Impacts of Fire Regime on Small Mammal Communities in California

California is considered a biodiversity hotspot based on its high species richness and endemism, and the considerable threats to the ecosystems that support these species. The Sierra Nevada region of California harbors much of the state's biodiversity, including unique small mammal communities which represent an important component of healthy ecosystems. Recently, the Sierra Nevada has experienced increasing amounts of climate change, drought, vegetation change, and alterations to fire regime. Small mammals are known to be sensitive to changes in climate and habitat. However, the long-term influence of these changes is poorly understood. In this study, we aim to understand the drivers of small mammal communities, with a focus on fire regime. We make use of historical surveys and resurveys from the Sierra Nevada, which provide high-quality species occurrence data from the last century. We compare richness and community turnover to fire regime, climate, and habitat variables using single- and multiple-variable linear regression. Results indicate that historic fire regime plays a key role explaining variation in modern small mammal richness across the Sierra Nevada. In the multivariate models, time since last fire and fire return interval significantly impact richness, and inclusion of these variables improve model fit compared to models with just climate and habitat. Additionally, time since last fire, number of fires and fire return interval have significant effects on small mammal community turnover. Overall, our work contributes knowledge of the factors influencing small mammal communities in an era of global change. TITLE : Impacts of Fire Regime on Small Mammal Communities in California

California is considered a biodiversity hotspot based on its high species richness and endemism, and the considerable threats to the ecosystems that support these species. The Sierra Nevada region of California harbors much of the state’s biodiversity, including unique small mammal communities which represent an important component of healthy ecosystems. Recently, the Sierra Nevada has experienced increasing amounts of climate change, drought, vegetation change, and alterations to fire regime. Small mammals are known to be sensitive to changes in climate and habitat. However, the long-term influence of these changes is poorly understood. In this study, we aim to understand the drivers of small mammal communities, with a focus on fire regime. We make use of historical surveys and resurveys from the Sierra Nevada, which provide high-quality species occurrence data from the last century. We compare richness and community turnover to fire regime, climate, and habitat variables using single- and multiple-variable linear regression. Results indicate that historic fire regime plays a key role explaining variation in modern small mammal richness across the Sierra Nevada. In the multivariate models, time since last fire and fire return interval significantly impact richness, and inclusion of these variables improve model fit compared to models with just climate and habitat. Additionally, time since last fire, number of fires and fire return interval have significant effects on small mammal community turnover. Overall, our work contributes knowledge of the factors influencing small mammal communities in an era of global change.

community ecology, species richness, species turnover, spatial distribution, ecological modeling, spatio-temporal

The Sierra Nevada region harbors much of California’s biodiversity, including 401 vertebrate species (Graber 1996) and over 3,000 vascular plants (Murphy et al. 2004). Over the last century, this region has undergone and continues to experience significant ecosystem change as a result of climate change and related impacts (Anderson 2004, Westerling et al. 2011, Schwartz et al. 2015, Stephens et al. 2018). Warming and drought have resulted in faunal species’ range shifts and declines (Tingley et al. 2009, Rowe et al. 2015, Roberts. et al. 2019b) as well as significant tree mortality, resulting in an increasing prevalence of severe wildfire (Stephens et al. 2018, Roberts et al. 2019a, 2019b). However, our understanding of how fire regime change has impacted ecosystems remains incomplete (Schwartz et al. 2015, Buechi et al. 2018, Stephens et al. 2021). Fire maintains the structure and heterogeneity of Sierra Nevada forests (Anderson 2004, Roberts et al. 2015). Pre-historic forests in California were maintained through low-to-mid severity cultural burning by Native Americans (Klimaszewski-Patterson and Mensing 2020), along with natural lightning fires; 20 th century fire suppression disrupted both Native American forest management and natural ecological processes (Calkin et al. 2015, Vachula et al. 2018, Stephens et al. 2021). This led to an acceleration in fire severity and area of high-severity fire starting in the 1980s and continuing through the 2000s, with future estimates predicting further escalations (Calkin et al. 2015, Buechi et al. 2018, Vachula et al. 2018). High-severity fires are generally categorized as being “stand-replacing” (Stevens et al. 2017, pg 29) with >70-90% tree mortality—a significant change in tree density, canopy cover and forest type—while low to moderate-severity fires are characterized by changes in shrub cover and understory, with 20-70% tree basal area mortality (Stevens et al. 2017). Predictions of Sierra Nevada climate and wildfire trajectories (Hayhoe et al. 2008, Westerling et al. 2011, Stephens et al. 2021) illustrate the need to better understand the combined effects of environmental change—drought, warming, habitat change, and fire—on forest ecosystems. One key component of Sierra Nevada ecosystems are small mammals (Kelt et al. 2013, 2017), who play many roles within Sierra Nevada ecosystems – as seed and fungal dispersers, prey, and habitat engineers (Copetto et al. 2006, Briggs et al. 2009). Changes to small mammal communities could thus significantly impact other aspects of Sierra Nevada ecosystems. Small mammals are also sensitive to fluctuations in vegetation and climate at both short and long time scales (Szpunar et al. 2008, Blois et al. 2010, Santos et al. 2017). For example, species distributions have shifted because of environmental changes, such as increase in temperature and drought, and land use change (Mortiz et al. 2008, Rowe et al. 2015, Santos et al. 2017). Small mammals are highly dependent on habitat structure, so vegetation changes can significantly influence mammal communities (Coppeto et al. 2006, Torre et al. 2015, Sollman et al 2015). In particular, small mammal community composition is strongly tied to macrohabitat characteristics, including canopy cover, shrub cover, and coarse woody debris (Converse et al. 2006, Coppeto et al. 2006, Sollman et al. 2015, Culhane et al. 2022). Global change drivers such as fire that influence these habitat variables may be particularly impactful for small mammals. In the Sierra Nevada over the last century, small mammal communities have undergone compositional change coincident with climate change (Moritz et