New Published Research: Informing Flatwoods Salamander Conservation with High-Resolution Elevation Data

Views:

123 views
Example of LiDAR data used in this research. Warm colors indicate higher elevations (trees and shrubs) while cool colors indicate lower elevations (bare ground or grasses). The entire image is made up of over 100 million individual points, each with an associated elevation value. The data can be processed to estimate both bare ground elevation and vegetation height across a continuous grid. – Houston Chandler
Example of LiDAR data used in this research. Warm colors indicate higher elevations (trees and shrubs) while cool colors indicate lower elevations (bare ground or grasses). The entire image is made up of over 100 million individual points, each with an associated elevation value. The data can be processed to estimate both bare ground elevation and vegetation height across a continuous grid. – Houston Chandler
The Importance of Ephemeral Wetlands for Amphibian Conservation

Ephemeral wetlands are characterized by regular cycles of flooding and drying, typically in response to short-term weather events. These wetlands have long been recognized as critical habitats for amphibians because regular drying limits the presence of fish and other top predators. In the southeastern United States, ephemeral wetlands often support high amphibian diversity, provide important habitat for imperiled species, and promote landscape connectivity and ecosystem health through seasonal movement of animals, water, and nutrients into and out of wetland basins. Thus, understanding the ecology of ephemeral wetlands can provide important insights into amphibian conservation and help inform management decisions for imperiled species.

 
Wetland Hydrology and Vegetation: Key to Amphibian Reproduction

For many amphibian species using ephemeral wetlands as breeding habitats, the most important characteristics influencing reproductive success are wetland hydrology and vegetation structure. In general, wetlands must hold water for long enough and that water must submerge preferred egg-laying and larval habitats if reproduction is going to be successful. Hydroperiod (the length of time a wetland has standing water) has long been used as the primary metric describing hydrology in amphibian studies. However, hydroperiod presents a relatively vague picture of the day-to-day changes in flooded habitats, especially when considered as a binary indicator of wet versus dry. 

For large wetlands or wetlands with variable vegetation characteristics, changes in the amount of standing water could have large impacts on the amount of habitat available to larval amphibians or to the quality of that habitat. Despite the potential importance to amphibians, these within wetland fluctuations in flooded area have rarely been examined, likely due to the difficulty in measuring such metrics.

Flatwoods salamander conservation - salamander captured entering a breeding wetland in Florida.
A flatwoods salamander captured entering a breeding wetland in Florida. – Houston Chandler
Using Remote Sensing to Monitor Flatwoods Salamander Habitats

As part of my dissertation work, and in collaboration with researchers at Virginia Tech, I set out to examine whether advances in remote sensing data could increase our ability to monitor wetlands used as breeding habitats by Reticulated Flatwoods Salamanders (Ambystoma bishopi). The goal of this work was to expand on some of my previous research (Chandler et al. 2015, 2016, 2017) to create a more holistic view of wetland hydrology and vegetation structure.

A characteristic pine flatwoods landscape with an overstory of pines and an understory of thick herbaceous vegetation. Areas like these were studied for flatwoods salamander conservation. Image by Houston Chandler
A characteristic pine flatwoods landscape with an overstory of pines and an understory of thick herbaceous vegetation. The line between wetland and upland can sometimes be difficult to define in such landscapes. – Houston Chandler

One of the challenging aspects of working in wetlands in the Florida Panhandle is that the entire landscape is incredibly flat. The line between wetland and upland is often blurred, challenging assessments of wetland habitats. The critical first step for this project was finding an efficient method to measure wetland bathymetry (the shape of the basin). While it is possible to measure bathymetry by hand, field-derived estimates are time consuming and challenging in landscapes with such little overall elevation change. Bathymetry is a fascinating metric because it allows you to spatially map standing water within a wetland. Combined with water level monitoring data (which we had for many wetlands in this study area), the metric for understanding wetland hydrology shifts from hydroperiod to flooded area.

Leveraging LiDAR Data to Map Wetlands

For this project, we took advantage of increases in available remote sensing data describing the Earth’s surface. Specifically, Light Detection and Ranging (or LiDAR as it’s commonly referred to) uses a laser, typically mounted on a small aircraft or drone, to map the topography of a landscape. LiDAR data returns a point cloud of various elevations, often at high spatial resolution (multiple elevation readings within a single square meter). This data can then be used to visualize not only the ground surface but also vegetation or structures above the ground (see example in the top image).

Conceptual diagram linking wetland bathymetry (A) to the relationship between water depth (stage) and flooded area (B). Once this relationship has been determined, historical water level data can be converted to estimates of flooded area over time (C). At any point in time, flooded area metrics can also be displayed spatially across the wetland to visualize specific areas that were flooded. Gray bars in C represent flatwoods salamander breeding seasons. – Houston Chandler
Conceptual diagram linking wetland bathymetry (A) to the relationship between water depth (stage) and flooded area (B). Once this relationship has been determined, historical water level data can be converted to estimates of flooded area over time (C). At any point in time, flooded area metrics can also be displayed spatially across the wetland to visualize specific areas that were flooded. Gray bars in C represent flatwoods salamander breeding seasons. – Houston Chandler

We used high-resolution LiDAR data for one of the remaining flatwoods salamander strongholds to map wetland bathymetry and vegetation structure across 30 ephemeral wetlands. In 17 of these wetlands, we had multiple years of water level monitoring data to examine trends in flooded area over time. We ultimately used these data to 1) quantify depth-to-flooded-area relationships across the landscape, 2) examine trends in flooded area (within wetlands and through time), and 3) map potential flatwoods salamander habitat.

