1 Introduction

Global climate change poses immense challenges to ecosystems and biodiversity conservation (Simmons and Jenkins 2007; Sutton et al. 2022). Projections suggest that the Earth is warming at an unprecedented rate, not seen in the last 40 million years (Simmons et al. 2004). Previous studies indicate that more species are expected to face range contractions of suitable habitats than those experiencing range expansions in the future (Araújo and Guisan 2006; Eustace et al. 2021; Cheng et al. 2023; Sutton et al. 2022). Understanding species-specific, climate-driven distributional shifts of suitable habitats is therefore a vital prerequisite for designing and implementing realistic and flexible conservation strategies (Sutton et al. 2022).

Sub-Saharan Africa is considered the most vulnerable to projected environmental changes, such as global warming and increased drought conditions, which are expected to have far-reaching consequences for ecosystems and landscapes (Simmons et al. 2004; Ogada et al. 2016; Simmons and Jenkins 2007; Phipps et al. 2017). While attention has been traditionally focused on how climate change affects human landscapes (Nhemachena et al. 2020; Thinda et al. 2020; Praveen and Sharma 2019; Fischer et al. 2005), it is becoming increasingly clear that wildlife species and their habitats are similarly under threat, heightening the risk of species extirpation (Sutton 2020; Román-Palacios and Wiens 2020; Mpakairi et al. 2020; Sáenz-Jiménez et al. 2021). Severe extinction risks are particularly evident for tropical species with narrower tolerance ranges, whose climatically suitable habitats are shifting or shrinking (Simmons et al. 2004; VanDerWal et al. 2013).

Among wildlife species, avian scavengers are perhaps some of the most vulnerable to environmental disturbances (Kendall et al. 2018; Margalida and Ogada 2018; Simmons et al. 2004). Previous studies have shown that climate change, working in concert with other environmental factors, can drastically alter avian scavenger habitats and prey availability, resulting in severe population declines (Simmons and Jenkins 2007; Román-Palacios and Wiens 2020). To address this challenge, there is need to develop ecological models that can provide insights into how current and future geographic and environmental spaces for threatened species might be impacted by environmental changes. Such information is important for guiding the design and implementation of effective strategies to monitor and preserve these species now and in the future (Hannah et al. 2007).

Previous studies have primarily focused on proximate threats to African vultures, including poisoning episodes, poaching, habitat loss (such as roosting and breeding sites), food shortages, and collisions with human infrastructure (Margalida and Ogada 2018; Santangeli et al. 2019; Ogada et al. 2022; Ogada et al. 2016; Safford et al. 2019; Kane et al. 2022). However, despite the IUCN identifying climate change as a major threat to numerous species, the potential impacts of projected climate change scenarios on the geographic ranges of avian scavengers such as the African White-backed Vulture (Gyps africanus), (hereafter AWbVs), remain largely under-explored (Phipps et al. 2017; Eustace et al. 2021). Like many other species, avian scavengers such as the AWbVs are not immune to the challenges posed by climate change, which may have unprecedented consequences for their persistence in ecosystems and their role in supporting human health (Simmons and Jenkins 2007).

The AWbV is a huge, long-lived obligate scavenger that is broadly distributed across selected African countries such as Zimbabwe, South Africa, Botswana, Namibian, Zambia, Kenya, Tanzania, Mozambique and some parts of Western Africa (Anderson 2004; Botha et al. 2017; Mundy et al. 1992). This species plays a vital role in ecosystems by providing a wide range of services, including removing carcasses from the environment, curbing the spread of infectious diseases by rapidly disposing of infected carrion, reducing greenhouse gas emissions from decaying carcasses, facilitating nutrient cycling, and supporting recreational and ecotourism activities (Henriques et al. 2020; Margalida and Ogada 2018; Van Den Heever et al. 2021). Classified as critically endangered (BirdLife International 2018), the conservation of the AWbVs is now a global priority (Safford et al. 2019; Santangeli et al. 2019; Bildstein 2022; Kane et al. 2022). However, one of the overarching challenges in prioritising vulture conservation is the lack of spatially explicit knowledge regarding their current geographic and environmental ranges and how these may be affected by projected climate change scenarios (Shrestha et al. 2021; Román-Palacios and Wiens 2020; Kane et al. 2022).

