Using sutures to attach miniature tracking tags to small bats for multimonth movement and behavioral studies

Determining the detailed movements of individual animals often requires them to carry tracking devices, but tracking broad-scale movement of small bats (<30 g) has been limited by transmitter technology and long-term attachment methods. This limitation inhibits our understanding of bat dispersal and migration, particularly in the context of emerging conservation issues such as fatalities at wind turbines and diseases. We tested a novel method of attaching lightweight global positioning system (GPS) tags and geolocating data loggers to small bats. We used monofilament, synthetic, absorbable sutures to secure GPS tags and data loggers to the skin of anesthetized big brown bats (Eptesicus fuscus) in Colorado and hoary bats (Lasiurus cinereus) in California. GPS tags and data loggers were sutured to 17 bats in this study. Three tagged bats were recaptured 7 months after initial deployment, with tags still attached; none of these bats showed ill effects from the tag. No severe injuries were apparent upon recapture of 6 additional bats that carried tags up to 26 days after attachment; however, one of the bats exhibited skin chafing. Use of absorbable sutures to affix small tracking devices seems to be a safe, effective method for studying movements of bats over multiple months, although additional testing is warranted. This new attachment method has the potential to quickly advance our understanding of small bats, particularly as more sophisticated miniature tracking devices (e.g., satellite tags) become available.     

Testing traditional assumptions about regional migration in bats

While some bats cover long distances during migration, moving thousands of kilometers, most migratory bats are considered regional migrants, thought to move relatively short distances (<?500 km) between hibernacula and maternity sites. However, behavior can vary considerably among species and our understanding of these movements has largely been limited to banding studies or detailed tracking of small numbers of bats by aircraft. Inferring population-wide behavior from small samples is difficult and can introduce bias. We tagged 108 Indiana bats (Myotis sodalis) in the Midwestern US and used a regional network of radiotelemetry receivers to study movement patterns. With this dataset, we tested the following traditional generalizations about regional migrants: (1) bats move away from hibernacula in spring in all directions with known maternity roosts, e.g., in a star-like pattern; (2) bats follow linear landscape features; (3) long-distance movements are uncommon; and (4) autumn migration comprises a single movement from summer maternity site to winter hibernaculum. In spring, bats left the hibernaculum immediately and primarily moved north despite available maternity roosts in all directions. We found no evidence that bats follow rivers, the predominant linear element in the landscape. Only six tagged bats traveled >?100 km, suggesting that longer-distance movements may be outliers. In autumn, only two bats visited multiple known hibernacula, and after swarming, some females moved >?100 km to areas without known hibernacula. Common generalizations about regional migrant movements may not be representative of population behavior and care should be taken with respect to management decisions based on those assumptions.     

Site fidelity of Svalbard polar bears revealed by mark-recapture positions

Quantifying the degree of site fidelity in polar bears’ (Ursus maritimus) use of denning and mating areas in spring is of considerable interest for both basic and applied purposes. We analyzed 276 spring-to-spring movements (displacements) of 178 polar bears obtained from mark-recapture in the period 1987–2010 in Svalbard, Norway. Male and female subadults and adults showed site fidelity (only subadult females non-significantly) when their movements were compared to a scenario of random movements between all capture locations. The median observed displacement was 47.6 km (bootstrapped 95 % CI: 38.6–57.8 km), significantly smaller than the median potential displacement for random movements of 206.3 km (bootstrapped 95 % CI: 187.3–219.6 km). Subadult females tended to have the longest displacements, followed by adult males and adult females. However, large individual variation both in displacement distances and recapture frequency tended to blur age and sex differences. Analysis restricted to one movement for each bear showed that the difference between adult males and adult females was small and non-significant. This indicates that findings based on telemetry, which is almost always restricted to females, may be relatively representative of the whole adult Barents Sea population in the spring season.     