al. 2008, Rowe et al. 2015). In the early 1900s, Joseph Grinnell and colleagues conducted surveys that generated a baseline of mammal species occurrences throughout the Sierra Nevada (Moritz et al. 2008). These sites were resurveyed in the early 2000s by the University of California, Berkeley-led Grinnell Resurvey Project, generating a high-quality, temporally significant dataset of Sierra Nevada small mammal change (Moritz et al. 2008, Rowe et al. 2015). Across three regions, 25 of 34 species analyzed shifted their elevational ranges, a majority in ways that concurred with climate change (Rowe et al. 2015). A subsequent study on 8 small mammal species determined that vegetation in addition to climate helped predict range shifts (Santos et al. 2017), and thus likely also explained changes in small mammal communities. Several other studies have demonstrated connections between habitat and small mammal communities in the Sierra Nevada (Copetto et al. 2006, Kelt et al. 2013, Sollman et al. 2015), supporting linkages between small mammal community change and global change. In comparison to habitat and climate, the long-term impact of fire on small mammal communities is poorly understood. While a connection between elevational range shifts and historic fire suppression in the Sierra Nevada has been postulated (Rowe et al. 2015), relatively few analyses have examined the long-term impact of alterations in natural fire regime on small mammal communities. Studies on prescribed fire in the Sierra Nevada (e.g., Converse et al. 2005, 2006, Amacher et al. 2008, Monroe and Converse 2008) provide valuable insight into small mammal responses to fire, but do not address the full extent of potential wildfire impacts, especially high-severity wildfire. Several recent studies have explored the connection between wildfire, habitat change, and Sierra Nevada small mammals. These studies have concentrated on immediate to short-term responses to fire (Roberts et al. 2008, Borchert et al. 2014, Culhane et al. 2022), with a maximum time since fire of 15 years (Roberts et al. 2015). Additionally, many of these studies have focused on individual species, while only a few have focused on the full small mammal community (Roberts et al. 2015, Culhane et al. 2022). Research in other regions, such as in Australia and other parts of North America, has identified factors relating to the long-term influence of fire on small mammal communities (Krefting and Ahlgren 1974, Kelly et al. 2011, Fontaine and Kennedy 2012, Griffiths et al. 2014). A similar understanding of fire impacts on small mammal communities is needed in the California Sierra Nevada, where fire is an important, prevalent, and increasingly impactful factor within the landscape. In this study, we explore the drivers of small mammal richness in the Sierra Nevada, focusing on three sets of variables with established or inferred relationships with small mammals: fire, habitat and climate. We specifically focus on fire because impacts of climate and vegetation on small mammal communities are already well-established (Carey and Wilson 2001, Copetto et al. 2006, Rowe et al. 2015), though we include all three sets of variables in most of our analyses because of their intertwined nature. Because elevation is the dominant physical gradient within the Sierra Nevada, we first ask: 1) How does small mammal richness change across latitudinal (regional) and elevation gradients? Next, we examine the drivers of small mammal richness using two complementary multivariate modeling approaches. To inform these models, we examine univariate relationships between richness and environment and collinearity amongst variables, and then ask the following questions: 2) What combination of variables best explains variation in small mammal richness across the Sierra Nevada? 3) When multiple variables are included in models, do we see substantially different relationships than with individual variables? 4) Does the inclusion of fire variables explain significantly more variation in small mammal richness than variation explained by vegetation and climate variables? Finally, having established the importance of fire on spatial variation in small mammals, we examine if fire regime has impacted small mammal community turnover over the last century. We ask: 5) Is there a relationship between compositional turnover and fire?

OVERALL APPROACH We relied on data collected by the Grinnell Survey and Resurvey Project, an effort in the early 2000s to document change in California vertebrate species occurrences (Mortiz et al. 2008). The Grinnell Project established three elevational transects in the northern, central, and southern regions of the Sierra Nevada and southern Cascade mountain ranges in California: the Lassen transect (LA; sampled from 2006-2009), the Yosemite transect (YO; sampled 2003-2006), and the Sequoia transect (SS; sampled 2008-2012). Each area was originally sampled by Joseph Grinnell and colleagues from 1904-1940s and included detailed metadata that allowed the resurvey team to approximate the locations of trapping sites, making the two datasets comparable. This dataset includes high-quality small mammal occurrence data in areas minimally disturbed by human impacts. Moreover, the availability of both historic and modern small mammal data allows tracking of change over time. The Grinnell Project based many of their original analyses on lifezones (Figure 1; Moritz et al. 2008), but different lifezones encompass different amounts of area, which may affect comparison of community-level metrics through species-area effects (Rosenzweig 1995). Thus, we defined sample areas as evenly-spaced 200m elevation bands along the approximately 0-3,600m elevational gradient in each region, which should minimize (though not eliminate; Moradi et al. 2020, Karger et al. 2011) this effect. Lassen included data from 0-2600m, Yosemite 0-3400m and Sequoia 0-3600m, resulting in 48 total sample areas across California. This approach captures biologically and environmentally relevant variation across elevation, a dominant gradient in the Sierra Nevada. Additionally, calculating small mammal richness in elevation bands buffers richness against hyper-local effects of individual occurrences and short-term sampling, since richness emerges at a regional scale over longer time scales (White et al. 2010, Shuai et al. 2017). SMALL MAMMAL DATA Modern Richness The small mammal data collected by the Grinnell Project team was described in Mortiz et al. (2008) and Rowe et al. (2015). Small mammal species were documented across 166 sites within the three regions (LA, YO, SS). At each site, 40-150 traps were placed for 2-4 nights, including both Sherman and Tomahawk traps to capture a range of body sizes, and species occurrences were recorded. Subsequently, the authors developed occupancy models for each species with enough data to estimate total elevational range limits (Rowe et al. 2015). We calculated richness within each sample area based on the reported modern upper and lower elevation limits for each species (Rowe et al. 2015, Supplementary Material 1). We chose this method over summing observed presences from