Over the last decade, species distribution modelling (SDM) has emerged as a powerful tool for addressing information gaps regarding the current and future distributional ranges of climatically suitable habitats (Booth 2018; Valavi et al. 2022; Sutton et al. 2021). SDM operates on the premise that suitable environmental conditions exist where species are currently found. By establishing a relationship between species occurrence data and the environmental variables in these locations, SDM allows for the prediction of geographically and environmentally suitable spaces for species across time and space (Phipps et al. 2017; Phillips and Dudík 2008; Booth 2018). This predictive capability is instrumental in updating current species distributional ranges and projecting their future distributions under various climate change scenarios (Sutton 2020; Elith and Leathwick 2009; Booth 2018; Kaky et al. 2020). In addition, SDM can guide research efforts and identify potential reintroduction sites for threatened species (Sutton 2020). In this regard, SDM is considered a critically important tool for developing management strategies that focus on climatically suitable environments where the target species are most likely to thrive (Kumar Jha and Jha 2020; Sutton 2020). When combined with very high-precision Global Positioning System (GPS) data, SDM can yield invaluable insights that support spatially explicit conservation interventions, such as species reintroduction, translocation programmes, and ecosystem restoration (Valavi et al. 2022; Phipps et al. 2017).

The primary aim of this study was to apply the MaxEnt and BIOCLIM SDMs to conduct the first assessment of how climatic constraints may influence the current and future geographic distribution of a sample of six AWbVs tracked within Zimbabwe's Hwange ecosystem. Specifically, it has four objectives:

First, we evaluate how current and projected climate scenarios for 2050 and 2070 may affect the geographical distribution of climatically suitable environments for the AWbVs. Second, we investigate the potential for range expansion, contraction, or shifts under these future climate scenarios and identify areas expected to remain climatically stable despite projected changes. Third, we assess whether these geographical shifts in suitable environments are concentrated within or outside the boundaries of gazetted protected areas, using the Transfrontier Conservation Areas (TFCA) network of Southern Africa as a framework to infer the relevance of protected areas under projected climate scenarios. Finally, we use climate envelope plots generated through the BIOCLIM model (Booth 2018) to demonstrate the extent of AWbVs tolerance ranges for relevant bioclimatic variables. Acquiring reliable information on potential range shifts, contractions, or expansions is fundamentally important for providing anticipatory conservation planning tools that can support effective wildlife management strategies (Hannah et al. 2007; Hannah 2011).

2 Materials and Methods

2.1 Study Area

The study was conducted in Hwange National Park, located in western Zimbabwe, which forms part of the expansive Kavango–Zambezi Transfrontier Conservation Area (KAZA–TFCA; Figure 1). Hwange National Park is home to a diverse array of wildlife species, including African elephant (Africana loxodonta), lion (Panthera leo), buffalo (Syncareous caffer), greator kudu (Tragelaphus strepsiceros), impala (Aepyceros melampus), and zebra (Equus burchelli) (Mpakairi et al. 2020). These animals serve as the primary food source for vultures inhabiting the ecosystem (Mundy et al. 1992). The AWbVs, one of the prominent vulture species in the semi-arid Hwange ecosystem, is often given top priority in conservation interventions due to its rapidly declining population (Safford et al. 2019; Santangeli et al. 2019; Mundy et al. 1992).

3 Results

The models were constructed using thirteen uncorrelated bioclimatic variables and 89,809 georeferenced occurrence data points collected from six AWbVss. The tracked vultures exhibited a broad geographic range spanning six countries: Mozambique, Zambia, Zimbabwe, Namibia, Botswana and South Africa (Figure 2).