Assessing noninvasive hair and fecal sampling for monitoring the distribution and abundance of river otter

Monitoring the distribution and abundance of populations is an important component of efforts to meet management or conservation goals. Although the objectives for such studies are easy to define, cost-effective, precise, and accurate estimates are often elusive. We tested the efficacy and compared the cost-effectiveness of methods for estimating the number and recording the distribution of river otter (Lontra canadensis). We genotyped otter hair sampled using two noninvasive instruments and compared those results with a hypothetical study design based on DNA extracted from fecal matter. Patterns of distribution generated from DNA collected at latrine sites were then compared to observations of otter collected using VHF radiotelemetry. We achieved a high probability of genotyping river otter with a small number of hairs (i.e., 59.0 % probability of producing a genotype with 1 guard hair and >5 under hair samples) collected using wire body snares and knaplock hair snags. Body snares were more effective at collecting otter hair, but there was relatively little additional cost to using both sampling instruments. Genotyped hair resulted in a high multi-year recapture rate (61.9 %). Hair collection and genotyping was the most cost-effective method for monitoring populations of river otter ($168.50 US/datum) followed by radiotelemetry ($264.50 US/datum), and the extraction of DNA from fecal matter ($266.00 US/datum). However, the noninvasive techniques did not represent the full distribution and fine-scale movements of otter, as observed using radiotelemetry. There has been much recent reporting of the efficacy of fecal matter as a source of DNA for conducting mark–recapture population estimates for mesocarnivores. Our data suggested that collecting DNA in hair may be a more cost-effective and efficient approach.     

Some like it cold: summer torpor by freetail bats in the Australian arid zone

Bats are among the most successful groups of Australian arid-zone mammals and, therefore, must cope with pronounced seasonal fluctuations in ambient temperature (T _a), food availability and unpredictable weather patterns. As knowledge about the energy conserving strategies in desert bats is scant, we used temperature-telemetry to quantify the thermal physiology of tree-roosting inland freetail bats (Mormopterus species 3, 8.5 g, n = 8) at Sturt National Park over two summers (2010–2012), when T _a was high and insects were relatively abundant. Torpor use and activity were affected by T _a. Bats remained normothermic on the warmest days; they employed one “morning” torpor bout on most days and typically exhibited two torpor bouts on the coolest days. Overall, animals employed torpor on 67.9 % of bat-days and torpor bout duration ranged from 0.5 to 39.3 h. At any given T _a, torpor bouts were longer in Mormopterus than in bats from temperate and subtropical habitats. Furthermore, unlike bats from other climatic regions that used only partial passive rewarming, Mormopterus aroused from torpor using either almost entirely passive (68.9 % of all arousals) or active rewarming (31.1 %). We provide the first quantitative data on torpor in a free-ranging arid-zone molossid during summer. They demonstrate that this desert bat uses torpor extensively in summer and often rewarms passively from torpor to maximise energy and water conservation.     

The third dimension of bat migration: evidence for elevational movements of Miniopterus natalensis along the slopes of Mount Kilimanjaro

Bats are important ecosystem service providers, and therefore most relevant for both lowland and highland habitats, particularly in the tropics. Yet, it is poorly understood to what extent they perform large-scale movements, especially movements along mountain slopes. Here, we studied the movement ecology of the potentially migratory species Miniopterus natalensis at Mount Kilimanjaro in Tanzania. We analysed stable isotope ratios of C (δ¹³C), N (δ¹⁵N) and H (δ²H) in keratin of sedentary frugivorous and insectivorous bats captured between 800 and 2,400 m above sea level to establish elevational gradients of stable isotope ratios in consumer tissues. We expected correlations between stable isotope ratios of the non-exchangeable portion of H in fur keratin and the elevation of capture site, but not necessarily for δ¹³C and δ¹⁵N. Yet, in bats of both feeding ensembles, we found δ¹⁵N of fur keratin to correlate positively with the elevation of capture sites but not δ²H. In frugivorous bats, δ¹³C increased with increasing capture elevation as well. By looking at intra-individual variation of δ¹³C and δ¹⁵N in fur keratin and wing membrane tissues of sedentary Rhinolophus cf. clivosus and of the potentially migratory species M. natalensis, we gathered evidence that M. natalensis migrates seasonally between low and high elevations along the slopes of Mount Kilimanjaro. Finally, based on an isoscape origin model we estimated that M. natalensis captured before and after the cold period at around 1,800 m above sea level originated from around 1,400 m a.s.l. or lower. Thus, we received convergent results in support of seasonal elevational movements of M. natalensis, probably in search for cold hibernacula at higher elevations of Mount Kilimanjaro.     

Deeply torpid bats can change position without elevation of body temperature

Because body temperature is tightly coupled to physiological function, hibernating animals entering deep torpor are typically immobile. We analysed thermal behaviour and locomotory activity of hibernating greater mouse-eared bats Myotis myotis and found two types of movement behaviour related to body temperature, i.e. movement at high fur temperature and at low fur temperatures (Tflow; <5 °C). First Tflow movements appeared at the beginning of March and often occurred during long torpor bouts. In most cases, Tflow events represented slow displacements between clusters of bats. In several cases, however, departure or arrivals from and into clusters was also recorded without any elevation in body temperature. Distance travelled, flight duration and speed of locomotion during Tflow events was lower than in high fur temperature events. Such behaviour could allow bats to save energy long-term and prolong torpor bouts. Tflow movement in torpid bats significantly changes our understanding of basic hibernation principles and we strongly recommend further studies on the subject.     

