Feral cat home-range size varies predictably with landscape productivity and population density

Transcription

1 Journal of Zoology. Print ISSN Feral cat home-range size varies predictably with landscape productivity and population density A. J. Bengsen 1, D. Algar, G. Ballard 3,, T. Buckmaster 5, S. Comer 6, P. J. S. Fleming 1,, J. A. Friend 7, M. Johnston, H. McGregor 8, K. Moseby 9 & F. Zewe 1 Vertebrate Pest Research Unit, Biosecurity NSW, NSW Department of Primary Industries, Orange, NSW, Australia Science and Conservation Division, Department of Parks and Wildlife, Woodvale, WA, Australia 3 Vertebrate Pest Research Unit, Biosecurity NSW, NSW Department of Primary Industries, Armidale, NSW, Australia School of Environmental and Rural Sciences, University of New England, Armidale, NSW, Australia 5 Invasive Animals Cooperative Research Centre and Institute for Applied Ecology, University of Canberra, Canberra, ACT, Australia 6 South Coast Region, Department of Parks and Wildlife, Albany, WA, Australia 7 Science and Conservation Division, Department of Parks and Wildlife, Albany, WA, Australia 8 Mornington Wildlife Sanctuary, Australian Wildlife Conservancy, Derby, WA, Australia 9 Ecology and Environmental Science, University of Adelaide, Adelaide, SA, Australia Keywords environmental productivity; feral cats; fpar; home-range size; landscape productivity; predator control; polygyny; solitary carnivore. Correspodence Andrew J. Bengsen, Vertebrate Pest Research Unit, Biosecurity NSW, NSW Department of Primary Industries, Locked bag 66, Orange, NSW 8, Australia. Tel: ; Fax: andrew.bengsen@dpi.nsw.gov.au Editor: Andrew Kitchener Received 3 November 1; revised 13 July 15; accepted 9 July 15 doi:1.1111/jzo.19 Abstract An understanding of the factors that drive inter-population variability in home-range size is essential for managing the impacts of invasive species with broad global distributions, such as the feral domestic cat (Felis catus). The assumption that home-range sizes scale negatively with landscape productivity is fundamental to many spatial behaviour models, and inter-site variation in landscape productivity has often been invoked to explain the vast differences in feral cat home-range sizes among different regions. However, the validity of this explanation has not been tested or described. We used regression models to examine the ability of remotely sensed landscape productivity data, average body weight and population density to explain differences in the size of feral cat home ranges estimated across a diverse collection of sites across the globe. As expected for a solitary polygynous carnivore, female cats occupied smaller home ranges in highly productive sites, and range sizes of male cats scaled positively with those of females. However, the relationship between range size and productivity broke down at highly seasonal sites. Home-range size also scaled negatively with population density, but there was no clear relationship with average body weight. The relationships we describe should be useful for predicting home-range sizes and for designing effective feral cat control and monitoring programmes in many situations. More generally, these results confirm the importance of landscape productivity in shaping the spatial distribution of solitary carnivores, but the nature of the relationship is more complicated than is often appreciated. Introduction Effective management of invasive animal populations requires a sound understanding of the ways that animals use space and landscape. Information about traits such as habitat selection, movement rates, excursive movements and the size and form of home ranges can be important determinants of the scale, location and intensity at which damage mitigation or monitoring operations should be conducted (Goltz et al., 8; Moseby, Stott & Crisp, 9; Bengsen, Butler & Masters, 11, 1). Reliable information on spatial behaviour is also critical to understanding many of the ecological functions or impacts of pest populations, through predator prey interactions (Alterio, Moller & Ratz, 1998), interspecific competition (Wilson et al., 1) and disease transmission (Sparkes et al., 1). Feral domestic cats (Felis catus) are an important invasive predator of native fauna in many parts of the world (e.g. Medina et al., 11; Abbott et al. 1; Oppel et al. 1). As expected for an obligate carnivore distributed over a broad range of biomes, feral cats show high interpopulation variability in movement patterns and in the size of geographic areas that they regularly use (i.e. their home ranges, sensu Burt, 193; Moseby et al., 9; Doherty, Bengsen & Davis, 1). Consequently, management programmes to mitigate their impacts must be customized to 11 Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London

