises and indigo snakes may ultimately be successful in this region. Additionally, our results demonstrate how challenging it can be to identify suitable indigo snake habitat, especially in southern Florida. While including winter temperature in our model improved the results, model performance was still best in southern Georgia and northern Florida where indigo snakes have more stringent habitat requirements. Similarly, it is challenging to effectively identify summer habitat use in indigo snakes because they are rarely encountered during this time of year. Thus, our model is likely a better representation of winter habitat, particularly in the northern half of the range. Finally, the results from this work are now available to use in ongoing indigo snake conservation and management projects, and there are many potential applications. For example, results could be used to identify areas of potentially suitable habitat where indigo snake records are lacking, informing future survey efforts and filling in existing data gaps.
The full publication is available through the link below and on the Publications Page of our website.
Chandler, H.C., C.L. Jenkins, and J.M. Bauder. 2022. Accounting for geographic variation in species-habitat associations during habitat suitability modeling. Ecological Applications:e2504.
Crawford, B.A., J.C. Maerz, and C.T. Moore. 2020. Expert-informed habitat suitability analysis for at-risk species assessment and conservation planning. Journal of Fish and Wildlife Management 11:130–50.
U.S. Fish and Wildlife Service. 2019. Species Status Assessment report for the Eastern Indigo Snake (Drymarchon couperi). Version 1.1, Atlanta, GA.