the trapping data, as occupancy modeling accounts for sampling bias that may occur during surveys and controls for false absences (Rowe et al 2015). Compositional Turnover We calculated compositional turnover between the historic and modern datasets using Jaccard dissimilarity based on presence/absence data, relying on occupancy-modeled elevation range limits for each species in the historic and modern samples. Compositional turnover represents a more in-depth representation of community change compared to richness change, as richness-only measurements can obscure species-specific shifts. ENVIRONMENTAL DATA We characterized the environment of each sample area based on fire regime, climate, and vegetation (Table 1). For each sample area, we generated a series of random points at evenly spaced 5km intervals within the sample area and extracted the environmental values at each point. Environmental values were averaged within each sample area for the three transects, except for number of fires, which were summed, and fire return interval, which was based upon number of fires. Fire Regime We quantified fire regime using several metrics, including time since last fire (TSLF), number of fires occurring across the timeframe for each area (num_fires), and fire return interval (FRI). To best represent the fire regime across the last century, we included fire data from 1908 (see description of the FRAP dataset below) through the first year of small mammal surveys in each transect. The number of fires occurring within a timeframe is a common metric to describe fire regime, because this substantially influences vegetation growth and behavior (Safford and Water 2014). Similarly, time since last fire is related to the state of the current landscape as a result of the most recent fire (Safford and Water 2014). In addition to mean time since last fire (mean_TSLF), we also calculated the standard deviation (SD_TSLF) to represent the variance between the most recent fire occurrences within the sample area. Fire return interval is calculated using number of fires, representing the amount of time between fires at a site (Safford and Water 2014). For our calculations, we used the California Department of Forestry and Fire Protection: Fire Resource Assessment Program dataset (FRAP 2022), which includes fire perimeter data from 1878 to the present, with complete data after 1908 (Safford and Water 2014). We extracted fire data for every fire within 1km of the random points within each sample area. We included all fires from the start of the complete dataset in 1908 to the modern small mammal resurvey periods for each transect (YO and standard deviation of the time since last fire (mean_TSLF, SD_TSLF) based on the most recent fire at each point within a sample area. Number of fires (num_fire) was calculated as the total number of fires experienced throughout the sample area over the timeframe (1908 – first surveying year). We calculated fire return interval using the following equation: FRI = (Number of years/Number of fires) +1, in which the number of years corresponds to the length of the timeframe. All calculations are based on equations described in Safford and Water (2014). Vegetation For vegetation, we used the 2001 version of the Landscape Fire and Resource Management Planning Tools program LANDFIRE (LANDFIRE 2001). This database was created in 2002 to inform fire management and support habitat and wildfire research. Specifically, we downloaded the variables forest canopy height and forest canopy cover at each random point. This data includes vegetation data through the year 2001, just prior to when the small mammal surveys were conducted (2003 – 2012). Canopy height represents the average height of forested regions, and canopy cover represents the percent cover of the tree canopy. Climate For climate, we relied on modern PRISM data collected at 4km resolution (PRISM Group, 2014). For each of the transect years (YO: 2003-2006, LA: 2006-2008, SS: 2008-2012) we downloaded annual mean temperature (tmean) and total annual precipitation (ppt) at each random point, calculated the average annual temperature and average total precipitation across the sampling period, then averaged each variable across all random points within a sample area. ANALYSES We compared modern small mammal richness to environmental and spatial variables to determine the most significant factors impacting small mammal communities. Additionally, we compared community turnover to fire regime variables to examine effects of fire on small mammals over the last century. VARIATION ACROSS ELEVATION AND REGION We first determined trends in richness across region and elevation, then compared each environmental variable to elevation (Supplemental Figure 1). Elevation was quantified as the midpoint of each elevation band for a sample area. Next, we used ANOVA (region/transect) and linear regression (elevation) to identify significant variation in richness or environment across gradients. This indicated whether any of the observed relationships in single-variable models were simply proxies for elevation or transect. DRIVERS OF CONTEMPORARY SMALL MAMMAL RICHNESS Univariate Relationships between Richness and Environment We first used linear regression to examine whether individual relationships between richness and environmental variables followed expected relationships (Supplemental Table 1). We then examined collinearity between environmental variables using absolute Pearson correlation coefficients (Supplemental Figure 2). Multivariate Relationships between Richness and Environment Based on our understanding of collinearity (Supplemental Figure 2), individual variable and richness relationships, and biological factors likely to most strongly influence small mammal richness (Supplemental Table 1), we developed two sets of multivariate regression models. The first multivariate model included small mammal richness as the independent variable and all other variables, including elevation and region, as predictors. The second multivariate model was based on variables developed through principal components analysis (PCA). We used principal component analysis to reduce dimensionality and address collinearity among variables. Our aim with this was to identify if a combination of multiple variables better represented richness than any single variable relationship, even after accounting for collinearity. We excluded region from the PCA model because this variable is categorical. We then developed a multivariate model with small mammal richness as the independent variable and the first four principal component axes as the predictor variables. For each approach, we used AIC to simplify the initial model to a final best-fit model, using AIC corrected to account for small sample sizes (AICc). RELATIONSHIP BETWEEN FIRE REGIME AND SMALL MAMMAL COMMUNITY CHANGE To examine the relationship between fire and community change through time, we developed univariate and multivariate regression models comparing Jaccard dissimilarity of the small mammal community between the historic and modern sampling periods within each sample area and each of the fire regime variables.