The trans-boundary importance of artificial bat hibernacula in managed European forests

Many European migratory bat species hibernate in large hollow trees, a decreasing resource in present day silviculture. Here, we report on the importance of man-made hibernacula to support trans-boundary populations of noctule bats (Nyctalus noctula), a species that performs seasonal long distance movements throughout Europe. In winter, we surveyed nine bat roosts (eight artificial and one natural) in Germany and collected small tufts of fur from a total of 608 individuals. We then measured the stable isotope ratios of the non-exchangeable hydrogen in fur keratin and estimated the origin of migrants using a refined isoscape origin model that included information on expected flight distances and migration directions. According to the stable isotope signature, 78 % of hibernating bats originated from local populations. The remaining 22 % of hibernacula occupants originated from distant populations, mostly from places in northern or eastern countries such as Sweden, Poland and Baltic countries. Our results confirm that many noctule bats cross one or several political borders during migration. Data on the breeding origin of hibernating noctule bats also suggest that artificial roosts may not only be important for local but also for distant populations. Protection of natural and artificial hibernacula in managed forests may support the trans-boundary populations of migratory bats when hollow trees are scarce in managed forests.     

Postnatal development in the big-footed bat,Myotis macrodactylus: wing morphology,echolocation calls,and flight

We studied the postnatal development of wing morphology and echolocation calls during flight in a free-ranging population of the big-footed bat, Myotis macrodactylus, using the mark-recapture methodology. Young bats were reluctant to move until 7 days of age and started fluttering at a mean age of 10 days. The wingspan and wing area of pups followed a linear pattern of growth until 22 days of age, by which time the young bats exhibited flapping flight, with mean growth rates of 0.62 mm/day and 3.15 mm²/day, for wingspan and area, respectively, after which growth rates decreased. Pups achieved sustained flight at 40 days of age. Of the three nonlinear growth models (logistic, Gompertz, and von Bertalanffy), the logistic equation provided the best fit to the empirical curves for wingspan and wing area. Neonates emitted long echolocation calls with multiple harmonics. The duration of calls decreased significantly between flutter (19 days) and flight (22 days) stages. The peak and start frequency of calls increased significantly over the 3-week period of development, but the terminal frequency did not change significantly over the development period.     

Bat species richness in live fences and in corridors of residual rain forest vegetation at Los Tuxtlas, Mexico

Fragmentation of lowland tropical rain forests has resulted in loss of animal and plant species and isolation of remaining populations that puts them at risk. At Los Tuxtlas. Mexico, lowland rain forests are particularly diverse in the bat fauna they contain and while most of the forests have been fragmented by human activity, many of the fragments still harbor diverse assemblages of bat species. To assess the effectiveness of corridors, among other options, to ameliorate the negative effects of fragmentation, we investigated bat species richness and relative abundance in one 6 km long section of live fences (LF) bordering a dirt road and in three 6 km long sections of residual forest vegetation along the sides of three permanent streams (BS. MS. HS). Netting of bats resulted in the capture of 967 bats. At the LF site we captured 12 bat species. 15 at the BS site. 18 at the MS site and 23 at the HS site. Species richness was associated with average area of forest fragments within a 1000 m band on each side of each corridor (r = 0.97. p = 0.01). Only 28% of the species were common among sites. Frugivorous and insectivorous species accounted for 48% each of bat captures while nectarivores accounted for 3%, sanguinivores for 0.5% and carnivore-frugivores for 0.5% Edge habitat species such as Pteronotus parnelli and Sturnira lilium accounted for 50% of the captiures. Frugivorous species such as Carollia brevicauda. Vampyrodes caraccioli. Dermanura phaeotis, D. toltecus and A. jamaicensis accounted for another 25% of bat captures. Recaptures of bats indicated bat movements from forest fragments to corridors and between corridors, with recapture distances ranging from 200 to 2000 m. Within corridor recaptures separated by several months from the original recapture date indicated individual bat revisitation to these sites. We discuss the value of these corridors to bats as stepping stones in the fragmented landscape.     

