Assessing multi-scale habitat relationships and responses to forest management for cryptic and uncommon herpetofauna in the Missouri Ozarks, USA

Highlights

• Modeled capture probabilities using multiple habitat and harvest related covariates.

• Multiple responses to stand-level harvest methods for uncommon herpetofauna.

• Models showed habitat associations for multiple herpetofauna species.

• Herpetofauna responses to management can be used when creating management plans.

• Cave salamanders are associated with streams but can use extensive terrestrial habitat.

Abstract

Behaviorally cryptic or uncommon herpetofauna are often understudied due to the extensive effort it requires to obtain adequate data for statistical analysis. Additionally, potential impacts from forest management on these already small or difficult to study populations may have a dissimilar effect in comparison to more common species. To address this, we examined species-specific responses of uncommon herpetofauna within the Missouri Ozarks to even-aged and uneven-aged silvicultural systems at multiple scales, as well as their habitat relationships. Using capture histories collected over 23 years (1992–2014) on the Missouri Ozark Forest Ecosystem Project (MOFEP) we examined the cumulative effects of two harvest entries (1996 and 2011) at both the local- (stand-level) and landscape-scale (compartment-level) for eight uncommon herpetofauna, which includes one toad, two salamanders, one skink, and four snakes. We modeled capture probabilities with respect to multiple habitat- and harvest-related covariates. Three species showed compartment-level declines in the post-treatment period, however only two of the declines appeared to be related to forest management; the decline for the third species was observed in both treatment and control compartments suggesting that the cause was environmental related. In contrast to compartment-level responses, we observed stand-level responses in five species, mostly positive. In general, our observed declines were minimal and currently we have no concerns that forest management will lead to the loss of any of the uncommon herpetofauna species considered here. Our models showed habitat relationships for multiple species, which aids our understanding of species’ life history strategies and can also guide future management efforts.

Introduction

There is a continued need to understand how disturbances created by forest management impact amphibian and reptile (collectively known as herpetofauna) communities at multiple scales (Block and Morrison, 1998, Liu, 1993, Sing et al., 2018), as well as how species-specific habitat relationships may result in a range of responses to disturbance (Connette and Semlitsch, 2013, Peterman et al., 2013a, Peterman et al., 2013b, deMaynadier and Hunter, 1995, Todd and Andrews, 2008). Forest management is often constrained by rare or endangered species, therefore ecological studies focusing on the effects of management on potentially at-risk herpetofauna species (i.e., less common and/or those with specific habitat relationships) is crucial information for managers and can reduce impacts on already small or isolated populations. Often this information is not available for a species until after declines are realized or a species receives protection (Dobson, 2005, Ingersoll et al., 2013), or the effect is only measured at an ecologically small scale (Corn and Bury, 1989, deMaynadier and Hunter, 1995, Semlitsch et al., 2008, Semlitsch et al., 2009). Species rarity can often be tied to unique life history strategies, habitat requirements, physiological constraints, secrecy, or low detection (Durso et al., 2011, Parker and Plummer, 1987, Stearns, 1992, Suzuki et al., 2007), and often little is known about these herpetofauna species due to extensive surveying that is needed to collect robust datasets (Gibbons et al., 2000). Rare or endangered species often rely on specific fine scale habitat features for persistence (Block and Morrison, 1998, Peterman and Semlitsch, 2013), which can intensify the impacts of forest management, therefore it is important for managers to consider habitat requirements for a variety of species. Additionally, habitat requirements at the landscape scale can also have an important role in population persistence due to metapopulation dynamics (McCullough, 1996, Schmidt and Pellet, 2005, Stoddard and Hayes, 2005, Wiens, 1996). Understanding what habitat and harvest-related factors are associated with species-specific presence, persistence, and/or population decline can help direct land management (e.g., develop priority geographies for species) based on easily measured habitat variables (Hamer and Mahony, 2010, Kroll et al., 2008, Noon et al., 2012, Schmidt and Pellet, 2005, Stoddard and Hayes, 2005).