VARIATION ACROSS ELEVATION AND REGION One-way ANOVA demonstrated that total richness did not significantly differ across transects (Figure 2A; F(2, 45) = 1.941, p > 0.05). Generally, richness increased with elevation (Figure 2B-D), with significant richness-elevation relationships in the Lassen (p DRIVERS OF SMALL MAMMAL RICHNESS Individual Relationships between Richness and Environmental Variables Average annual precipitation was positively associated with small mammal richness (p <0.001, adj. R 2 = 0.16), while temperature was negatively associated with richness (p <0.001, adj. R 2 = 0.20) (Figure 3A, B). Canopy cover and canopy height had strong positive relationships with small mammal richness (CC: p < 0.001, adj. R 2 = 0.34; CH: p < 0.001, adj. R 2 = 0.43) (Figure 3C, D). There was not a significant relationship between small mammal richness and time since last fire (mean_TSLF, SD_TSLF), number of fires, or fire return interval (Figure 3E-H). Multi-variable Relationships between Richness and Environmental Variables For the All-Variables Model, after variable selection from the initial multivariate model (Table 1), six models fell within 2 ∆AICc units of the top model (Table 2). The best model included canopy height, fire return interval, mean_TSLF, SD_TSLF, temperature mean, and region (p < 0.001, adj.R 2 = 0.61, AICc =212.7). In this model, fire return interval, SD_TSLF and temperature had negative relationships with richness, while canopy height, mean_TSLF, and region had positive relationships. The next best model was nearly the same, with temperature instead of elevation included in the model (p <0.001, adj.R 2 = 0.60, AICc =213.28). Canopy height was included in all six models, while temperature was present in all but the second model, indicating strong covariation between temperature and elevation. Fire variables were included a majority of the models, including the top three. In the All-Variables PCA Model, the first PCA axis explained 39% of the variation and primarily represented climate, with temperature mean, precipitation, and elevation midpoint having the highest loadings (Supplemental Tables 2 & 3). The second PCA axis explained 24% of variation and primarily represented vegetation, as canopy cover, canopy height and fire return interval were most strongly loaded on this axis. The third PCA axis explained 14% of variation, with mean_TSLF and SD_TSFL having the highest loadings, predominantly representing time since last fire. The fourth PCA axis explained 13% of the total variation, and mean_TSLF, temperature mean, and elevation midpoint had the highest loadings. Together, these four variables explain 91% of the variation in environmental variables among sample areas (Supplemental Table 2); we thus used the first four principal components for the PCA model. The top PCA model included PC1, PC2, and PC3 as predictors of richness (Table 2; p < 0.001, adj. R 2 = 0.48, AICc = 219.8). However, three other models performed within 2 ∆AICc units of the top model (Table 2). COMMUNITY TURNOVER Univariate Models There were significant positive relationships between community turnover and mean_TSLF (p < 0.01, adj.R 2 = 0.12), SD_TSLF (p < 0.05, adj.R 2 = 0.07), and number of fires (p < 0.05, adj.R 2 = 0.08). There was no significant relationship between turnover and fire return interval. Multivariate Model The top model for community turnover included positive relationships with number of fires and mean_TSLF (Table 2; p <0.0001, adj.R 2 = 0.30, AICc = -83.4). A second model was within 2 ∆AICc units of the top model, and it included positive relationships with number of fires, fire return interval and mean_TSLF (Table 2; p <0.01, adj.R 2 = 0.29, AICc = -81.7).

A century of fire suppression in the Sierra Nevada has resulted in changes to fire regimes that will substantially influence ecosystems going forward (Calkin et al. 2015, Buechi et al. 2018, Vachula et al. 2018). We sought to better understand how fire impacts small mammal communities, both isolated from, and in tandem with, vegetation and climate. Metrics of fire emerged as important variables in both small mammal richness and community turnover. Overall, variation in small mammal richness was best explained by including fire, climate and vegetation rather than just climate and vegetation alone. OVERALL IMPACTS OF FIRE, VEGETATION, AND CLIMATE ON SMALL MAMMALS Fire suppression has negatively impacted Sierra Nevada forests in two main ways: it has 1) reduced heterogeneity among sites and 2) increased the total amount of fuel (thus increasing canopy cover), leading to high-severity megafires (Calkin et al. 2015, Buechi et al. 2018, Vachula et al. 2018). Given the importance of heterogeneity and canopy cover on small mammals (Coppeto et al. 2006, Borchert et al. 2014, Sollman et al. 2014), we expected to see positive relationships between fire regime variables and small mammal richness (Supplemental Table 1); however, we didn’t observe any significant relationships between modern richness and individual fire variables. Other studies have also found no or unclear relationships between richness and fire, though there is some evidence that frequent low-medium severity fires generate habitats that support small mammal communities (Roberts et al. 2008, Roberts et al. 2015, Culhane et al. 2022, Mike Conner et al. 2022). The importance of fire, however, emerged in the multivariate models, where fire was a necessary variable explaining richness in most of the top models. The best multivariate models contained representatives of all variable categories: fire, climate and vegetation. The top model included fire return interval and both mean and standard deviation of time since last fire. While these variables did not have significant univariate relationships with richness, when combined with temperature and canopy cover, they significantly contributed to explaining variation in small mammal richness. Fire is therefore clearly important, but our results also reaffirm the importance of vegetation and climate on small mammal richness. The best multivariate model and five of the top six models included canopy height and temperature, in addition to other variables. Vegetation, especially canopy height, emerged as the strongest single-variable indicator of small mammal richness. Both temperature and precipitation were also significantly associated with small mammal richness (Figure 3A, B), though they did not explain as much variation in richness as canopy height and cover (Adj. R 2 of 0.20 and 0.16 for temperature and precipitation versus 0.43 and 0.34 for canopy height and cover, respectively; Figure 2). Together, these results are consistent with previous research that identified both climate and vegetation as driving small mammal richness in the Sierra Nevada (Rubidge et al 2011, Kelt et al. 2017, Santos et al. 2017). Below, we explore each of these variables (fire, climate, and vegetation) in more depth. We also found that community turnover over the last century was best explained by multiple fire variables (Table 2). Small mammal communities experienced less change when there was more frequent, consistent and recent fire – a regime likely similar to the original survey period, before fire suppression efforts (Calkin et al. 2015, Vachula et al. 2018, Stephens et al. 2021). IMPORTANCE OF INDIVIDUAL VARIABLES Four environmental variables were consistently included in the top models: fire return interval, time since last fire, canopy height, and temperature. We originally expected that richness