Frogs are an integral component of most food webs and ecosystems because their complex life cycle enables them to influence energy exchanges within aquatic and terrestrial environments (Cuthbert et al. 2022). They are generally semi-aquatic, requiring water for reproduction purposes while using adjacent riparian and terrestrial habitats as adults (Du Preez & Carruthers 2017). As such, waterbodies (i.e. floodplain pans, vleis, reservoirs and rivers) and their riparian zones typically support an elevated diversity and communities of frogs at the landscape level.

The Kruger National Park (KNP) is one of the largest protected areas in Africa (Eckhardt, Van Wilgen & Biggs 2000), and is characterised by a sprawling network of rivers, streams and floodplain pans (Riddell et al. 2019). For these reasons, the area has been the focus of many scientific studies to elucidate the biodiversity associated with these aquatic ecosystems (Broomker 1994; Majdi et al. 2022; Pienaar 1968; Wolmarans & De Kock 2006). When coupled with eyewitness accounts from rangers, scientists and visitors, KNP has amassed an impressive species list spanning most taxa. The incorporation of technological advances and the popularisation of citizen science platforms has also improved our understanding of species assemblages within the region. This is particularly true of the frogs, with historical publications (e.g. Pienaar, Passmore & Carruthers 1976) and present citizen science platforms (e.g. FrogMap 2023; iNaturalist 2023) contributing to a robust dataset of records spanning the entire park.

Irrespective of the high number of amphibian records and their apparent importance, differences in the locality data of frogs emerge when comparing baseline literature sources for KNP. For example, the Animal Demographic Unit (ADU) (FrogMap 2023), which collates data from a multiplicity of sources, lacks records that have been listed elsewhere (Pienaar et al. 1976; Vlok et al. 2013). This is largely because the FrogMap (2023) has yet to incorporate additional museum and literature records into its database since the publication of the Frog Red Atlas (Minter et al. 2004). While the data certainly exist, it is challenging to obtain and interpret for conservation practitioners, who lack advanced knowledge in amphibian systematics and biogeography. This apparent disconnect between data sources is more pronounced in the northern sections of KNP, which is problematic given the conservation significance of this region of the park (Dzurume, Dube & Shoko 2022).

In the north of the KNP, the Makuleke Contractual National Park (MCNP) has received much interest from aquatic ecologists in recent years because of its expansive network of inter-connected permanent and ephemeral pans (Dyamond 2017; Kock 2017; Malherbe et al. 2019). The MCNP is a conservation area in the northern reaches of the KNP that has shared conservation management between South African National Parks (SANParks) and the community of Makuleke who own the land (Figure 1). The park is a product of a newly evolving conservation strategy termed ‘community-based natural resource management’ (CBNRM) (Reid & Turner 2013). The floodplain pan system, collectively referred to as the Makuleke Wetlands, is a Ramsar-declared wetland system, which can be found mainly within the geographical limits of the MCNP, with several pans located just south of the Luvuvhu River. Although relatively well sampled by aquatic ecologists, the area has received less attention from an amphibian point of view, with much of the frog-centric work being localised south of the Luvuvhu River.

To address the knowledge gap, we aim to investigate the frog diversity within the Makuleke Wetlands region using published literature, citizen science platforms, museum records and active field surveys. In doing so, we will develop an updated baseline to inform conservation practices.

Field surveys of the study region resulted in the documentation of 18 species comprising 10 families of frogs (Table 1; Figure 2) Voucher specimens representing the amphibian diversity of the region were taken, resulting in the cataloguing of 83 specimens into the PEM herpetological collection (Online Appendix Table 1-OA1). The collection was also supplemented with an additional 58 DNA samples, resulting in a total of 141 novel DNA samples for the region (that can be used in future phylogenetic and conservation research). Using published literature and citizen science records, the most comprehensive species lists for the region came from ‘pre-2004’ with 28 species being recorded in the region. This was followed by ‘post-2004’ with 24 species and ‘this study’ with 18 species. Using all available records from the literature, citizen science platforms and this study, the total species list for the MCNP was found to be 30 species (Table 1). One of the more notable records from our field surveys was the collection of a Natal sand frog – Tomopterna natalensis within the study area (confirmed using keys from Du Preez & Carruthers 2017), as it represented the first record of the species for QDS 2231AD, and the first confirmed record of the species for the region. Phylogenetic reconstruction of Afrixalus using the maximum likelihood algorithm (ML) revealed a well-supported sister relationship between GPN02 + PEM A12181 and A. aureus Genbank Accession Number - PP913975) (Figure 3).