The predictive performance of the models was robust, with a high overall mean AUC_TEST of 0.89, indicating that the MaxEnt model possessed strong statistical power to accurately predict the potential distribution of the climatically suitable space for the AWbVs.

Among the bioclimatic variables, annual mean temperature (bio1), isothermality (bio3), maximum temperature of warmest month (bio5), and precipitation of wettest month (bio13) contributed the most to the model.

The current continuous distribution model predicts that most climatically suitable habitats for the AWbVs are concentrated within the KAZA region, which covers countries such as Zimbabwe, Botswana, Zambia, and Namibia (Figure 3). Less suitable ranges are predicted to expand beyond the KAZA region, mainly towards the southwest, encompassing parts of western South Africa. Notably, a significant proportion of these currently less suitable climatic habitats is projected to lie outside legally protected areas (Figure 3). The most suitable range is estimated to cover an area of 378,707 km² (Figure 3).

While the current model presents a restricted climatically suitable geographic range for the AWbVs the scenario shifts considerably under the projected 2050 and 2070 climate models. Across all climate models and GHG emission scenarios, the climatically suitable habitats for the AWbVs are expected to expand substantially. By 2050 and 2070, the largest portion of Southern Africa is predicted to fall under the most suitable category. These expanded suitable spaces are projected to encompass countries such as South Africa, Botswana, Zimbabwe, Namibia, Zambia and Mozambique (Figure 4). Importantly, there is an indication that protected areas will remain climatically stable and continue to support suitable conditions for the AWbVs across all future climate scenarios (Figure 4).

The projections indicate a multi-directional range expansion, with climatically suitable habitats extending into human-dominated landscapes. Most suitable habitats are predicted to become more contiguous, while isolated pockets of less suitable or unsuitable habitats are concentrated in the northern parts of Southern Africa across all GHG emission scenarios (Figure 4).

Suitability distribution maps (Figure 4) generated using MaxEnt reveal slight variations in the predicted suitable spaces across different climate models and GHG emission scenarios. The projections exhibit a high degree of similarity, with a similarity index exceeding 0.93, highlighting the robustness of the model outputs across various scenarios.

When compared to the current predicted range, our results suggest that all future climate scenarios are expected to result in massive expansions of climatically suitable habitats. The HadGEM2-AO-2070-4.5 model predicted the highest percentage increase, at 1487%, while the CCSM4-2050-4.5 model projected the smallest percentage increase, at 1229% (Table 2). None of the models predicted any loss in climatically suitable space for the AWbW. For 2050, the mean percentage gain in suitable habitat under the lower-emission 4.5 scenario across all GCMs is projected to be 1300%, slightly decreasing to 1287% for the high-emission, worst-case RCP 8.5 scenario. For 2070, the mean percentage gain under RCP 4.5 is estimated at 1399%, with a marginal decrease to 1378% under RCP 8.5. Thus, for both the 2050 and 2070 climate scenarios, results suggest that the mean percentage gain under the worst-case scenario is slightly lower compared to the lower-emission scenario (Table 2).

Climatic tolerance ranges for the AWbVs indicate a narrow tolerance for bio3, ranging from 54 to 59 J/K. while for bio1, it can tolerate temperatures between 21°C and 24°C. For bio10, the range extends between 25°C and 28°C, and for bio12, it exhibits an equally narrow tolerance, spanning 350–750 mm (Figure 5).

4 Discussion

Understanding the current and future potential distribution of climatically suitable habitats for endangered species is crucial for developing realistic and targeted conservation strategies (Sutton 2020). Although perfect knowledge of present and potential future geographic ranges is unattainable, applying SDM tools provides useful baseline information to guide ecosystem restoration, population viability assessments, and species reintroduction and translocation programmes (Hannah et al. 2007; Kaky et al. 2020; Phipps et al. 2017). To our knowledge, this study represents one of the first assessments of the potential impacts of climate change scenarios on the spatial and temporal distribution of climatically suitable habitats for the AWbVs.