would positively correlate with smaller fire return intervals, as this likely corresponds with a patchy, heterogenous landscape (Supplemental Table 1), an expectation our models uphold. Previous studies conducted in frequent-fire forests support our results that fire regime impacts small mammal occurrence; Borchert et al (2014) showed that years-after-fire significantly impacted small mammal composition in the California San Bernadino Mountains while Lindenmayer et al (2016) demonstrated that time since last fire and fire frequency significantly impacted the presence of Australian terrestrial rodent species. Kelly et al. (2011) also showed that Australian rodent species were impacted by time since fire, though only one species had a positive relationship with burnt vegetation. However, Monamy and Fox (2000) identified that the effects of time since last fire were mitigated by vegetation regrowth, with some mammal species recolonizing quickly after fire, concurrent with higher vegetation density. Other studies on prescribed fire treatments have shown relationships between species abundance and interval between fires (Converse et al. 2005, Mike Conner et al 2022). Generally, too-frequent fire can reduce availability of habitat such as coarse woody debris, but without fire, habitat quality tends to degrade (Converse et al. 2005, Mike Conner et al. 2022,). Our single-variable results generally reconstructed established relationships between richness and vegetation, climate, and elevation in California. Canopy height, canopy cover and precipitation were positively associated with small mammal richness, while richness decreased with higher temperatures. This matches our current understanding of the importance of these variables on richness in the Sierra Nevada. The influence of climate and vegetation on small mammals, however, has considerable nuance. Previous studies have disagreed on the extent to which different variables impacts small mammals, with many highlighting climate (and often temperature) as the primary indicator of small mammal distributions (Mortiz et al. 2008, Rubidge et al 2011, Rowe et al. 2015, Santos et al. 2017). For example, Rowe et al. (2015) observed species shifts coinciding with temperature but not precipitation change. Likewise, Santos et al. (2017) identified that climate had greater effects on species ranges than habitat. Studies exploring habitat characteristics of vegetation show that some species prefer higher canopy openness and shrub density (Coppeto et al. 2006, Sollman et al. 2014), while certain specialist species prefer greater canopy cover (Meyer et al. 2009, Kelt et al. 2013), in line with variation in individual niches. These mixed results on the influence of variables among different species (Carey and Wilson 2001, Coppeto et al. 2006, Sollman et al. 2014) can make community-wide trends difficult to parse. Since our study explored richness on a regional scale, it makes sense that, overall, canopy cover and height would have a positive effect on small mammals, especially given the close relationship between vegetation and elevation life zones. In particular, canopy height is associated with older, more mature forests, which tend to support greater richness (Meyer et al. 2009, Kelt et al. 2013, Sollman et al. 2014). CAVEATS Fire is an increasingly dominant force within the Sierra Nevada and worldwide. This study contributes a better understanding of the intertwined impacts of fire, climate, and vegetation on small mammal communities. As our focus was on the importance of fire regime, we did not explore climate and vegetation variables as deeply as other studies. We relied primarily on annual temperature and precipitation, while other studies have included variables related to climate seasonality (Rubidge et al 2011). In addition, we chose not to use habitat type as a vegetation characteristic due to the difficulty of averaging this categorical value across sites, instead relying on numeric vegetation metrics. However, the climate and vegetation metrics we included have demonstrable and strong influences on small mammals (Sollman et al. 2014, Rowe et al. 2015, Santos et al. 2017), so our reliance on just temperature and precipitation for climate, and canopy height and cover for vegetation, likely captures the dominant features explaining richness variation at a regional scale. We also focused on overall small mammal richness to capture an important and underexplored metric. Many studies observing the effects of wildfire on small mammals, especially in California, focus on abundance rather than richness, as demonstrated by a meta-analysis summarizing the impacts of fire on small mammals (Griffiths et al. 2014). However, focusing on abundance or species distributions tends to prioritize the most common species, as rare species (often also specialists) are often excluded from analyses (e.g., Roberts et al. 2008, Borchert et al. 2014, Chia et al. 2016). For our richness metric, we compressed data into elevation bands to best describe landscape-level richness. Due to our reliance on occupancy-modeled elevation ranges, which are themselves affected by rarity, we still did not capture the full small mammal community richness. However, while richness is a blunt tool, it captures a cross-section of the community. Despite these caveats, our results conformed with previously established relationships between richness and vegetation, climate, and elevation in California (Coppeto et al. 2006, Mortiz et al. 2008, Kelt et al. 2013, Rowe et al. 2015) indicating that our data captured the dominant gradients in both mammal communities and environment in the Sierra Nevada, as well as relationships among these variables.

AND NEXT STEPS A century of fire suppression in the Sierra Nevada has resulted in changes to fire regimes that will likely substantially impacts small mammal communities going forward (Calkin et al. 2015, Buechi et al. 2018, Vachula et al. 2018, Stephens et al. 2021). We sought to understand how fire influences small mammal communities, both isolated from, and in tandem with, vegetation and climate. Overall, fire was an important factor explaining small mammal richness and community turnover in almost every analysis we completed, and the addition of fire variables improved models versus those with just climate and vegetation alone. We observed higher small mammal richness at sites with shorter fire return intervals. Over the last century, small mammal communities that experienced more time since last fire and a higher number of fires showed greater turnover. Our results thus support hypotheses that small mammals prefer frequent-fire, heterogenous landscapes (Robert et al. 2015, Kelt et al. 2017). However, a positive association between fire and small mammal diversity is tied to the assumption that a frequent-fire regime creates patchy, irregular landscapes. With the increase in high-severity stand-replacing megafires, many landscapes are experiencing complete loss rather than healthy turnover (Williams et al. 2022), and high-severity fire negatively impacts canopy cover and canopy height, important variables for small mammal richness. Recent research demonstrates when, how, and if areas with high severity wildfires will recover (Zwolak and Foresman 2007, Culhane et al. 2022), but more research is needed to completely understand the effects of high severity fire on small mammals. Because of this, future studies should endeavor to include fire when researching small mammal community change over space and time.