2 A. J. Bengsen et al. Feral cat home-range size and landscape productivity suit specific local conditions (Dickman, Denny & Buckmaster, 1; Doherty et al., 1). However, the high variability in home-range size among populations makes it difficult to predict optimal spatial scales and intensities for population control programmes, and direct estimation of feral cat home-range size is beyond the financial and logistical capacity of many management programmes. An understanding of the mechanisms underlying variation in feral cat home-range size is therefore needed to provide a foundation for understanding and managing their impacts across the globe. Feral cats that are not dependent on anthropogenic resources are generally solitary and exhibit a polygynous mating system, in which males seek access to multiple females (Liberg et al., ). Female home-range sizes should be determined by the availability of prey as well as the distribution of other females (Liberg et al., ; Powell, ). Highly productive sites should provide a greater range and abundance of lower order consumers (i.e. prey) than less productive sites (McNaughton et al., 1989; Ferguson, Currit & Weckerly, 9), thereby allowing female cats to meet their food requirements within smaller home ranges (McLoughlin & Ferguson, ). However, home ranges should be larger where productivity is clustered or highly seasonal (Konecny, 1987; Powell, ), so relationships between range size and productivity may be mediated by spatial or temporal heterogeneity. Highly productive sites should also support greater population densities (Liberg et al., ). Average homerange size of female cats should, therefore, scale negatively with the magnitude and consistency of landscape productivity, and also with population density. Negative relationships between productivity and homerange size have been observed in several carnivores (Herfindal et al., 5; Nilsen, Herfindal & Linnell, 5; Ferguson et al., 9). However, the form of these relationships has been inconsistent among species (Nilsen et al., 5). Although environmental productivity has often been invoked to explain inter-population variability in feral cat home-range size (e.g. Moseby et al., 9; Buckmaster, 11), its importance has not been evaluated and the form of any relationship between range size and productivity has not been described. In a polygynous mating system, adult males are expected to structure their space use to maximize mating opportunities with multiple females (Sandell, 1989). Average adult male home-range sizes should, therefore, be greater than female home-range sizes (Liberg et al., ; Powell, ) and greater than required for metabolic requirements alone (Sandell, 1989). Hence, average adult male home-range size should relate indirectly to landscape productivity through its effect on female home-range size. This study combines data on feral cat home-range sizes from published and unpublished sources with remotely sensed landscape productivity data to estimate relationships between home-range size and productivity among feral cat populations across the globe. Specifically, we test the following expectations: (1) female home-range size scales negatively with indices of landscape productivity; () the effects of productivity on female range size are mediated by spatial and temporal variation in productivity; (3) for any given body weight, male home range is consistently larger than female home range; () previously established relationships between population density and home-range size will persist when challenged with data from more recent studies. The resulting information should help conservation managers and researchers to make initial predictions about home-range sizes that can be used to inform the design of control and monitoring programmes to mitigate the damage caused by feral cats. It will also further our understanding of the importance of landscape productivity and population density in the spatial organization of cats and other carnivores. Materials and methods Ecological data collection We collected data from 1 studies that reported average feral cat home-range sizes from 7 sites or situations across the globe (Fig. 1 and Supporting Information Table S1). We focussed on feral cats: wild-living cats in self-sustaining populations independent of humans. We excluded studies in which cats were tracked for < 15 days because cats may not have had time to reveal the full extent of their movements over a time scale relevant to typical feral cat control operations. Most average home-range sizes were estimated using 95% or 1% minimum convex polygon estimates (MCP95, n = 1; MCP1, n = 16); the remainder used kernel-based (n = 7), modified minimum (n = ) or homerange length methods (n = 1). Some studies reported homerange estimates from different seasons (summer/autumn/ winter/spring or breeding/non-breeding) within the same sites, in which case we treated each set of seasonal estimates as a single case. Most studies were from the Australasian region or large islands. We collated mean homerange size estimates and body weight estimates for each sex from studies that provided body weight data, and population density