Knowing the way home: strong philopatry of a highly mobile insect species, Brenthis ino

Mobility is crucial for the maintenance of viable metapopulations, but quantitative data to evaluate risks due to insufficient individual mobility of focal insect species are mostly lacking. We selected the butterfly Brenthis ino, a species typically confined to wet fallow grasslands in Central Europe and performed a mark–release–recapture study in a 3.2 ha study area with one big and one small patch of suitable habitat from 22 June to 23 July 2010. The position of each butterfly capture was measured with a GPS and transferred into a GIS. In total, we marked 984 individuals in 1,545 capture events and estimated that the cumulative population size was 2,400 individuals. The initial increase of adult males proceeded much faster than for females, similar to the protandrous population build-up known from other butterflies. Moved distances for both sexes usually did not exceed 80 m, and about 40 % of all individuals used less than 2 % of the available suitable habitat. All individuals switching to the other patch returned later to their patch of origin, confirming that B. ino is highly philopatric. We conclude that low effective mobility in B. ino produces much smaller home ranges than suggested by merely observing flight activities in the field, and that low tendencies towards long-distance movements significantly hamper the maintenance of metapopulations when patch density decreases due to landscape fragmentation.     

Non-lethal method of DNA sampling in euglossine bees supported by mark–recapture experiments and microsatellite genotyping

Non-lethal sampling methods are of great interest for conservation genetic studies to prevent the death of individuals in populations that are threatened or in decline. With this aim, we tested a non-lethal method of partial antennae removal for DNA sampling in two euglossine bee species: Euglossa cordata and Eulaema nigrita. We validated the survival of the individuals through mark–recapture experiments during 16 months. The quality and quantity of the tissue for DNA analysis was verified through amplification and genotyping of nine and eleven microsatellite loci, respectively. Our results from the mark–recapture experiments showed equal recapture rates of individuals with intact and removed antennae (E. cordata χ² = 2.492, df = 1, p = 0.114; E. nigrita χ² = 1.683, df = 1, p = 0.194). Microsatellite loci were successfully genotyped in 97.1 and 97.6 % of the E. cordata and E. nigrita individuals, respectively. Our results validate the feasibility of using antennae tissue for DNA genetic analysis without compromising the survival of individual bees.     

Summer heterothermy in Rafinesque’s big-eared bats (Corynorhinus rafinesquii) roosting in tree cavities in bottomland hardwood forests

Many small mammals are heterothermic endotherms capable of maintaining an elevated core body temperature or reducing their thermoregulatory set point to enter a state of torpor. Torpor can confer substantial energy savings, but also incurs ecological costs, such as hindering allocation of energy towards reproduction. We placed temperature-sensitive radio transmitters on 44 adult Rafinesque’s big-eared bats (Corynorhinus rafinesquii) and deployed microclimate dataloggers inside 34 day roosts to compare the use of torpor by different sex and reproductive classes of bats during the summer. We collected 324 bat-days of skin-temperature data from 36 females and 4 males. Reproductive females employed fewer torpor bouts per day than non-reproductive females and males (P < 0.0001), and pregnant and lactating females had higher average (P < 0.0001) and minimum (P < 0.0001) skin temperatures than non-reproductive females. Pregnant females spent less time torpid (P < 0.0001) than non-reproductive females, but lactating females used relatively deep, long torpor bouts. Microclimates varied inside tree species with different configurations of entrances to the roost cavity (P < 0.0001). Bats spent more time torpid when roosting in water tupelo (Nyssa aquatica) trees possessing only a basal entrance to the cavity (P = 0.001). Of the tree species used as roosts, water tupelo cavities exhibited the least variable daytime and nighttime temperatures. These data demonstrate that use of summer torpor is not uniform among sex and reproductive classes in Rafinesque’s big-eared bat, and variation in microclimate among tree roosts due to species and structural characteristics facilitates the use of different thermoregulatory strategies in these bats.     