Amacher, A. J., Barrett, R. H., Moghaddas, J. J. and Stephens, S. L. 2008. Preliminary effects of fire and mechanical fuel treatments on the abundance of small mammals in the mixed-conifer forest of the Sierra Nevada. - Forest Ecology and Management 255: 3193–3202. Anderson, R. S. 2004. Response of Sierra Nevada vegetation and fire regimes to past climate changes. - In: Murphy, Dennis D. and Stine, Peter A., editors. Proceedings of the Sierra Nevada Science Symposium. Gen. Tech. Rep. PSW-GTR-193. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture: 47-50 in press. Blois, J. L., McGuire, J. L. and Hadly, E. A. 2010. Small mammal diversity loss in response to late-Pleistocene climatic change. - Nature 465: 771–774. Borchert, M. I., Farr, D. P., Rimbenieks-Negrete, M. A. and Pawlowski, M. N. 2014. Responses of Small Mammals to Wildfire in a Mixed Conifer Forest in the San Bernardino Mountains, California. - soca 113: 81–95. Briggs, J. S., Wall, S. B. V. and Jenkins, S. H. 2009. Forest rodents provide directed dispersal of Jeffrey pine seeds. - Ecology 90: 675–687. Buechi, H., Weber, P., Heard, S., Cameron, D., Plantinga, A. J., Buechi, H., Weber, P., Heard, S., Cameron, D. and Plantinga, A. J. 2021. Long-term trends in wildfire damages in California. - Int. J. Wildland Fire 30: 757–762. Calkin, D. E., Thompson, M. P. and Finney, M. A. 2015. Negative consequences of positive feedbacks in US wildfire management. - Forest Ecosystems. 2:9. doi: 10.1186/s40663-015-0033-8. in press. Carey, A. B. and Wilson, S. M. 2001. Induced Spatial Heterogeneity in Forest Canopies: Responses of Small Mammals. - The Journal of Wildlife Management 65: 1014–1027. Chia, E. K., Bassett, M., Leonard, S. W. J., Holland, G. J., Ritchie, E. G., Clarke, M. F. and Bennett, A. F. 2016. Effects of the fire regime on mammal occurrence after wildfire: Site effects vs landscape context in fire-prone forests. - Forest Ecology and Management 363: 130–139. Converse, S. J., White, G. C., Farris, K. L. and Zack, S. 2006a. Small Mammals and Forest Fuel Reduction: National-Scale Responses to Fire and Fire Surrogates. - Ecological Applications 16: 1717–1729. Converse, S. J., Block, W. M. and White, G. C. 2006b. Small mammal population and habitat responses to forest thinning and prescribed fire. - Forest Ecology and Management 228: 263–273. Coppeto, S. A., Kelt, D. A., Van Vuren, D. H., Wilson, J. A. and Bigelow, S. 2006. Habitat Associations of Small Mammals at Two Spatial Scales in the Northern Sierra Nevada. - Journal of Mammalogy 87: 402–413. Culhane, K., Sollmann, R., White, A. M., Tarbill, G. L., Cooper, S. D. and Young, H. S. 2022. Small mammal responses to fire severity mediated by vegetation characteristics and species traits. - Ecology and Evolution. 12(5): e8918 in press. Fontaine, J. B. and Kennedy, P. L. 2012. Meta-analysis of avian and small-mammal response to fire severity and fire surrogate treatments in U.S. fire-prone forests. - Ecol Appl 22: 1547–1561. FRAP 2022. Historical Fire Perimeters 2022: GIS Mapping and Data Analytics | CAL FIRE. Graber, D. M. 1996. Sierra Nevada Ecosystem Project final report to Congress : Status of Terrestrial Vertebrates.: Ch.33. Graumlich, L. J. 1993. A 1000-Year Record of Temperature and Precipitation in the Sierra Nevada. - Quaternary Research 39: 249–255. Griffiths, A. D., Brook, B. W., Griffiths, A. D. and Brook, B. W. 2014. Effect of fire on small mammals: a systematic review. - Int. J. Wildland Fire 23: 1034–1043. Hayhoe, K., Cayan, D., Field, C. B., Frumhoff, P. C., Maurer, E. P., Miller, N. L., Moser, S. C., Schneider, S. H., Cahill, K. N., Cleland, E. E., Dale, L., Drapek, R., Hanemann, R. M., Kalkstein, L. S., Lenihan, J., Lunch, C. K., Neilson, R. P., Sheridan, S. C. and Verville, J. H. 2004. Emissions pathways, climate change, and impacts on California. - Proceedings of the National Academy of Sciences 101: 12422–12427. Karger, D. N., Kluge, J., Krömer, T., Hemp, A., Lehnert, M. and Kessler, M. 2011. The effect of area on local and regional elevational patterns of species richness. - Journal of Biogeography 38: 1177–1185. Kelly, L. T., Nimmo, D. G., Spence-Bailey, L. M., Haslem, A., Watson, S. J., Clarke, M. F. and Bennett, A. F. 2011. Influence of fire history on small mammal distributions: insights from a 100-year post-fire chronosequence. - Diversity and Distributions 17: 462–473. Kelt, D. A., Van Vuren, D. H., Johnson, M. L., Wilson, J. A., Innes, R. J., Jesmer, B. R., Ingram, K. P., Smith, J. R., Bigelow, S. W., Burnett, R. D. and Stine, P. A. 2013. Small mammals exhibit limited spatiotemporal structure in Sierra Nevada forests. - Journal of Mammalogy 94: 1197–1213. Kelt, D. A., Sollmann, R., White, A. M., Roberts, S. L. and Van Vuren, D. H. 2017. Diversity of small mammals in the Sierra Nevada: filtering by natural selection or by anthropogenic activities? - Journal of Mammalogy 98: 85–93. Klimaszewski-Patterson, A. and Mensing, S. 2020. Paleoecological and paleolandscape modeling support for pre-Columbian burning by Native Americans in the Golden Trout Wilderness Area, California, USA. - Landscape Ecol 35: 2659–2678. LANDFIRE 2001. Canopy Fuel: Forest Canopy Cover, Forest Canopy