estimates from 7 studies. Landscape productivity Landscape productivity can be indexed using several remotely sensed data sources that describe vegetation biomass and function. We used the fraction of photosynthetically active radiation absorbed by vegetation (fpar) to characterize landscape productivity at each study site in which feral cat movement data were collected after. fpar provides a robust indicator of productivity (Coops et al., 8) and consistency with home-range studies in other carnivores (Herfindal et al., 5; Nilsen et al., 5). Data were collected at 1 km resolution by NASA s MODIS sensor, which began operating in. For each study that provided data on feral cat home ranges after the year, we downloaded local monthly fpar data for the last 1 months of the study period, using the MODIS Reprojection Tool Web Interface (USGS, Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London 113

3 Feral cat home-range size and landscape productivity A. J. Bengsen et al. Figure 1 Locations of 7 study situations that provided data on feral cat home-range sizes including the situations (triangles) at which remotely sensed data on environmental productivity (fpar) were used to estimate relationships between range size and environmental productivity (some sites obscured by overlap). Global feral cat distribution (grey shading) adapted from Long (3). 1). We excluded studies that were conducted around human settlements or on small islands (<5 km ) because cats were likely to receive resource subsidies from people or from marine environments through predation on seabirds. We clipped each dataset to a 1 km buffer around the centre of each study site and removed any cells that were covered by water, providing a maximum 31 km area. We then calculated five different indices to characterize different aspects of productivity at each site: (1) fpar min : an index of average productivity during the least productive month calculated by identifying the least productive month for each site (i.e. the lowest mean fpar value, estimated across all cells) and taking the mean fpar value for that month. () fpar mean : an index of average annual productivity for each site obtained by creating a new data layer comprising the mean fpar value for each cell estimated over 1 months, and then taking the grand mean. (3) CV month : an index of temporal variation in productivity for each site by creating a new data layer comprising the coefficient of variation (CV) for each cell estimated over 1 months, and then taking the mean across all cells. () CV min : an index of spatial variation in productivity obtained by calculating the CV of fpar values across all cells for the fpar min layer, and (5) CV mean : an index of spatial variation in productivity obtained by calculating the CV of the fpar mean layer. Although we indexed productivity at each site during the time that cats were tracked, an individual s home range is likely to be shaped by earlier experiences and conditions and there will be a time lag before any changes in productivity induce changes in cat behaviour. Indices are therefore intended only to characterize differences in the magnitude and distribution of landscape productivity among sites at temporal and spatial scales relevant to the system. Determinants of female home-range size We used seven linear regression models, weighted by sample size, to estimate relationships between female homerange size (log-transformed for normality) and fpar min and fpar mean across study situations for which fpar data were available. Explanatory variables were centred by subtracting the mean. A negative relationship between fpar min and home-range size would imply that female cats structured their space use according to resource availability during the least productive time of year, whereas a negative relationship with fpar mean would indicate that an estimate of total annual productivity would be a better predictor. We also included two-way interactions between fpar min and CV month or CV min, and interactions between fpar mean and CV month or CV mean to evaluate the hypothesis that relationships between productivity and home-range size are mediated by temporal or spatial variability in productivity. We calculated the marginal effect of CV month or CV mean on fpar min and fpar mean to describe any such mediating effect. Collinearity was considered problematic for any explanatory variable with a variance inflation factor > 3 (Zuur et al., 9). We used second-order Akaike s information criterion (AICc) to estimate the relative strength of support for each model. AICc provides a greater penalty for extra parameters than AIC to reduce the risk of selecting over-fitted models with small sample sizes (Burnham & Anderson, ). Parameters were estimated from the best supported model, as the different 11 Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London