The field surveys combined with historical records produced the most comprehensive species list for the area to date, with a total of 30 frog species within the MCNP. In addition to increasing the known species diversity of the area, this study highlights discrepancies in our understanding of the biodiversity of a region when individual sources are utilised in isolation.

The most comprehensive source, Minter et al. (2004), which filtered into FrogMap (2023), collated data from citizen scientists and museum records to provide nationwide biodiversity baselines. While Minter et al. (2004) acknowledged the presence of an Afrixalus sp. north of the Letaba River, these records were not attributed to species level and did not make their way into any recent publications. In Pienaar et al. (1976), however, there are several records of Afrixalus brachycnemis brachycnemis flanking the Luvuvhu River, with one historical record within QDS 2231AD. Since the publication of Pienaar et al. (1976), Afrixalus b. brachycnemis has been split into three species, namely, Afrixalus delicatus, Afrixalus crotalus and Afrixalus aureus (Pickersgill 1984). The only record since Pienaar et al. (1976) is an iNaturalist record from 2022 (www.inaturalist.org/observations/59764353), initially identified as Afrixalus aureus based on its proximity to specimens of the same species from southern KNP. Additionally, our sample (PEM A12181) was also assumed to belong to this species based on its close geographic proximity to the iNaturalist record and all records from Pienaar et al. (1976), which had been recognised as A. aureus in Minter et al. (2004) and Vlok et al. (2013). A superficial look at the morphology and the results of gene blasting (high similarity of PEM A12181 to KZN A. aureus) also suggested A. aureus. Closer inspection of the specimen, however, revealed it to be a different species of Afrixalus entirely, based on the absence of a dorsal tibia band.

Before this study, there were no publicly available records of A. crotalus on GenBank, but based on our phylogenetic reconstruction, the record from Gorongosa National Park (GPN02), which was identified as A. delicatus in Portik et al. (2019), is a misidentified A. crotalus. Delicate leaf-folding frog – Afrixalus delicatus is a widely distributed species, occurring from eastern South Africa, into northern Tanzania. Snoring leaf-folding frog – Afrixalus crotalus is a small, cryptic leaf-folding frog, which according to its original description (Pickersgill 1984) is distributed from eastern Zimbabwe northwards to southern Malawi and eastwards through central Mozambique to the coast. The availability of a more robust dataset has revealed a more expansive distribution that stretches from southern Zambia to the northernmost tip of South Africa (Channing & Rödel 2019). In addition to being morphologically similar, both species are sympatric in central Mozambique, which can confound field identification for these animals (Channing & Rödel 2019).

Although the presence of A. crotalus, in South Africa, was alluded to in Channing and Rödel (2019), this is yet to be confirmed with any verifiable records. The well-supported relationship between GPN02 + PEM A12181 and A. aureus coupled with the morphological characteristics in our sample, and the iNaturalist record lends strong support to the presence of A. crotalus in northern South Africa. While we were unable to confirm the IDs of the northern Afrixalus specimens from Pienaar et al. (1976), it is highly likely that these specimens will also represent specimens of A. crotalus. This would suggest that A. crotalus is restricted to the northern fringes of Limpopo within South Africa while A. aureus does not extend north of the Letaba River.

Sclerophrys gutturalis is a widespread and generalist toad species that is broadly expected to occur in the region based on the proximity of records on FrogMap (2023), north and west of the target QDS’s and the distribution maps supplied by both Du Preez & Carruthers (2017) and Channing and Rödel (2019). In Vlok et al. (2013), S. guturalis was recorded within the region. Interestingly, neither the combined efforts from our study nor previous research (excluding Vlok et al. 2013) managed to document the species within the MCNP. Sclerophrys gutturalis is one of only two species absent from the pre-2004 list, even though the species is considered locally abundant throughout much of its range. Unlike S. pusilla, which is common in low-lying areas, S gutturalis is less common in these areas and tends to be abundant in the higher-lying, western parts of KNP (Pienaar et al. 1976). Given that much of our sampling was focussed on lower-lying areas, near ephemeral pans, it is unsurprising that we did not encounter S. gutturalis here. Increased sampling in the western parts of the MCNP would likely produce S. gutturalis records.

