Estimating salamander dispersal distances to inform management decisions


Reticulated Flatwoods Salamander – Houston Chandler

I have written previously about the importance of maintaining and restoring connectivity at a landscape scale to reduce the negative effects of fragmentation on natural communities.  Even though the importance of maintaining connectivity is well-understood in the conservation community, direct measures of connectivity between animal populations that can inform specific management objectives are often lacking.  Measuring dispersal between populations in the field can be logistically difficult, expensive, and/or time consuming.  For example, radio telemetry may be unsuitable for small or rare species and analyzing population genetics at a landscape scale can be expensive and time consuming.  Think about how challenging these two questions would be to answer.  How do you follow a small frog or salamander hundreds of meters across a complex landscape, especially when they spend time underground?  Out of hundreds or thousands of individuals, how do you know which individuals will be the dispersers?

Pond-breeding amphibians are often described as one of the best examples of a metapopulation (i.e., a group of populations that are close enough for individuals to occasionally interact).  Multiple amphibian breeding ponds often occur in somewhat close proximity to one another, allowing individuals to occasionally move between wetlands.  Adults generally have high site fidelity once they are established at a wetland (forming the population at that site) but sub-adult individuals occasionally disperse, either intentionally or unintentionally, to adjacent wetlands.  These dispersal processes are important for species breeding in ephemeral wetlands because wetlands that are currently not occupied can be colonized by individuals moving across the landscape.  However, as described above these processes are difficult to measure and quantify in a way that can guide management decisions. 

Distance to the nearest occupied flatwoods salamander breeding wetland.

As part of ongoing work with collaborators in the Fish and Wildlife Conservation Department at Virginia Tech, we recently published a manuscript examining connectivity between and habitat characteristics within Reticulated Flatwoods Salamander (Ambystoma bishopi) breeding wetlands.  We employed spatially explicit statistical models that quantified the distance at which colonization events were occurring across several flatwoods salamander breeding wetlands.  Our results indicated that the probability of salamanders colonizing wetlands decreased sharply with increasing distance between wetlands, and the chances of colonization occurring was negligible at distances greater than 1.5 km (see figure below).  This is important because it refines previously published dispersal distances for flatwoods salamanders, which were as high as 1,700 m.  Other research on flatwoods salamanders has indicated that movements over 1 km are extremely rare, and our estimate of an average dispersal distance of approximately 230 m is similar to those observed in other species of ambystomatid salamanders.  These results indicate that wetlands separated by several hundred meters would rarely be colonized by individuals from surrounding wetlands.

The relationship between distance to the nearest wetland and the probability of a colonization event occurring.

Previous work has indicated that flatwoods salamanders depend on fire-maintained wetlands with thick herbaceous vegetation covering the wetland basins.  This vegetation acts as a site for egg deposition and provides cover and foraging opportunities for larval salamanders.  Our analysis indicated that the only significant environmental predictor for site occupancy was the total amount of suitable egg laying habitat.  In fact, sites with less than 0.2 ha of suitable egg laying habitat were never occupied by flatwoods salamanders over the course of the study (2009–2017).  These results provide further evidence that restoring fire-maintained vegetation structure is crucial to successfully managing flatwoods salamander populations.

Overall, the results of this research have strong implications for the ongoing management of flatwoods salamander breeding wetlands at sites across their range.  The chances of a colonization event increased with decreasing distance to other wetlands, and management activities should be focused on occupied wetlands and wetlands directly adjacent to these occupied sites.  These management activities should attempt to increase the amount of herbaceous vegetation in the wetland basin by clearing woody midstory vegetation, removing accumulated duff layers, and reintroducing fire to wetland basins, especially during the growing season when fire effects are more significant.  In addition to identifying specific wetlands that are ideal for expanded management opportunities, these results can also inform the delineation of broader management units for the species’ recovery planning.  Management units should focus on groups of wetlands well-within the 1,500 m threshold where salamanders can move between different wetlands via natural processes.  Clusters of wetlands outside of this range can be considered separate management units or metapopulations that enhance the overall resiliency of flatwoods salamanders.

Larval Reticulated Flatwoods Salamander – Houston Chandler

The research described here was published in the journal Aquatic Conservation: Marine and Freshwater Ecosytems. If you are interested in receiving a copy of the manuscript, please email to request a copy.