4 A. J. Bengsen et al. Feral cat home-range size and landscape productivity models each had definite and different interpretations regarding relationships among the variables (Burnham & Anderson, ). Male and female range size We used linear regression to estimate the relationship between average male and female home-range size across 3 study situations that provided >1 home-range estimate for both sexes. Both variables were log-transformed for normality and to provide a linear relationship. We used data from situations that provided body weight estimates for both sexes to estimate relationships between average body weight and range size for male and female cats. Log-transformed range size was specified as the response variable in five nested linear mixed effects models ranging from a global model with an interaction between sex and average body weight to a null model. Site was specified as a random effect because home ranges of both sexes were estimated within sites. Support for each model was assessed using AICc. We used marginal and conditional R to assess the explanatory power of the best supported model (Nakagawa & Schielzeth, 13). Home-range size and population density Liberg et al. () found that range size for female and male feral cats scaled negatively with population density, suggesting that home-range sizes might be roughly predicted from estimated population densities. We remodelled the data collated in Liberg et al. () to estimate precision and 95% prediction intervals. To assess the predictive ability of our remodelling of the original data, we calculated mean squared prediction error (MSPE) using 17 new cases (8 females, 9 males; details provided in Supporting Information Table S1) with data on home-range sizes and population densities from nine datasets that were not included in Liberg et al. s () study. We used a paired sample t-test to test whether differences between predicted and observed home-range sizes in the new sample were different from zero at a =.5. All analyses were conducted in the R statistical environment (v 3.1.1, R Core Team 1). Results Mean female home-range size across the cases for which useful fpar data were available ranged from 1.16 to 3. km (median =.36 km ). Mean and minimum fpar indices showed substantial variation among sites (fpar mean mean = , fpar min mean =.37.97), as did seasonality (CV month mean = ). Most sites characterized by high seasonality also had below average productivity (fpar mean and fpar min ), whereas sites characterized by low seasonality occurred across the range of productivity indexed in this study. All studies that used MCP1 estimates occurred at highly seasonal sites, but Table 1 Model selection criteria for seven models explaining the effects of environmental productivity (fpar min, fpar mean ) on female feral cat home-range size in study situations Model terms b 1, b, b 3 the spread of the remaining data suggested that this was more likely to be a source of noise than bias. Determinants of female home-range size Variation in female home-range size was best described by the model in which fpar mean varied conditionally with the temporal variation in productivity index, CV month. This was the only model in the set with substantial support (D AICc <, Table 1). Variation in female range size was described by the equation: lnðfemale range size½km ŠÞ ¼ a þ b 1 fpar mean þ b CV month þ b 3 ðfpar mean CV month Þ (1) where a (intercept) = , b 1 =.1.9, b =..1 and b 3 =.., and fpar mean and CV month were centred by subtracting means of 35.5 and 37.95, respectively (adjusted R =.51). The marginal effect of fpar mean on log-transformed female range size was described by the equation: marginal effect ¼ðb 1 þ b 3 CV month ÞfPAR mean () Thus, female range size decreased sharply with increasing average annual productivity (fpar mean ) at sites where there was little seasonal variation in productivity (low CV month ), moderately at sites with average seasonal variation (average CV month ), and tended to increase at sites with high seasonal variation (high CV month ; Fig. a, c). The 95% confidence intervals of the marginal effect of fpar mean on female range size excluded zero at centred CV month values < 5 and > (Fig. d), corresponding with raw values of.3 (5% of studies) and.6 (17% of studies), respectively. Male and female home-range size K AICc D i x i Log(L) fpar mean,cv month, fpar mean 9 CV month fpar mean,cv mean, fpar mean 9 CV mean fpar min,cv month, fpar min 9 CV month Null fpar mean fpar min fpar min,cv min, fpar min 9 CV min All variables are log-transformed. K, number of parameters; AICc, second-order Akaike s information criterion; D i, difference in AIC between model i and the most supported model; x i, Akaike weight: weight of evidence relative to the other models in the set; Log(L), log-likelihood. Median female home-range size at those sites that provided data for both sexes was.6 km [interquartile range Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London 115