When comparing the specimen records following the completion of the amphibian red atlas (Minter et al. 2004), several species are yet to be documented again. These include Breviceps mossambicus, Cacosternum boettgeri, Hildebrandtia ornata and Hyperolius pusillus. While the absence of the latter three species is surprising given their presence in adjacent QDS’s, the absence of B. mossambicus is likely a product of their cryptic nature. Their terrestrial lifestyle coupled with their specialised reproductive strategy (water-independent tadpole stage) (Du Preez & Carruthers 2017; Passmore & Carruthers 1995) precludes these animals from water edges, which decreases encounter probability. Additionally, rain frogs often call from concealed positions in highly heterogeneous terrestrial environments near burrows. Combined with their preference to call during or after bouts of rain, this makes the discovery of these animals very challenging, with trained herpetologists often resorting to triangulation to locate these animals (Passmore & Carruthers 1995). Although infrequently encountered by past researchers, the environment and microhabitats offered by MCNP are ideal for Breviceps spp. In addition to their cryptic nature, the taxonomy of Breviceps spp. is currently in flux and the specimens regarded as B. mossambicus in Pienaar et al. (1976) are likely B. passmorei as B. mossambicus is not expected to occur in South Africa (Nielsen et al. 2018).

The presence of Tomopterna natalensis from just north of Pafuri Research Camp, on the boundary of the MCNP was not only the first confirmed record of the species for QDS 2231AD, but the first record for the MCNP. The species has been recorded south of the study area on iNaturalist (www.inaturalist.org/observations/71876167; www.inaturalist.org/observations/70820437), further corroborating our morphological identification of the specimen. Tomopterna spp. encountered during the field trips included Tomopterna krugerensis, Tomopterna adiastola and Tomopterna marmorata. The latter two were omitted from our study survey list as they were recovered close to the Punda Maria Camp Site, outside of the designated area for this study. The same applies to Phrynobatrachus natalensis, which was also only recovered in the Punda Maria Camp Site during our survey of the region. These samples have, however, been included in Online Appendix Table 1-OA1 for completeness. Although not all documented within the MCNP, all four Tomopterna spp. and P. natalensis are expected to be found within the Ramsar-declared wetland system based on previous surveys of the region (Table 1).

While the field surveys conducted during this study could not locate and confirm the presence of all the frogs previously recorded from the area, we are confident that a more robust sampling strategy (more time and more sites) would reveal all 30 species. While frogs are often associated with water, their ecologies and behaviours are complex as they also use terrestrial landscapes during their adult stages. Seasonality and climatic conditions coupled with topography and a varying suite of abiotic and biotic conditions work synergistically to create the ideal conditions and microhabitats for different frog species. As we could not cover all sections of the park and focussed on the ephemeral pan systems, it is unsurprising that certain species were not encountered.

These findings, coupled with the smaller size and lower perceived value of anurans, lend support to the possibility of new distributional records and even new species for the area given more sampling effort. This means that although well sampled, knowledge of the area is outdated and thus not subject to the same scrutiny and insights as modern findings. Following this line of reasoning, it is safe to assume that there is still much to uncover in the MCNP. Prior to this study, Hylambates maculatus, Xenopus muelleri and Sclerophrys pusilla (FrogMap 2023) had not been recorded in their quarter degree cells since 1960, 1973 and 1996, respectively. This highlights the importance of these studies in confirming species presence in areas through time, which empowers conservation practitioners with the tools necessary to make informed decisions regarding species assemblages. Lastly, in addition to providing a robust species and updated list for the MCNP, this study highlights the taxonomical changes (Table 1), since Pienaar et al. (1976), and provides the most up-to-date nomenclature for biologists, conservations and managers working in KNP.