Strategic Habitat Conservation for Beach Mice: Estimating Management Scenario Efficiencies

Abstract

The Perdido Key beach mouse (Peromyscus polionotus trissyllepsis), Choctawhatchee beach mouse (P. p. allophrys), and St. Andrew beach mouse (P. p. peninsularis) are 3 federally endangered subspecies that inhabit coastal dunes of Alabama and Florida, USA. Conservation opportunities for these subspecies are limited and costly. Consequently, well‐targeted efforts are required to achieve their downlisting criteria. To aid the development of targeted management scenarios that are designed to achieve downlisting criteria, we developed a Bayesian network model that uses habitat characteristics to predict the probability of beach mouse presence at a 30‐m resolution across a portion of the Florida Panhandle. We then designed alternative management scenarios for a variety of habitat conditions for coastal dunes. Finally, we estimated how much area is needed to achieve the established downlisting criterion (i.e., habitat objective) and the amount of effort needed to achieve the habitat objective (i.e., management efficiency). The results suggest that after 7 years of post‐storm recolonization, habitat objectives were met for Perdido Key (within its Florida critical habitat) and Choctawhatchee beach mice. The St. Andrew beach mouse required 5.14 km2 of additional critical habitat to be protected and occupied. The St. Andrew beach mouse habitat objective might be achieved by first restoring protected critical habitat to good dune conditions and then protecting or restoring the unprotected critical habitat with the highest predicted probability of beach mouse presence. This scenario provided a 28% increase in management efficiency compared to a scenario that randomly protected or restored undeveloped unprotected critical habitat. In total, when coupled with established downlisting criteria, these quantitative and spatial decision support tools could provide insight into how much habitat is available, how much more is needed, and targeted conservation or restoration efforts that might efficiently achieve habitat objectives. © 2020 The Wildlife Society.

Decline in annual survival of American oystercatchers wintering in Florida linked to extreme high tides

“…Loss of late‐succession sand dune plant habitats and fragmentation of coastal dune systems due to frequent storm events coupled with sea‐level rise is predicted to cause beach mouse ( Peromyscus spp.) population declines (Feagin et al 2005, Yuro 2011, Chen et al 2014, Cronin et al 2021). Climate change and sea‐level rise will have effects on marsh bird populations; however, the timing and scale of effects may be highly variable regionally and among species (Woodrey et al 2012, Nuse et al 2015, Hunter et al 2017).…”

Integrating Bayesian Networks to Forecast Sea‐Level Rise Impacts on Barrier Island Characteristics and Habitat Availability

“…At moderate spatiotemporal scales (i.e., decadal and 30‐m landscape scales), BNs have been developed to explore the potential for coastal landscape dynamic response to SLR (Lentz et al., 2016); SLR‐driven evolution in barrier island characteristics (e.g., elevation, beach width along shore‐normal transects spaced in 50‐m intervals; Gutierrez et al., 2015); and storm‐driven changes to barrier island characteristics (50‐m transects; Plant & Stockdon, 2012; Plant et al., 2016; Wilson et al., 2019). Finally, BNs of seasonal habitat availability given coastal landform conditions (e.g., vegetation density, substrate type) have been developed for species like piping plovers (5‐m cells; Zeigler et al., 2021) and beach mice ( Peromyscus poliontus subspecies; 30‐m cells; Cronin et al., 2021). In addition to spanning spatial and temporal scales, these BNs are capable of simultaneously incorporating data from a variety of sources (e.g., empirical data and expert opinion) and formats (e.g., continuous and categorical variables), capturing hierarchical and complex correlations among variables, and accommodating incomplete data.…”