5 Feral cat home-range size and landscape productivity A. J. Bengsen et al. a low seasonality b average seasonality ln (female range size [km ]) 3 1 ln (female range size [km ]) 3 1 ln (female range size [km ]) 3 1 Mean productivity index (fpar mean ) c high seasonality Mean productivity index (fpar mean ) Marginal effect of fpar mean ± 95% CI MCP95 MCP1 KDE95 Mean productivity index (fpar mean ) d Seasonality index (CV month ) Figure Relationships between mean annual productivity (fpar mean index) and female feral cat home-range size, mediated by seasonal variation in productivity (CV month index): (a) low variation (seven least variable sites and regression slope for CV month index of 15); (b) average variation (eight intermediate sites and regression slope for CV month index of 6); (c) high variation (eight most variable sites and regression slope for CV month index of 5); and (d) the marginal effect of annual productivity (fpar mean ) on range size at different levels of seasonal variability (CV month ). fpar mean and CV month have been centred by subtracting means of 35.5 and 37.95, respectively. Different symbols in panels a c indicate different home-range estimators used in individual studies. The broken vertical line in d represents the median CV month value. ln (male range size [km ]) ln (female range size [km ]) Table Model selection criteria for five models explaining the effects of sex and mean body weight on log-transformed home-range size estimates for feral cats in study situations Model terms b 1, b, b 3 K AICc D i x i Log(L) Weight Weight, sex Sex Weight, sex, weight 9 sex Null K, number of parameters; AICc, second-order Akaike s information criterion; D i, difference in AIC between model i and the most supported model; x i, Akaike weight: weight of evidence relative to the other models in the set; Log(L), log-likelihood. Figure 3 The relationship between log-transformed male and female feral cat home-range size estimated from 3 studies across the globe. (IQR) = 3.6], whereas median male range size was 5.1 km (IQR = 8.17). Log-transformed average male home range increased with log-transformed average female home range (F 1,1 = 6.91, P <.1): ln½male range size ðkm ÞŠ ¼ :95ð:13 SEÞ þ :71ð:9 SEÞ ln½female range sizeðkm ÞŠ (3) (adjusted R =.6, Fig. 3). However, differences in homerange size diminished as female home-range size increased because home-range size was log-transformed. 116 Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London

6 A. J. Bengsen et al. Feral cat home-range size and landscape productivity ln range size (ha) ln range size (ha) ln density (no. cats km ) Model-averaged parameter estimates from the best supported models (D AICc < ) describing the relationship between body weight and range size (Table ) indicated that log-transformed range size tended to increase with body weight at a similar rate for both sexes. Average home-range sizes for males tended to be slightly larger than those of females of a similar weight: ln½range size ðkm ÞŠ ¼ :9ð:5Þþ:7ð:18Þsex male :8ð:16 WeightðkgÞ () However, the fixed effects of sex and body weight explained very little of the variation in range size (marginal R =.). Explanatory power was improved when the random site effect was considered (conditional R =.69). Range size and population density Female Male This study Liberg et al. () Figure Relationships between log-transformed range size for female and male feral cats from 17 recent estimates (crosses), contrasted with points (solid circles), regression lines and 95% prediction intervals (shaded) estimated in a previous study (Liberg et al., ). We estimated standard errors of.6 and.8 for the slopes of Liberg et al. s () relationships between population density and female and male home-range size, respectively. The 17 new cases that provided data on estimated range sizes and population densities fitted the form of the relationships described by Liberg et al. (), although two female cases fell outside the prediction interval of the original relationship (Fig. ), and the MSPE of the female cases (.979) was larger than the mean squared error (MSE) of the original model (.31). The MSPE of the male cases (.369), however, was smaller than the MSE of the original model (.668), indicating that the new data lay closer to predicted values than the original data. Observed female and male home ranges in the new datasets were similar to those predicted by the existing models (paired t-test: female t 7 = 1.1, P =.3; male t 8 =.3, P =.7). Discussion The assumption that home-range size scales negatively with landscape productivity is fundamental to several influential models of spatial ecology (e.g. McNab, 1963; Harestad & Bunnell, 1979; Lindstedt, Miller & Buskirk, 1986). Our study is the first that we are aware of to show that home ranges of an invasive species with a broad global distribution scale negatively with landscape productivity, and the first to describe the relationship. Home-range sizes of feral cats varied widely across the situations we examined, as did the range of