The predicted distribution models developed in this study show that the current climatically suitable geographic ranges for the AWbVs are largely confined to protected areas within the KAZA region, a known stronghold for the species (Botha et al. 2017; Anderson 2004). This suggests that protected areas are indeed safe havens for the AWbVs (Shrestha et al. 2021; Phipps et al. 2017). Previous studies have shown that AWbVs populations tend to persist in protected areas that offer climatically suitable habitats for foraging, breeding, and roosting (Virani et al. 2010; Buechley et al. 2022; Zvidzai et al. 2022; Kendall et al. 2018). These findings reaffirm the importance of protected territories as biodiversity strongholds that warrant substantial investment in conservation resources to ensure they continue fulfilling their conservation mandate (Shrestha et al. 2021; Buechley et al. 2022).

Projected climate scenarios for 2050 and 2070 suggest that the legally protected areas assessed in this study are expected to remain stable and maintain their climatic suitability, thereby serving as refugia for critically endangered AWbVs. However, due to their spatially fixed and static boundaries, these protected areas may not be sufficient to provide effective protection for critically endangered species such as the AWbVs, whose ranges are projected to expand considerably under future climate scenarios (Eustace et al. 2021; Buechley et al. 2022). This finding contrasts with previous studies that predicted a reduction in the proportion of protected area space expected to remain climatically suitable under projected climate change scenarios (Phipps et al. 2017; Hannah et al. 2007).

The predicted massive expansion of climatically suitable ranges into a mosaic of human-dominated landscapes, as indicated by all model scenarios in this study, contrasts with earlier studies that have projected potential contractions in the suitable geographic ranges of many species due to climate changes (Sutton 2020; Simmons et al. 2004; VanDerWal et al. 2013; Phipps et al. 2017; Shrestha et al. 2021). Although AWbVs face serious threats, the predicted expansion of suitable geographic spaces offers a positive outlook for wildlife managers. This suggests that, under the scenarios explored in this study, the AWbVs could emerge as one of the species that is expected to benefit under future climate change scenarios (Hannah et al. 2007).

However, caution is warranted when interpreting these findings. The exclusive use of bioclimatic variables in model fitting may oversimplify the complex ecological interactions that influence species distributions (Sutton 2020). While bioclimatic factors provide a useful framework for predicting broad patterns, they cannot fully capture the interplay of habitat characteristics, ecological dynamics, and anthropogenic influences.

Bioclimatic variables such as annual mean temperature, isothermality, maximum temperature of warmest month, and precipitation of wettest month emerged as the most important predictors in the models in this study. Temperature and other temperature-derived variables are particularly influential in the movement ecology of large avian species, including African vultures (Campbell 2015; Harel, Horvitz, and Nathan 2016). Previous studies have shown that as large, soaring birds with wide ranges, AWbVs rely heavily on favourable meteorological conditions, such as high temperature, thermal uplift, and convective energy, to locate the unpredictable food resources in the form of carrion (Dodge et al. 2014; Duerr et al. 2015; Safi et al. 2013; Harel, Horvitz, and Nathan 2016).

Moreover, future climate scenarios for the South predict lower rainfall levels, which could support the expansion of suitable habitats for the AWbVs. High rainfall has been shown to negatively impact flight efficiency, foraging success, and breeding outcomes of vulture species (Simmons and Jenkins 2007; Harel, Horvitz, and Nathan 2016). This is corroborated by previous studies that have observed an inverse relationship between vulture breeding success and the amount of rainfall (Virani et al. 2010; Virani et al. 2012).

Assuming that vultures in this study track climatically suitable habitats, this could lead to an escalation in human-wildlife conflicts, presenting new local conservation challenges, though potential ecosystem benefits may also emerge (Henriques et al. 2018). Conservation challenges may arise because human-dominated landscapes are often viewed as incompatible with biodiversity conservation objectives (Buechley et al. 2022). For example, as AWbVs shift into these landscapes to follow climatically suitable spaces, they may be vulnerable to several threats, including poisoning, poaching, scarcity of food, and a lack of suitable roosting and nesting habitats (Ogada, Keesing, and Virani 2012).