 

Wetland Bathymetry and Flooded Area Dynamics

Overall, we found that the relationship between wetland depth and flooded area varied across wetlands as a function of wetland size and bathymetry. A one-centimeter increase in depth could generate increases in flooded area raging from 100s to 1000s of square meters, depending on the wetland. Because wetland basins were shallow overall (just 1.1 m elevation gain from the lowest to highest point, on average) and not typically bowl-shaped, wetlands tended to have many disjointed flooded patches at low water levels. As water levels rose, these patches eventually coalesced into a single large flooded area. 

Most wetlands were dominated by a single large patch of water at approximately 50% of their maximum depth. Importantly, the fragmentation of a single flooded area into multiple patches could elevate larval flatwoods salamander densities, increasing competition, shifting habitat suitability for larvae in certain patches, and impacting growth or survival. Different patches may, in fact, experience different hydroperiods all within the same breeding season! In years where wetland water levels fluctuate significantly, the effects of variable patch sizes could occur repeatedly over a single breeding season.

Relationships between flooded area (stage) and the number of discrete flooded patches in pine flatwoods wetlands. All patches considered were at least 1 square meter. The colors in the bottom plot represent total wetland size (< 2 hectares = blue, > 2 hectares = orange). – Houston Chandler
Relationships between flooded area (stage) and the number of discrete flooded patches in pine flatwoods wetlands. All patches considered were at least 1 square meter. The colors in the bottom plot represent total wetland size ( 2 hectares = orange). – Houston Chandler
Mapping Suitable Habitat for Flatwoods Salamanders

We used the resulting hydrologic characterizations along with vegetation metrics (also derived from the LiDAR data) to map potentially suitable flatwoods salamander habitat. We based habitat classifications on whether areas within wetlands were predicted to be flooded at least 50% of the time (hydrologically suitable) and whether vegetation heights were less than 1 meter (indicating herbaceous vegetation used as egg laying and larval habitat). 

We found that these maps generally aligned well with previous on-the-ground assessments of habitat, while allowing us to cover a much larger area and consider hydrology in addition to vegetation. Such spatial delineations of habitat can be used to guide future management actions by identifying specific areas of wetlands where vegetation restoration would have the largest benefit. It can often be challenging to make such decisions, especially in large wetlands.

Maps demonstrating how LiDAR data can be used to map potential flatwoods salamander habitat, both from a hydrologic and vegetation perspective. Black polygons indicate areas that were delineated in the field as having high-quality vegetation for flatwoods salamanders. However, such areas do not necessarily overlap with parts of a wetland that are most likely to hold water. – Houston Chandler
Maps demonstrating how LiDAR data can be used to map potential flatwoods salamander habitat, both from a hydrologic and vegetation perspective. Black polygons indicate areas that were delineated in the field as having high-quality vegetation for flatwoods salamanders. However, such areas do not necessarily overlap with parts of a wetland that are most likely to hold water. – Houston Chandler
Conclusion and Future Research Opportunities

This project was primarily a proof of concept highlighting how relatively new data sources could be applied to better understand ephemeral wetland ecology and amphibian conservation. There are ample opportunities for additional research following these general themes. For example, the effects of within-wetland variation in flooding and habitat quality should be linked to the population biology of the species breeding in these wetlands. Importantly, we found only a weak correlation between hydroperiod and flooded area, suggesting that these metrics describe different parts of the hydrologic cycle. Considering flooded area as a metric will likely increase our understanding of amphibian ecology. Ultimately, integrating species ecology with factors controlling hydrologic regimes within wetlands will be fundamental to making informed conservation and management decisions in these systems.

The research discussed here was recently published in the journal Wetlands. The full publication can be viewed at the following link.

Literature Cited

Chandler, H. C., C. A. Haas, and T. A. Gorman. 2015. The effects of habitat structure on winter aquatic invertebrate and amphibian communities in pine flatwoods wetlands. Wetlands 35:1201–1211.

Chandler, H. C., A. L. Rypel, Y. Jiao, C. A. Haas, and T. A. Gorman. 2016. Hindcasting historical breeding conditions for an endangered salamander in ephemeral wetlands of the southeastern USA: Implications of climate change. PLoS ONE 11:e0150169.

Chandler, H. C., D. L. McLaughlin, T. A. Gorman, K. J. McGuire, J. B. Feaga, and C. A. Haas. 2017. Drying rates of ephemeral wetlands: Implications for breeding amphibians. Wetlands 37:545–557.