environmental and ecological conditions. For example, cats occurred as the dominant terrestrial predator at some sites but were sympatric with different predators at others. Prey species also varied widely across studies. Nonetheless, sites where the annual fpar coefficient of variation was less than 3% showed a predictable negative relationship between female range size and an index of average annual productivity, the strength of which decreased as seasonality increased (Fig. a,b). These sites included a wide range of landscape types, such as deserts, coastal heathlands and semi-arid or temperate woodlands and forests (some sites described in Robley et al., 8; Moseby et al., 9; Johnston et al., 1; Buckmaster, 11; Bengsen et al., 1; Johnston, 1; Johnston et al., 13; McGregor et al. 15). The breakdown of the strong negative effect of annual productivity on female range size at sites with highly seasonal productivity is not predicted by existing home-range size optimization models. Inconsistent effects of seasonality on range size have been reported for some carnivores (Nilsen et al., 5), but the mechanisms behind these relationships are unclear and interaction effects have not been described. Given the wide variation in the marginal effect of annual productivity on female range size at these highly seasonal sites (Fig. d), the apparent switch from a negative relationship at sites with low seasonality to a positive relationship is open to question. Nonetheless, it is clear that factors other than mean annual productivity became more important as seasonality increased, or that the methods we used to characterize annual productivity may have been unsuitable at these sites. For example, we estimated productivity over a 31 km circular area at each site; however, some studies at highly seasonal sites only tracked cats within a limited section of the site, such as valley floors Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London 117

7 Feral cat home-range size and landscape productivity A. J. Bengsen et al. and river beds within sites that experienced significant snow cover at higher altitudes (e.g. Harper, 7; Recio & Seddon, 13). Consequently, the productivity that was actually available to cats that provided home-range estimates was likely to have been greater than the low values averaged across the broader area. Species occupying temporally variable habitats might be expected to maintain fixed home ranges that minimize the risk of starvation during periods of low productivity because a flexible strategy, in which ranging activity contracts and expands in response to changes in resource availability, could be prohibitively costly to maintain (Lima, 198; Sandell, 1989). Consistent with this expectation, Ferguson et al. (9) found that home-range sizes of female bobcats (Lynx rufus) scaled negatively with an index describing productivity during the least productive month, although they did not compare their model against any alternatives. The weak support for a similar model in our study suggests that annual productivity is a more practical predictor of home-range size in female feral cats. Similarly, space use and home-range size of feral cats should be influenced by the spatial variation and patchiness of resource distribution (Konecny, 1987); however, our study did not identify an important role for spatial variation in landscape productivity as a determinant of female cat home-range size. The observation that average male home ranges tended to be larger than those of females in the same population, at least at small home-range sizes (Fig. 3), is consistent with previous studies and with expectations for a polygynous mating system (Liberg et al., ). Furthermore, disproportionately large home-range sizes, consistent with wide-ranging behaviour by territorial or itinerant individuals in a polygynous mating system, have been reported for some male cats within populations (e.g. Moseby et al., 9; Bengsen et al., 1). Although average home-range sizes for male cats tended to be slightly larger than those of females for a given body weight (equation ), suggesting that males tended to have larger home ranges than required for metabolic needs alone, there was a great deal of unexplained variance in the weight:home-range size relationship. Landscape productivity probably exerts a greater influence over home-range size than body mass in this widely distributed species, and further attempts to estimate allometric relationships across vastly different feral cat populations are unlikely to be fruitful. Our results supported Liberg et al. s () finding that average home-range size declined with increasing population density. Liberg et al. () believed this relationship was most likely due to range size and population density both correlating with a third variable, namely food availability, but were unable to test this hypothesis. Feral cats tend to be opportunistic generalist feeders (Doherty et al., 15), so it is often difficult to