However, the expansion of suitable ranges may also create opportunities for coexistence and mutualistic relationships between vultures and humans (Henriques et al. 2018). Vultures may provide essential ecosystem services that enhance human well-being, such as waste disposal, nutrient cycling, greenhouse gas reduction, and regulating the potential spread of diseases (Margalida and Ogada 2018; Santangeli et al. 2019; Buechley et al. 2022).

Some studies have shown that organisms forced out of their current habitats may not necessarily become immediately threatened by new environments (Simmons et al. 2004; Margalida and Ogada 2018). Instead, certain species may quickly adapt and establish viable populations in new settings (Simmons and Jenkins 2007). For example, research on priority conservation areas for vultures, such as the Hooded Vulture (Necrosyrtes monachus), has shown that significant proportions of their habitat lie outside formally protected areas. This suggests that some vulture species may indeed adapt and thrive in human-dominated landscapes (Buechley et al. 2022; Henriques et al. 2018).

Although other studies have demonstrated poleward shifts in climatically suitable habitats for various bird species, including some African vulture species (Simmons et al. 2004; Phipps et al. 2017; Sutton 2020; VanDerWal et al. 2013), the results of this study suggest a contrasting trend. Here, we observed a multi-directional expansion of suitable climate spaces for AWbVs under future climate scenarios. This predicted expansion encompasses countries such as Zimbabwe, Botswana, South Africa, Namibia, Zambia, and Mozambique, which are already recognised as established strongholds for the AWbVs populations (Mundy et al. 1992).

Previous analyses, projecting poleward shifts in species' suitable ranges were mainly based on global temperature metrics, which are closely linked to elevation and latitude (VanDerWal et al. 2013). In contrast, this study used interactions between temperature, precipitation, and other derived bioclimatic variables, which have often been overlooked in previous studies. This methodological difference could possibly explain the contrasting results. The findings of this study are corroborated by Sutton (2020), who observed that within the tropics, climate change may not necessarily drive poleward geographic range shifts but instead result in multi-directional range shifts of suitable habitats (VanDerWal et al. 2013).

Climatically suitable distribution models based on different GCMs and RCP scenarios show a high similarity index (> 93), which is also visibly evident (Figure 4). The high degree of similarity across climate change scenarios suggests the relative stability of the MaxEnt model, a result consistent with findings from other studies (Kaky et al. 2020; Valavi et al. 2022). This consistency underscores the reliability of the MaxEnt model for guiding the design and implementation of wildlife management strategies relevant to both current and future climate trajectories.

Although climate is often considered the primary constraint on the distribution of many species at a landscape scale, we acknowledge the limitations of this assumption. As such, there are good grounds for future studies to integrate the effects of bioclimatic variables with other environmental factors, such as the availability and distribution of carrion, to enhance the accuracy of SDM tools. While this study suggests a potential expansion of climatically suitable habitats for the AWbVs, these findings should be interpreted with caution. The study is based on a relatively small sample size (n = 6) from a limited geographical region, which may restrict the generalisability of the results. Future research could benefit from using a larger sample size from a broader geographic scale to develop more robust species distribution models. Nevertheless, this study provides valuable initial insights into the potential impacts of climate change on the geographic distribution of the critically endangered AWbVs. These findings can serve as a baseline for improving the current and future conservation status of this species.

Author Contributions

Mark Zvidzai: conceptualisation, fieldwork activities, methodology, software, formal data analysis, writing the original manuscript, reviewing manuscript, software, validation, visualisation, data curation. Fadzai M. Zengeya: conceptualisation, supervision, data analysis reviewing the manuscript, visualisation, validation. Mhosisi Masocha: conceptualisation, supervision reviewing manuscript, visualisation, validation. Amon Murwira: conceptualisation, supervision funding acquisition, resources, reviewing of manuscript. Paradzayi Tagwireyi: conceptualisation, fieldwork activities, reviewing and editing the article, visualisation.

Conflicts of Interest

The authors declare no conflicts of interest.