Height. Lindenmayer, D. B., Blanchard, W., MacGregor, C., Barton, P., Banks, S. C., Crane, M., Michael, D., Okada, S., Berry, L., Florance, D. and Gill, M. 2016. Temporal trends in mammal responses to fire reveals the complex effects of fire regime attributes. - Ecological Applications 26: 557–573. Meyer, M., Kelt, D. and North, M. 2009. Microhabitat Associations of Northern Flying Squirrels in Burned and Thinned Forest Stands of the Sierra Nevada. - The American Midland Naturalist 157: 202–211. Mike Conner, L., Holland, A. and Morris, G. 2022. Fire exclusion and fire return interval affect small mammal populations in longleaf pine forests. - Forest Ecology and Management 520: 120352. Monamy, V. and Fox, B. J. 2000. Small mammal succession is determined by vegetation density rather than time elapsed since disturbance. - Austral Ecology 25: 580–587. Monroe, M. E. and Converse, S. J. 2006. The effects of early season and late season prescribed fires on small mammals in a Sierra Nevada mixed conifer forest. - Forest Ecology and Management 236: 229–240. Moradi, H., Fattorini, S. and Oldeland, J. 2020. Influence of elevation on the species–area relationship. - Journal of Biogeography 47: 2029–2041. Moritz, C., Patton, J., Conroy, C., Parra, J., White, G. and Beissinger, S. 2008. Impact of a Century of Climate Change on Small-Mammal Communities in Yosemite National Park, USA. - Science (New York, N.Y.) 322: 261–4. Murphy, D. D., Fleishman, E. and Stine, P. A. 2004. Biodiversity in the Sierra Nevada. - In: Murphy, Dennis D. and Stine, Peter A., editors. Proceedings of the Sierra Nevada Science Symposium. Gen. Tech. Rep. PSW-GTR-193. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture: 167-174 in press. PRISM Group, Oregon State University, https://prism.oregonstate.edu, data created 4 Feb 2014, accessed 18 April 2025. Roberts, S. L., van Wagtendonk, J. W., Miles, A. K., Kelt, D. A. and Lutz, J. A. 2008. Modeling the Effects of Fire Severity and Spatial Complexity on Small Mammals in Yosemite National Park, California. - fire ecol 4: 83–104. Roberts, S. L., Kelt, D. A., van Wagtendonk, J. W., Miles, A. K. and Meyer, M. D. 2015. Effects of fire on small mammal communities in frequent-fire forests in California. - Journal of Mammalogy 96: 107–119. Roberts, C. P., Donovan, V. M., Wonkka, C. L., Powell, L. A., Allen, C. R., Angeler, D. G., Wedin, D. A. and Twidwell, D. 2019a. Fire legacies in eastern ponderosa pine forests. - Ecology and Evolution 9: 1869–1879. Roberts, L. J., Burnett, R., Tietz, J. and Veloz, S. 2019b. Recent drought and tree mortality effects on the avian community in southern Sierra Nevada: a glimpse of the future? - Ecol Appl 29: e01848. Rosenzweig, M. L. 1995. Species Diversity in Space and Time. - Cambridge University Press. Rowe, K. C., Rowe, K. M. C., Tingley, M. W., Koo, M. S., Patton, J. L., Conroy, C. J., Perrine, J. D., Beissinger, S. R. and Moritz, C. 2015. Spatially heterogeneous impact of climate change on small mammals of montane California. - Proceedings of the Royal Society B: Biological Sciences 282: 20141857. Safford, H. D. and Water, K. M. V. de 2014. Using fire return interval departure (FRID) analysis to map spatial and temporal changes in fire frequency on national forest lands in California. - Res. Pap. PSW-RP-266. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station. 59 p in press. Schwartz, M. W., Butt, N., Dolanc, C. R., Holguin, A., Moritz, M. A., North, M. P., Safford, H. D., Stephenson, N. L., Thorne, J. H. and van Mantgem, P. J. 2015. Increasing elevation of fire in the Sierra Nevada and implications for forest change. - Ecosphere 6: art121. Shuai, L.-Y., Ren, C.-L., Yan, W.-B., Song, Y.-L. and Zeng, Z.-G. 2017. Different elevational patterns of rodent species richness between the southern and northern slopes of a mountain. - Sci Rep 7: 8743. Sollmann, R., White, A. M., Gardner, B. and Manley, P. N. 2015. Investigating the effects of forest structure on the small mammal community in frequent-fire coniferous forests using capture-recapture models for stratified populations. - Mamm Biol 80: 247–254. Stephens, S. L., Collins, B. M., Fettig, C. J., Finney, M. A., Hoffman, C. M., Knapp, E. E., North, M. P., Safford, H. and Wayman, R. B. 2018. Drought, Tree Mortality, and Wildfire in Forests Adapted to Frequent Fire. - BioScience 68: 77–88. Stephens, S. L., Thompson, S., Boisramé, G., Collins, B. M., Ponisio, L. C., Rakhmatulina, E., Steel, Z. L., Stevens, J. T., Wagtendonk, J. W. van and Wilkin, K. 2021. Fire, water, and biodiversity in the Sierra Nevada: a possible triple win. - Environ. Res. Commun. 