estimate food availability directly (though see Cruz, Glen & Pech, 13; Recio & Seddon, 13). However, recent technologies such as camera trapping are beginning to overcome many of the historic barriers to estimating cat population densities (Bengsen et al., 11; McGregor et al., 15). Evaluations of relationships between population density and landscape productivity may provide some coarse insights into relationships between population density, home-range size and resource availability as more population density estimates become available in future. Conclusion As expected for solitary carnivores, average home-range sizes of feral cats were greatest in areas with consistently poor landscape productivity and smallest in sites with consistently high productivity. The strength of this relationship across populations occurring in a wide range of environmental and ecological contexts highlights the importance of landscape productivity as a driver of the movement patterns and spatial organization of feral cats. The relationships between home-range size and landscape productivity or population density estimated in this study should provide a useful starting point for approximating expected range sizes in many situations where such information is needed for the design of feral cat management operations. The ability to predict average home-range sizes, and to understand the relative importance of different environmental traits to the spatial ecology of feral cats, should be improved if future studies build on this work by relating their findings to local landscape productivity, or directly to the distribution and availability of prey (e.g. Recio & Seddon, 13). The remote-sensing data required to describe landscape productivity are easily accessible, and we encourage all workers investigating the spatial ecology of feral cats and other carnivores to consider the role of landscape productivity in their study systems. Acknowledgements A.J.B. was supported by a Meat and Livestock Australia postdoctoral fellowship. We thank three anonymous reviewers for comments improving an earlier draft. The authors declare that they have no conflict of interest. References Abbott, I., Peacock, D. & Short, J. (1). The new guard: the arrival and impacts of cats and foxes. In Carnivores of Australia: past, present and future: Glen, A.S. & Dickman, C.R. (Eds). Melbourne: CSIRO Publishing. Alterio, N., Moller, H. & Ratz, H. (1998). Movements and habitat use of feral house cats Felis catus, stoats Mustela erminea and ferrets Mustela furo, in grassland surrounding yellow-eyed penguin Megadyptes antipodes breeding areas in spring. Biol. Conserv. 83, Bengsen, A.J., Butler, J.A. & Masters, P. (11). Estimating and indexing feral cat population abundances using camera traps. Wildl. Res. 38, Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London

8 A. J. Bengsen et al. Feral cat home-range size and landscape productivity Bengsen, A.J., Butler, J.A. & Masters, P. (1). Applying home range and landscape use data to design effective feral cat control programs. Wildl. Res. 39, Buckmaster, A.J. (11) Ecology of the feral cat (Felis catus) in the tall forests of far east Gippsland. PhD thesis, University of Sydney. Burnham, K.P. & Anderson, D.R. (). Model selection and inference: a practical information-theoretic approach. New York: Springer-Verlag. Burt, W.H. (193). Territoriality and home range concepts as applied to mammals. J. Mammal., Coops, N.C., Wulder, M.A., Duro, D.C., Han, T. & Berry, S. (8). The development of a Canadian dynamic habitat index using multi-temporal satellite estimates of canopy light absorbance. Ecol. Indic. 8, Cruz, J., Glen, A.S. & Pech, R.P. (13). Modelling landscape-level numerical responses of predators to prey: the case of cats and rabbits. PLoS ONE 8, e735. Dickman, C.R., Denny, E. & Buckmaster, T. (1). Identification of sites of high conservation priority impacted by feral cats. Canberra: Report to the Department of Environment, Water, Heritage and the Arts. Doherty, T.S., Bengsen, A.J. & Davis, R.A. (1). A critical review of habitat use by feral cats and key directions for future research and management. Wildl. Res. 1, Doherty, T.S., Davis, R.A., van Etten, E.J.B., Algar, D., Collier, N., Dickman, C.R., Edwards, G., Masters, P., Palmer, R. & Robinson, S. (15). A continental-scale analysis of feral cat diet in Australia. J. Biogeogr., Ferguson, A.W., Currit, N.A. & Weckerly, F.W. (9). Isometric scaling in home-range size of male and female bobcats (Lynx rufus). Can. J. Zool. 87, Goltz, D.M., Hess, S.C., Brinck, K.W., Banko, P.C. & Danner, R.M. (8). Home range and movements of feral cats on Mauna Kea, Hawai i. Pac. Conserv. Biol. 1, Harestad, A.S. & Bunnell, F.L. (1979). Home range and body weight a re-evaluation. Ecology 6, 389. Harper, G.A. (7). Habitat selection of feral cats (Felis catus) on a temperate, forested island. Austral. Ecol. 3, Herfindal, I., Linnell, J.D.C., Odden, J., Nilsen, E.B. & Andersen, R. (5). Prey density, environmental