Introduction

Geophagy, the deliberate ingestion of earth, is common among free-ranging mammals (Krishnamani and Mahaney 2000) and rural human populations (Aufreiter et al. 1997). The activity tends to be localised at widely scattered mineral licks (Abrahams 1999). In general, licks are characterised by moderate to high clay and nutrient content (Klaus and Schmid 1998).

Although geophagy is widespread, the primary drivers of this behaviour remain poorly understood (Abrahams 1999). Lick use is costly for a number of reasons, including the energetic cost of reaching a lick, tooth wear due to sand consumption and the increased predation risk incurred at lick sites (Klaus et al. 1998). Geophagy must thus provide individuals with benefits to offset the costs. Medicinal benefits have been evoked to explain geophagy, including detoxification of secondary plant compounds (Klaus et al. 1998; Krishnamani and Mahaney 2000), countering the effects of acidosis (Krishnamani and Mahaney 2000) and remediation of diarrhoea (Krishnamani and Mahaney 2000). The most common hypothesis proposed, however, is that of nutrient supplementation (Abrahams 1999; Eksteen and Bornman 1990; Kreulen 1985; Mills and Milewski 2006). Licks vary considerably in their chemical and physical properties, and the benefits of geophagy are likely to vary between sites and among species and individuals.

Loskop Dam Nature Reserve, located in Mpumalanga, South Africa, was established in 1942 and encompasses 22,850 ha (including the 2,350 ha surface area of the dam). Although clinical nutritional deficiencies have not been recorded at Loskop Dam Nature Reserve, artificial lick blocks, purchased from agricultural suppliers, were provided for two dry seasons in the 1980s. They were under-utilized and consequently removed. However, the almost continual presence of wildlife within the vicinity of natural licks (Table 1) and the presence of earth within dung (pers. obs.) indicate that several natural licks are regularly used by wildlife in the reserve.

Table 1 Wildlife observed at lick sites within Loskop Dam Nature Reserve
Full size table

Materials and methods

Samples of lick earth (approximately 1 m deep in the regolith, n = 3) collected from the exposed face of a geophagic bank and adjacent topsoil (0–5 cm, n = 3) were collected from three sites in the reserve in 2005, namely Main Lick (Lick M), Klopperskloof Lick (Lick KK) and Rhenosterhoek Lick (Lick RH) (Fig. 1). [Eksteen and Bornman (1990) reported on results from Lick M and Lick KK.] In addition, kudu (Tragelaphus strepsiceros) dung (clayey pellets with no visible macro-organic matter) was collected from the base of the ‘lick wall’ at Lick M (Fig. 2), and earth samples were obtained from a kaolinitic lick (Lick LH) just outside the reserve where local people are known to collect lick earth and practise geophagy. All samples were analysed by inductively coupled plasma mass spectrometry (ICP-MS; 27 elements) at the University of Stellenbosch after digestion in nitric acid and hydrogen peroxide (EPA 2007). The method reports a high average percentage recovery (between 86% and 101%) for a range of elements. Replicate analyses of standard reference materials and other soil samples from South African savannas had relative standard deviations (calculated as the standard deviation divided by the mean and multiplied by 100) ranging from 0.1% to 29% across all elements (Table 2). Given the small sample size, statistical analysis of the data was not appropriate.

Fig. 1
figure 1

The geology of the Loskop Dam Nature Reserve and the location of the mineral licks sampled in the study

Full size image
Fig. 2
figure 2

The kudu dung collection site at the base of the ‘lick wall’ at Lick M

Full size image
Table 2 Element concentration (mg kg−1) of lick earths, adjacent topsoils and kudu dung at Loskop Dam Nature Reserve
Full size table

Results and discussion

Lick earth was geochemically distinct from the adjacent topsoils (Table 2). Micro-nutrient as well as Na concentration was greater in lick earth than in topsoil (B, Co, Zn, Se, Mo and Mn at Lick M, Cu at Lick KK and Na at both Lick KK and Lick RH). Differences between lick earth and topsoil varied according to site, suggesting that animals practise geophagy for different nutrients at different sites, or for reasons other than nutrition (e.g. antacid or plant toxin binding purposes). The concentrations of B, Co, Zn, Se, Mo and Mn in kudu dung were less than those in the lick earth collected at Lick M, which suggests that these nutrients were assimilated from the earth in the kudu’s gastro-intestinal tract. (This is based on the assumption that the dung samples were comprised predominantly of mineral lick material, a not unreasonable supposition given the absence of macro-organic material in the dung.)

In contrast to the above results, the kaolinitic earth collected by local people (Lick LH) had a lesser nutrient concentration than all topsoils sampled (Table 2). Kaolinite is an effective adsorbent of toxins, and lick earths with a great kaolinite concentration are often ingested by humans to relieve gastro-intestinal disorders (Aufreiter et al. 1997). The kaolinitic earth sample was poor in all nutrients, and consumption was thus probably for the gastro-intestinal benefits of the kaolinite.

The results suggest that geophagy at Loskop Dam Nature Reserve may be supplementing animals with: B, Co, Zn, Se, Mo and Mn (at Lick M); Cu (at Lick KK); and Na (at Lick KK and Lick RH). Selenium deficiency may be driving lick use at Lick M. This latter benefit is based on the observations that: (1) the Se concentration in topsoil adjacent to Lick M is almost three times less than the global average (0.33 mg kg−1) (Kabata-Pendias and Pendias 2001), suggesting that it is a particularly scarce nutrient; and (2) the Se concentration of Lick M is threefold that of the adjacent topsoil. This corroborates the findings of Mills and Milewski (2006) in the Ngorongoro Conservation Area, Tanzania, where large licks were exceptionally enriched in Se.

In contrast to our results, Eksteen and Bornman (1990) found that Na was enriched at Lick M as well as Lick KK. We ascribe the different results to variations in extraction methods employed. [Eksteen and Bornman (1990) used a 1:2.5 soil-to-water extract.] Although Mn is enriched in Lick M relative to the adjacent topsoil and the global average (437 mg kg−1) (Kabata-Pendias and Pendias 2001), it is unlikely that Mn deficiency, which is rare in mammals, is driving geophagy at this site (Hurley 1982). Given that Mn oxides are strong adsorbers of a range of elements, the general element enrichment at Lick M may be indicative of a zone of Mn oxide precipitation.

If licks, such as those at Loskop Dam Nature Reserve, are supplementing wildlife with micro-nutrients (as well as Na) important for fecundity and health, they may be important in conservation planning and tourism management (mineral licks could for example be used to provide opportunities for wildlife viewing). It is likely that, prior to reserve formation, roaming wildlife had access to various rock and soil types potentially supplying them with a range of scarce nutrients. The present availability of licks within small reserves, such as Loskop Dam Nature Reserve, may consequently be inadequate for conserving viable populations of certain wildlife species. Artificial supplementation of scarce micro-nutrients may therefore be necessary for long-term management of wildlife.