Data for: Effects of habitat management on rodent diversity, abundance, and virus infection dynamics

As anthropogenic factors continue to degrade natural areas, habitat management is needed to restore and maintain biodiversity. However, the impacts of different habitat management regimes on ecosystems have largely focused on vegetation analyses, with limited evaluation of downstream effects on wildlife. We compared the effects of grassland management regimes (prescribed burning, cutting/haying, or no active management) on rodent communities and the viruses they hosted. Rodents were trapped in 13 existing grassland sites in Northwest Arkansas, USA during 2020 and 2021. Rodent blood samples were screened for antibodies against three common rodent-borne virus groups: orthohantaviruses, arenaviruses, and orthopoxviruses. We captured 616 rodents across 5953 trap nights. Burned and unmanaged sites had similarly high abundance and diversity, but burned sites had a higher proportion of grassland species than unmanaged sites; cut sites had the highest proportion of grassland species but the lowest rodent abundance and diversity. A total of 38 rodents were seropositive for one of the three virus groups (34 orthohantavirus, three arenavirus, and one orthopoxvirus). Thirty-six seropositive individuals were found in burned sites, and two orthohantavirus-seropositive individuals were found in cut sites. Cotton rats and prairie voles, two grassland species, accounted for 97% of the rodents seropositive for orthohantavirus. Our study indicates that prescribed burns lead to a diverse and abundant community of grassland rodent species when compared to other management regimes; as keystone taxa, these results also have important implications for many other species in food webs. Higher prevalence of antibodies against rodent-borne viruses in burned prairies shows an unexpected consequence likely resulting from robust host population densities supported by the increased habitat quality of these sites. Ultimately, these results provide empirical evidence that can inform grassland restoration and ongoing management strategies.Funding provided by: National Science FoundationCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100008982Award Number: DEB 1911925Study sites Rodents were captured in grasslands throughout Benton and Washington Counties, Arkansas, USA. This area lies near the edge of the historical tallgrass prairie ecoregion, and like other tallgrass prairie ecosystems, most of the landscape has been altered by humans, with few remnant prairies remaining. Instead, many of the modern grasslands in this region are restored prairies or non-prairie grasslands. Trapping was conducted at 13 sites within six distinct grasslands (Figure 1). Neighboring sites within the same grassland were considered separate areas as they are distinctly managed and separated by physical barriers (i.e., roadway, riparian habitat, and/or firebreak). Although these barriers do not act as complete physical barriers, they limit rodent movement among sites and distinguish separately-managed parcels. Grasslands ranged in size from 6.7–32.6 ha, and distinct management sites within each grassland ranged in size from 1.5–23.6 ha. Site management was classified as one of three regimes: prescribed burning, reminiscent of natural ecosystem functioning (designated burn; five sites); haying, mowing, or other means of mechanical cutting, which result in managed yet artificial landscapes (designated cut; six sites); or no active management of vegetation, leading to heavy woody encroachment (designated unmanaged; 2 sites; Figure 1). Management regimes at the study sites have been continuous for several decades, and our results thus represent long-term effects of management regimes. Rodent trapping and sampling Rodent trapping was conducted once every two months at each site from June-November 2020 and April-July 2021. Because of the number of sites and distances between grasslands, sites were trapped in several groups over the course of the trapping month. To maintain consistency of time between trapping at each site, site groups were trapped in the same order each trapping month. For each trapping occasion, approximately 50 Sherman live traps (H. B. Sherman Traps, Inc.) were set for two consecutive nights approximately 10m apart in a series of transect lines (see Table 1 for deviations of trap numbers). Traps were baited with a mixture of millet and black oil sunflower seeds and set at dusk. Traps were checked and captured rodents were processed the following morning. Initially, all rodents were euthanized for tissue collection except for species classified as species of conservation need by Arkansas Game and Fish Commission (Reithrodontomys humulis, megalotis, and montanus). Due to permit limitations, we were unable to euthanize all individuals of abundant species (Reithrodontomoys fulvescens and Sigmodon hispidus) in fall 2020. Captured rodents were identified to species level based on morphology (pelage and lengths of ear, tail, head/body, and hind foot; Sealander & Heidt, 1990; Reid 2006). Visual inspection was used to determine sex and reproductive condition; males were considered to be reproductive if their testes were descended into the scrotum, and females were considered to be reproductive if their nipples were enlarged or lactating or if their vagina was perforate or plugged. Rodent blood samples were collected via either the submandibular vein directly into a microcentrifuge tube during processing