Indeterminate growth in long-lived freshwater turtles as a component of individual fitness

Although evidence that reptiles exhibit indeterminate growth remains equivocal and based on inadequate data, the assumption that they do is still widely accepted as a general trait of reptiles. We examined patterns of variation in adult growth using long-term mark-recapture data on 13 populations of 9 species representing 3 families of freshwater turtles located in South Carolina, Michigan, and Arizona in the USA and in Ontario, Canada. Across 13 study populations, growth rates of all adults and only those that grew averaged 1.5 and 1.9 mm/yr respectively. Sources of variation in growth rates included species, population, sex, age, and latitude. Most adults of both sexes with recapture intervals greater than 10 years grew, but across all populations an average of 19 % of individuals did not grow (some with recapture intervals up to 30 years). For known-age adults of three species, the highest growth rates occurred during the 10 years following sexual maturity, and the proportions of non-growing individuals increased with age. Growth rates of adults were on average 92 % lower than those of juveniles. Based on linear relationships of clutch size and body size of females at average juvenile and adult growth rates it would take 0.7 (0.2–1.2) years and 8.6 (min–max = 2.3–18.5) years, respectively, to grow enough to increase clutch size by one egg. The majority of within population variation in adult body size in 3 species appeared to be a combination of differences in ages at maturity and juvenile and early adult growth, rather than indeterminate growth. The results from our study populations indicate that increases in body size (and associated reproductive output) that results from indeterminate growth are not substantial enough to represent a major factor in the evolution of life histories in general or the evolution of longevity and aging specifically.     

Mark–recapture may reveal more about ecology than about population trends: Demography of a threatened ghost bat (Macroderma gigas) population

Since European settlement in Australia, the geographical range of ghost bats (Macroderma gigas) has contracted northwards. Ghost bats are thought to occur in disjunct populations with little interpopulation migration, raising concerns over the current status and future viability of the southernmost colony, which has also been threatened by mining activity. To address these concerns, demographic parameters of the southernmost colony were estimated from a mark–recapture study conducted during 1975–1981. Female bats gave birth to a single young in late spring, but only 40% (22–70%, 95% CI) of females bred in their second year, increasing to 93% (87–97%, 95% CI) for females ≥ 2 years old. Sixty‐five percent of juveniles caught were female. Annual adult survival ranged between 0.57–0.77 for females and 0.43–0.66 for males, and was lowest over winter–spring and greatest in autumn–winter. Juvenile survival for the first year ranged between 0.35–0.46 for females and 0.29–0.42 for males. Adult survival varied among seasons, was negatively associated with rainfall, but was not associated with temperature beyond being lower in late winter. Poor survival may result from the inferior daytime roosts that bats must use if water seepage forces them to leave their normal roosts. Although these age‐specific rates of fecundity and survival suggested a declining population, mark–recapture estimates of the population trend indicated stability over the study period. Counts at daytime roosts also suggested a population decline, but were considered unreliable because of an increasing tendency of bats to avoid detection. It is therefore likely that some assumptions in estimating survival were violated. These results provide a caution against the uncritical use of population projections derived from mark–recapture estimates of demographic parameters, and the use of untested indices as the basis for conservation decisions.     

Non-kin cooperation in bats

Many bats are extremely social. In some cases, individuals remain together for years or even decades and engage in mutually beneficial behaviours among non-related individuals. Here, we summarize ways in which unrelated bats cooperate while roosting, foraging, feeding or caring for offspring. For each situation, we ask if cooperation involves an investment, and if so, what mechanisms might ensure a return. While some cooperative outcomes are likely a by-product of selfish behaviour as they are in many other vertebrates, we explain how cooperative investments can occur in several situations and are particularly evident in food sharing among common vampire bats (Desmodus rotundus) and alloparental care by greater spear-nosed bats (Phyllostomus hastatus). Fieldwork and experiments on vampire bats indicate that sharing blood with non-kin expands the number of possible donors beyond kin and promotes reciprocal help by strengthening long-term social bonds. Similarly, more than 25 years of recapture data and field observations of greater spear-nosed bats reveal multiple cooperative investments occurring within stable groups of non-kin. These studies illustrate how bats can serve as models for understanding how cooperation is regulated in social vertebrates.     

Mortality of bats at wind turbines links to nocturnal insect migration?

This note is based on a literature search and a recent review of bat mortality data from wind farms in Europe (published elsewhere). We suggest that mortality of bats at wind turbines may be linked to high-altitude feeding on migrating insects that accumulate at the turbine towers. Modern wind turbines seem to reach high enough into the airspace to interfere with the migratory movements of insects. The hypothesis is consistent with recent observations of bats at wind turbines. It is supported by the observation that mortality of bats at wind turbines is highly seasonal (August–September) and typically peaks during nights with weather conditions known to trigger large-scale migratory movements of insects (and songbirds). We also discuss other current hypotheses concerning the mortality of bats at wind turbines.     