productivity and home range size in the Eurasian lynx (Lynx lynx). J. Zool. (Lond.) 65, Johnston, M. (1). Field assessment of the Curiosity â bait for management of feral cats after fire at Wilsons Promontory National Park. Heidelberg: Department of Sustainability and Environment. Johnston, M., Algar, D., Onus, M., Hamilton, N., Hilmer, S., Withnell, B. & Koch, K. (1). A bait efficacy trial for the management of feral cats on Dirk Hartog Island. Heidelberg: Department of Sustainability and Environment. Johnston, M., O Donoghue, M., Holdsworth, M., Robinson, S., Herrod, A., Eklom, K., Gigliotti, F., Bould, L. & Little, N. (13). Field assessment of the Curiosityâ bait for managing feral cats in the Pilbara. Heidelberg: Department of Sustainability and Environment. Konecny, M.J. (1987). Home range and activity patterns of feral house cats in the Galapagos Islands. Oikos 5, Liberg, O., Sandell, M., Pontier, D. & Natoli, E. (). Density, spatial organisation and reproductive tactics in the domestic cat and other felids.in The domestic cat: the biology of its behaviour: Turner, D.C. & Bateson, P. (Eds). Cambridge: Cambridge University Press. Lima, S.L. (198). Territoriality in variable environments: a simple model. Am. Nat. 1, Lindstedt, S.L., Miller, B.J. & Buskirk, S.W. (1986). Home range, time and body size in mammals. Ecology 67, Long, J.L. (3). Introduced mammals of the World: their history, distribution, and influence. Collingwood: CSIRO Publishing. McGregor, H.W., Legge, S., Potts, J., Jones, M.E. & Johnson, C.N. (15). Density and home range of feral cats in north-western Australia. Wildl. Res., McLoughlin, P.D. & Ferguson, S. (). A hierarchical pattern of limiting factors helps explain variation in home range size. Ecoscience 7, McNab, B.K. (1963). Bioenergetics and the determination of home range size. Am. Nat. 97, McNaughton, S., Oesterheld, M., Frank, D. & Williams, K. (1989). Ecosystem-level patterns of primary productivity and herbivory in terrestrial habitats. Nature 31, 1 1. Medina, F.M., Bonnaud, E., Vidal, E., Tershy, B.R., Zavaleta, E.S., Josh Donlan, C., Keitt, B.S., Corre, M., Horwath, S.V. & Nogales, M. (11). A global review of the impacts of invasive cats on island endangered vertebrates. Glob. Change Biol. 17, Moseby, K., Stott, J. & Crisp, H. (9). Movement patterns of feral predators in an arid environment implications for control through poison baiting. Wildl Res 36, 35. Nakagawa, S. & Schielzeth, H. (13). A general and simple method for obtaining R from generalized linear mixedeffects models. Methods Ecol. Evol., Nilsen, E.B., Herfindal, I. & Linnell, J.D.C. (5). Can intra-specific variation in carnivore home-range size be explained using remote-sensing estimates of environmental productivity? Ecoscience 1, Oppel, S., Burns, F., Vickery, J., George, K., Ellick, G., Leo, D. & Hillman, J.C. (1). Habitat-specific Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London 119

9 Feral cat home-range size and landscape productivity A. J. Bengsen et al. effectiveness of feral cat control for the conservation of an endemic ground-nesting bird species. J. App. Ecol. 51, Powell, R.A. (). Animal home ranges and territories and home range estimators. In Research techniques in animal ecology: Boitani, L. & Fuller, T.K. (Eds). New York: Columbia University Press. R Core Team (1) R: A language and environment for statistical computing. R Foundation for Statistical Computing. Available at: (accessed 1 August 1). Recio, M.R. & Seddon, P.J. (13). Understanding determinants of home range behaviour of feral cats as introduced apex predators in insular ecosystems: a spatial approach. Behav. Ecol. Sociobiol. 67, Robley, A., Ramsey, D., Woodford, L., Lindeman, M., Johnston, M. & Forsyth, D. (8). Evaluation of detection methods and sampling designs used to determine the abundance of feral cats. Heidelberg: Department of Sustainability and Environment. Sandell, M. (1989). The mating tactics and spacing patterns of solitary carnivores. In Carnivore behavior, ecology, and evolution: Gittleman, J.L. (Ed.). Ithaca: Cornell University Press. Sparkes, J.L., Fleming, P.J.S., Ballard, G., Scott-Orr, H., Durr, S. & Ward, M.P. (1). Canine rabies in Australia: a review of preparedness and research needs. Zoonoses. Public Hlth. 6, USGS (1) MODIS reprojection tool web interface (MRTWeb). US Geological Survey. Available at mrtweb.cr.usgs.gov/ (accessed May 1 to March 1). Wilson, R.R., Blankenship, T.L., Hooten, M.B. & Shivik, J.A. (1). Prey-mediated avoidance of an intraguild predator by its intraguild prey. Oecologia 16, Zuur, A., Ieno, E.N., Walker, N., Saveliev, A.A. & Smith, G.M. (9). Mixed effects models and extensions in ecology with R. New York: Springer. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Characteristics of sites used to describe relationships between feral cat home-range size and landscape productivity, sex and population density. 1 Journal of Zoology 98 (16) 11 1 ª 15 The Zoological Society of London