and immediately placed on ice or a heart sample that was placed into phosphate-buffered saline (PBS) during dissection (see below; Forbes et al., 2014). To promote efficiency and minimize handling time and associated distress to wild rodents, most rodent species were quickly euthanized via cervical dislocation without anesthetic; cotton rats were the exception due to their larger size and were anesthetized with inhalation isoflurane prior to cervical dislocation. Euthanized rodents were placed in individual labeled grip-lock bags and stored in a cooler with ice while in the field. Rodents that were not euthanized were ear-tagged and released at their point of capture following sample and data collection. Euthanized rodents were stored in a -20°C freezer and later dissected under a biosafety hood. Tissue samples were collected aseptically using clean forceps and scissors and placed in sterilized microcentrifuge tubes. Hearts were placed in PBS solution to permit serology assays. All samples and specimens were stored at -20°C. All animal handling and sampling procedures were approved by the University of Arkansas Institutional Animal Care and Use Committee (IACUC) protocol number 20028 and Arkansas Game and Fish Commission permit numbers 102820194 and 030820211. Additionally, sampling at Chesney Prairie Natural Area was also approved by Arkansas Natural Heritage Commission permit numbers S-NHCC-19-025 and S-NHCC-21-007. Assays to detect antibodies against rodent viruses Blood samples were tested for antibodies reactive to orthohantaviruses, arenaviruses, and orthopoxviruses using immunofluorescence assays (IFAs), as previously described (Kallio-Kokko et al., 2006; Kinnunen et al., 2011; Forbes et al., 2014). Briefly, samples were diluted in PBS and then incubated on slides with viral antigens followed by several wash cycles to remove unbound antibodies. Fluorescent polyclonal rabbit anti-mouse FITC conjugate was then added to the slides, which were again incubated and washed. Slides were examined under a fluorescence microscope for reactive antibodies. These serology assays are cross-reactive within broad virus groups and therefore are effective and efficient approaches for non-specific screening (e.g., Ogola et al., 2021). Data analyses All analyses were conducted in R 4.1.0 (R Core Team 2021). We used a chi-square test of independence to compare trapping success among management types. Renyi diversity profiles were used to compare several indices of rodent diversity among management regimes (Tóthmérész 1995; Kindt 2020). Additionally, an analysis of similarity (ANOSIM) test was used to determine if rodent community composition varied among management regimes using the Bray-Curtis dissimilarity index (Herlemann et al., 2016; Zorz 2019). Because rodents with antibodies against the focus virus groups were only detected in sites that were burned or cut, a chi-square test of independence was used to test for differences in total seroprevalence of all three viruses among habitat management regimes. Binomial generalized linear mixed models (GLMMs) with seropositivity as the response variable were then used to compare seroprevalence between burned and cut sites, with grassland and site identity as a nested random effect. Demographic data, including sex, reproductive status, abundance index (capture success), and their interactions were set as explanatory variables in the GLMMs. Because most seropositive cases were from rodents with antibodies against orthohantaviruses, we also used GLMMs to compare orthohantavirus seroprevalence alone among burned and cut sites. Two separate binomial GLMMs were used to analyze orthohantavirus seroprevalence from all sites within cotton rats and prairie voles (M. ochrogaster), as these two species accounted for the majority of seropositive rodents but have different life histories, including seasonal dynamics and mass (Brady and Slade, 2001). Large, reproductive male rodents are often disproportionately seropositive for orthohantaviruses (Douglass, 2007; Polop et al., 2010), so explanatory variables for species-level GLMMs included mass, sex, reproductive status, abundance index, and their interactions. Finally, a Poisson GLMM, again using seropositivity as the response variable, was used to compare seroprevalence by trap success at each site and trapping occasion, with trapping occasion as a random effect and grassland and site identity as a nested random effect. GLMMs were conducted using the lme4 package (Bates et al., 2022); all other statistical analyses were conducted using base R (R Core Team, 2021). It is worth noting that although sites were grouped according to management regime, some differences in management schedules, site history, and biogeochemical factors were unavoidable and created heterogeneity within group categories. In particular, three of the five burned sites were burned every three years and the other two were burned annually. These differences were unavoidable due to the study design, akin to a natural experiment. However, potential differences due to site heterogeneity within management categories were assessed to validate groupings; no differences in rodent abundance, rodent diversity, or seroprevalence were detected between burn frequencies (Appendix 1). Despite several replicates of burned and cut sites, only two unmanaged sites were available in this study, as these habitats change drastically with the onset of management and are prone to ecological succession in the prolonged absence of management