Drivers of leopard (Panthera pardus) habitat use and relative abundance in Africa's largest transfrontier conservation area

Highlights

• Leopards are widely distributed in Africa's largest transfrontier conservation area.

• N-mixture models shed light on factors driving relative abundance.

• Trophy hunting has a top-down impact on leopard relative abundance.

• Leopard conservation plans should strive to secure prey populations.

• Adaptive trophy hunting quotas should be implemented to ensure sustainable offtake.

Abstract

Transfrontier conservation areas (TFCAs) have the potential to provide havens for large carnivores while preserving connectivity across wider mixed-use landscapes. However, information on the status of species in such landscapes is lacking, despite being a prerequisite for effective conservation planning. We contribute information to this gap for leopard (Panthera pardus) in Africa, where the species is facing severe range contractions, using data from transect surveys of a 30,000km² area across Botswana and Zimbabwe in the Kavango-Zambezi (KAZA) TFCA. We used occupancy models to assess how biotic, anthropogenic, and management variables influence leopard habitat use, and N-mixture models to identify variables influencing the species' relative abundance. Leopard were detected in 184 out of 413 sampling units of 64km²; accounting for imperfect detection resulted in mean detection probability $\overline{\widehat{p}}$ = 0.24 (SD = 0.06) and mean probability of site use $\overline{\widehat{\mathrm{\psi }}}$ = 0.89 (SD = 0.20). Habitat use was positively influenced by prey availability and high protection. Relative abundance was best predicted by trophy hunting, which had a negative influence, while abundance was positively associated with high protection and availability of steenbok. Our findings suggest that securing prey populations should be a priority in conservation planning for leopard in Africa, and underline the necessity of preserving highly-protected areas within mixed-use landscapes as strongholds for large carnivores. Our findings also support calls for better assessment of leopard population density in trophy hunting areas, and illustrate the value of N-mixture models to identify factors influencing relative abundance of large carnivores.

Introduction

Despite being widely considered one of the most adaptable of the world's large felids, the leopard (Panthera pardus) is classified as vulnerable by the IUCN (Stein et al., 2016). In Africa, the species is primarily threatened by habitat loss and fragmentation (Di Minin et al., 2016b), prey depletion (Wolf and Ripple, 2016), and direct persecution by humans (Inskip and Zimmermann, 2009), which have collectively contributed to the species losing at least 48% of its historical range on the continent (Jacobson et al., 2016).

As apex predators, leopards and other large carnivore populations require large, connected landscapes and viable prey populations to thrive (Crooks et al., 2011). However, the connectivity beyond protected area borders required by these species is often at odds with the increased demand for land associated with growing and developing human populations (Di Minin et al., 2016b). The resulting habitat fragmentation has contributed to substantial range contractions for members of the large carnivore guild across the world (Wolf and Ripple, 2017). In light of this, there is a growing consensus that, while highly-protected areas remain a vital component of many large carnivore conservation strategies (Karanth and Chellam, 2009), small protected areas alone will not be sufficient to maintain viable populations of large carnivores into the future (Ripple et al., 2014). A shift from the traditional protected area-centric approach to a landscape-level approach to conservation holds particular promise for leopards in Africa, the majority of which are believed to occur outside strictly-protected areas (Hunter et al., 2013).

In Africa, transfrontier conservation areas (TFCAs) embody this movement, while emphasising the coupling of conservation with development initiatives. Defined as areas spanning international borders and encompassing multiple protected areas and land use types, TFCAs are managed as a single contiguous landscape for conservation (SADC, 1999). If effectively managed, TFCAs have the potential to protect large swathes of prime habitat and maintain connectivity within wider mixed-use landscapes, and thus provide havens for leopard within their contracting African range.

Until now, however, there has been relatively little effort to understand how leopards are faring across these larger conservation areas (Balme et al., 2014; Jacobson et al., 2016), with most research having been restricted to small, highly-protected reserves (but see Balme et al., 2010b; Henschel et al., 2011; Strampelli et al., 2018). There is therefore a need to increase knowledge of the species within the context of large, mixed-use landscapes like TFCAs. Understanding how leopards adapt to different components of these modern mosaic landscapes is essential to inform management and conservation planning, evaluate the conservation effectiveness of different land-use types, and ensure the persistence of Africa's leopard populations.

One land-use strategy of particular relevance to the leopard in Africa is trophy hunting, whereby a set quota of individuals may be hunted within a designated area each year. Trophy hunting has been used by conservation managers to attach economic value to wildlife areas by generating revenue for governments and local communities, and thus secure conservation benefits for natural habitats often unsuitable for photographic tourism (Lindsey et al., 2007). In the absence of viable conservation-oriented land use alternatives, trophy hunting is argued to therefore play an important role in preventing their conversion to agricultural land (Di Minin et al., 2016a; Dickman et al., 2019). However, the practice can have detrimental long-term impacts on hunted populations if carried out unsustainably or combined with other sources of anthropogenic mortality (Packer et al., 2010; Lindsey et al., 2013). Information on these impacts is lacking even for species which have attracted considerable attention from policymakers (e.g. lion, Panthera leo; Macdonald et al., 2017), and these gaps are greater still for leopard. Leopard populations within hunting areas should be assessed and monitored, and their habitat-use mechanisms understood, in order to ensure detrimental effects of hunting are avoided, and to inform sustainable and adaptive hunting management plans (Balme et al., 2010b).

We employed data from large-scale spoor surveys to investigate leopard status and habitat use across a mixed-use landscape within the southern Kavango-Zambezi (KAZA) TFCA. We used occupancy models to estimate the proportion of this landscape used by leopard, and assess how a suite of biotic, anthropogenic, and management variables influence habitat use for the species. We then employed novel N-mixture models to identify factors influencing relative abundance of leopard within the study area, and discuss the implications of our findings to highlight conservation priorities for leopard across modern African conservation landscapes.