3: 081004. Stevens, J. T., Collins, B. M., Miller, J. D., North, M. P. and Stephens, S. L. 2017. Changing spatial patterns of stand-replacing fire in California conifer forests. - Forest Ecology and Management 406: 28–36. Szpunar, G., Aloise, G., Mazzotti, S., Nieder, L. and Cristaldi, M. 2008. Effect of global climate change on terrestrial small mammal communities in Italy. - Fresenius Environ Bull in press. Tingley, M. W., Monahan, W. B., Beissinger, S. R. and Moritz, C. 2009. Birds track their Grinnellian niche through a century of climate change. - Proceedings of the National Academy of Sciences 106: 19637–19643. Torre, I., Gracia-Quintas, L., Arrizabalaga, A., Baucells, J. and Díaz, M. 2015. Are recent changes in the terrestrial small mammal communities related to land use change? A test using pellet analyses. - Ecological Research 30: 813–819. Urquizo, J. H., Barquez, R. M. and Ferro, I. 2022. The elevation gradient of small mammals richness and abundance in the dry Andes of northwestern Argentina. - Journal of Arid Environments 204: 104792. Vachula, R. S., Russell, J. M. and Huang, Y. 2018. New insights into the changing relationships between fire, humans, and climate in the Yosemite region of California’s Sierra Nevada. 2018: GC44C-03. Westerling, A. L., Bryant, B. P., Preisler, H. K., Holmes, T. P., Hidalgo, H. G., Das, T. and Shrestha, S. R. 2011. Climate change and growth scenarios for California wildfire. - Climatic Change 109: 445–463. White, E. P., Ernest, S. K. M., Adler, P. B., Hurlbert, A. H. and Lyons, S. K. 2010. Integrating spatial and temporal approaches to understanding species richness. - Philos Trans R Soc Lond B Biol Sci 365: 3633–3643. Williams, J. N., Safford, H. D., Enstice, N., Steel, Z. L. and Paulson, A. K. 2023. High-severity burned area and proportion exceed historic conditions in Sierra Nevada, California, and adjacent ranges. - Ecosphere 14: e4397. Zwolak, R. and Foresman, K. R. 2007. Effects of a stand-replacing fire on small-mammal communities in montane forest. - Can. J. Zool. 85: 815–822. FIGURES AND TABLES Figure 1: ( A) Map of California with elevation lifezones color-coded. Points represent the survey sites from the Grinnell Resurvey Project (Rowe et al. 2015). Figure 2: Relationship between small mammal richness and spatial variables. (A) Boxplot demonstrating relationship between region/transect and small mammal richness. (B - D) Linear regression models comparing small mammal richness and (B) Lassen (LA) transect elevation, (C) Yosemite (YO) transect elevation, and (D) Sequoia (SS) transect elevation. Figure 3: Single-variable linear regression models comparing modern small mammal richness and climate, vegetation and fire variables (Table 1). Variables are organized and color-coded by category: climate variables in blue on the top row (A, B), vegetation variables in green in the 2 nd row (C, D), and fire variables in red on the bottom two rows (E, F, G, H). TSLF = time since last fire, showing mean and standard deviation (sd). Figure 4: Linear regression models comparing community turnover to A) mean time since last fire (TSLF); B) standard deviation of TSLF; C) number of fires; and D) fire return interval. | Variables | Acronym | Dataset | | Fire | || | Number of Fires | num_fire | California Department of Forestry and Fire Protection: Fire Resource Assessment Program (1908-present) | | Fire Return Interval | FRI | | | Mean Time Since Last Fire Standard Deviation TSLF | mean_TSLF SD_TSLF | | | Vegetation | || | Forest Canopy Cover | CC | 2001 LANDFIRE database | | Forest Canopy Height | CH | | | Climate | || | Total Annual Precipitation | ppt | PRISM at 4km resolution, averaged over resurvey time periods per transect | | Mean Annual Temperature | tmean | | | Spatial | || | Elevation midpoint | elev | Grinnell Resurvey Project dataset | | Transect/Region | transect | | Model/Equation | p-value | adj. R 2 | AICc | | Modern Richness: All-Variables Model | ||| | 1. Richness = 12.76*** + 0.02*** CH + -0.08* FRI + 0.09** mean_TSLF + 2.37* RegionYO + 1.11 RegionSS + -0.11* sd_TSLF + -0.24** tmean | <0.001 | 0.61 | 212.7 | | 2. Richness = 8.39*** + 0.02*** CH + -0.09* FRI + 0.10** mean_TSLF + 0.001** elev + 2.50* RegionYO + 0.90 RegionSS + -0.11* sd_TSLF | <0.001 | 0.60 | +0.58 | | 3. Richness = 15.17*** + 0.02*** CH + -0.08* FRI + 0.06* mean_TSLF + -0.13* sd_TSLF + -0.30*** tmean | <0.001 | 0.57 | +1.14 | | 4. Richness = 11.17*** + 0.02*** CH + 1.69* RegionYO + 1.08 RegionSS + -0.18** tmean | <0.001 | 0.55 | +1.59 | | 5. Richness = 9.39*** + 0.02*** CH + 0.03 mean_TSLF + 2.75** RegionYO + 1.91* RegionSS + -0.14 tmean | <0.001 | 0.56 | +1.63 | | 6. Richness = 12.76*** + 0.02*** CH + -0.20** tmean | <0.001 | 0.52 | +1.67 | | Modern Richness: All-Variables PCA Model | ||| | 1. Richness = 12.73*** + 0.82*** PC1 + -1.00*** PC2 + -0.45 PC3 | <0.001 | 0.48 | 219.8 | | 2. Richness = 12.73*** + 0.82*** PC1 + -1.00*** PC2 | <0.001 | 0.46 | +0.15 | | 3. Richness = 12.73*** + 0.82*** PC1 + -1.00*** PC2 + -0.45 PC3 + -0.34 PC4 | <0.001 | 0.48 | +1.16 | | 4. Richness = 12.73*** + 0.82*** PC1 + -1.00*** PC2 + -0.34 PC4 | <0.001 | 0.46 | +1.26 | | Community Turnover: Fire Regime Variables Model | ||| | 1. Turnover = 0.08** + 0.005** num_fire + 0.002*** mean_TSLF | <0.01 | 0.30 | -83.4 | | 2. Turnover = 0.10* + 0.004* num_fire + -0.001 FRI + 0.003*** meanTSLF | <0.01 | 0.29 | +1.70 | Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License.

Authors Metrics & Citations Metrics Article Usage 301views 172downloads Citations Download citation Reina Warnert, Jessica Blois. Impacts of Fire Regime on Small Mammal Communities in California. Authorea. 05 June 2025. DOI: https://doi.org/10.22541/au.174913204.46170380/v1 DOI: https://doi.org/10.22541/au.174913204.46170380/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.