Estimating density of a forest-dwelling bat: a predictive model for Rafinesque’s big-eared bat

Data on species distribution and abundance are the foundation of population ecology. However, due to difficulties in surveying bats, abundance estimates for tree-roosting microchiropterans are non-existent. Therefore, our objective was to develop methods for estimating colony abundance and density, taking as our model Rafinesque’s big-eared bat Corynorhinus rafinesquii, a species of conservation concern found in cypress-gum swamps of the southeastern United States. We searched 123 transects at eight study sites in the Coastal Plain of Georgia, USA to locate and characterize diurnal summer roosts of C. rafinesquii. We modeled the relationship between the number of bat colonies and landscape-scale habitat variables with zero-inflated negative binomial regression and used Akaike’s information criterion to select the most parsimonious models. We generated a predictive density map to identify areas of high colony density and to estimate overall abundance. Colony density was predicted by the duration of wetland flooding, wetland width, and study site. Application of the regression model to a GIS indicated there were 3,734 colonies containing 6,910 adult bats on the eight study sites. Predicted density ranged from 0.07 colonies/ha and 0.07 adult bats/ha in saturated wetlands to 0.47 colonies/ha and 1.18 adult bats/ha in semi-permanently flooded wetlands. This study is the first to estimate density and abundance of forest-dwelling microchiropterans over a large area. Such data can serve as a baseline for future work on population trends in C. rafinesquii. In addition, our approach could be replicated for other bat species with moderately cryptic roosts.     

Bat diversity of a Cerrado habitat in central Brazil

The Cerrado is considered one of the 25 biodiversity hotspots in the world by conservation organizations. There are few studies on bat taxonomic groups for this biome. Herein we present a bat survey employing mist nets in the protected area of a private natural heritage reserve Reserva Particular do Patrimônio Natural Pousada das Araras, located in the west-central Brazil. We investigate the hypothesis that the Cerrado habitat complexity plays a role on the different structuring forces in bat ecological communities. Bats represent a diversity of trophic levels, and they occupy a wide range of available Cerrado habitats and microhabitats. The patterns and processes we discuss represent the factors influencing coexisting species of bats in different habitats and their implications for conservation. We captured 758 individuals of 25 species belonging to four families. Phyllostomidae was the dominant family, represented by 20 species (80%). The average recapture rate was 6.2%, and the species with most proportional recapture was A. caudifer. Greater species richness was observed among bats with predominantly insectivore habits, followed by frugivores. Glossophaga soricina was a dominant species, with about 30% of the captures. There was seasonal variation relating to the number of bats captured, with greater bat frequency occurring during the wet season, although some species occurred at higher rates during the dry season, such as C. perspicillata and S. lilium. Sanguivore bats were abundant, reflecting the availability of shelters and food supply (livestock) in the surrounding area. Compared to other surveyed areas, Pousada das Araras may be considered of high species diversity, supplanting the majority of known Cerrado studied areas. Apparently Cerrado favours the occurrence of insectivore species, with emphasis on the foliage-gleaning insectivores belong to the subfamily Phyllostominae. This study indicates that apparently the conservation of the Cerrado savanna fragments can support a considerable diversity of bat species.     

Hendra Virus Spillover is a Bimodal System Driven by Climatic Factors

Understanding environmental factors driving spatiotemporal patterns of disease can improve risk mitigation strategies. Hendra virus (HeV), discovered in Australia in 1994, spills over from bats (Pteropus sp.) to horses and thence to humans. Below latitude ? 22°, almost all spillover events to horses occur during winter, and above this latitude spillover is aseasonal. We generated a statistical model of environmental drivers of HeV spillover per month. The model reproduced the spatiotemporal pattern of spillover risk between 1994 and 2015. The model was generated with an ensemble of methods for presence–absence data (boosted regression trees, random forests and logistic regression). Presences were the locations of horse cases, and absences per spatial unit (2.7 × 2.7 km pixels without spillover) were sampled with the horse census of Queensland and New South Wales. The most influential factors indicate that spillover is associated with both cold-dry and wet conditions. Bimodal responses to several variables suggest spillover involves two systems: one above and one below a latitudinal area close to ? 22°. Northern spillovers are associated with cold-dry and wet conditions, and southern with cold-dry conditions. Biologically, these patterns could be driven by immune or behavioural changes in response to food shortage in bats and horse husbandry. Future research should look for differences in these traits between seasons in the two latitudinal regions. Based on the predicted risk patterns by latitude, we recommend enhanced preventive management for horses from March to November below latitude 22° south.