Important factors to consider when prescribing forest management is the potential for management to disrupt or alter resource availability, habitat and community structure, population sizes, and ecological processes (Ash, 1995, Phillips and Shure, 1990, Sutton et al., 2013). Forest management that resembles natural and historic stand-level disturbance will likely have less of an impact on the overall populations than anthropogenic disturbances that do not resemble historic processes (Drever et al., 2006, Franklin et al., 2007, Turner et al., 1989). Spatial arrangement, temporal application, as well as intensity of forest management are all important factors that could impact species presence and persistence. Although some species may benefit from openings created during harvest entries, such as clearcuts or group openings (Rota et al., 2017), other species require contiguous and mature forest with minimal edge (Greenberg, 2001, Moorman et al., 2011) and can be very sensitive to small changes in habitat structure due to small home ranges (Suzuki et al., 2007), physiology (Homan et al., 2003), and/or changes in the broader community (e.g., increase in predators or generalists near edges; Donovan et al., 1997, Polis, 1994). It is important to consider species temporally at both the larger landscape scale (i.e., the compartment) to understand the response of the broader population, as well as the finer scale (i.e., the stand) because this is where specific management activities and direct disturbance occur. Due to already small population sizes for select species, activities at the stand-level may quickly accumulate to impact the population at the landscape scale. However, most previous studies have only focused on stand-level impacts (i.e., changes in species composition or use of clearcuts; Corn and Bury, 1989, deMaynadier and Hunter, 1995, Semlitsch et al., 2008, Semlitsch et al., 2009, Sutton et al., 2013), which may not provide a broad enough look at the overall population or landscape.

Amphibians and reptiles are crucial ecosystem components due to their immense biomass (Burton and Likens, 1975, Semlitsch et al., 2014), important ecological functions and processes (Hocking and Babbitt, 2014, Valencia-Aguilar et al., 2013), and roles as intermediate consumers and predators (Polis, 1994, Pough et al., 1987). Response of amphibians and reptiles to forest management at the landscape scale, as well as finer scale habitat relationships and localized disturbance, are crucial components of habitat management at multiple levels (Suzuki et al., 2007). Due to factors such as their unique life history strategies and physiology, secretive nature, and minimal or sporadic surface activity, behaviorally cryptic or uncommon amphibian and reptile species can be especially difficult to monitor (Durso et al., 2011, Roloff et al., 2011, Ward et al., 2017). These factors may also contribute to why researchers often use targeted sampling for species of interest (Herbeck and Larsen, 1999, Peterman et al., 2011), group species by family or taxa (Rota et al., 2017, Wolf et al., 2016), or focus on common species or generalists (Renken et al., 2004, Todd and Rothermel, 2006). Focusing on more common or generalist species may increase the risk of masking forest management impacts on behaviorally cryptic or uncommon species, especially over longer temporal scales (Rompré et al., 2007). Additionally, many herpetofauna studies on short-term responses to forest management have been primarily focused on habitat selection immediately surrounding breeding habitat (e.g., ephemeral ponds and wetlands), which generally showed that both amphibians and reptiles avoided areas that were recently clearcut (Rittenhouse and Semlitsch, 2009, Semlitsch et al., 2008, Todd and Andrews, 2008, Todd et al., 2009). Although studies focused on breeding habitat and immediate responses to forest management provide important information for managers they do not consider nonbreeding habitat, the long-term use of harvested areas, or the potential for population recovery following disturbance due to their short timeframes, which may lead to different conclusions than long-term studies. Nonbreeding habitat provides an additional challenge for sampling already cryptic or uncommon herpetofauna and appropriate sample sizes can be difficult to obtain (Durso et al., 2011, Mazerolle et al., 2007). Here we investigate captures of uncommon herpetofauna throughout forested compartments (i.e., not constrained to breeding habitat) using a long-term dataset containing replicates, controls, and pre-treatment data collected as part of the Missouri Ozark Forest Ecosystem Project (MOFEP).

The Missouri Ozark Forest Ecosystem Project is a robust experiment designed to be completed over 100+ years to compare the long-term and landscape-scale effects of three forest management systems, including even-aged management (EAM), uneven-aged management (UAM), and no harvest management (NHM; control; Sheriff and He, 1997). To examine the impacts of forest management at multiple scales for uncommon herpetofauna species of the Missouri Ozarks, we experimentally assessed species-specific responses to these three forest management systems, as well as stand-level responses following two harvest entries. Additionally, to determine habitat relationships we assessed multiple covariates that can be easily measured and can be used to direct management.

The primary focus of these analyses is to determine species-specific habitat relationships and responses to forest management for previously unconsidered cryptic or uncommon species to add to the species-specific responses reported in Rota et al. (2017), as well as to further investigate select species with limited captures and unique life history strategies. We formulated the following broad hypotheses based on general life history strategies. We hypothesized that amphibians would show reduced captures in harvested stands with low basal area and have increased captures on North- and East-facing slopes, areas with higher flow accumulation, and closer proximity to stream and/or ponds based on breeding strategy. We hypothesized that reptiles would have increased captures in harvested stands with lower basal area relative to unharvested stands and increased captures on South- and West-facing slopes due to greater sun exposure compared to North- and East-facing slopes, with select species being tied to proximity to water sources.