The LNR (-24.14, 30.28) is an 18 000 ha provincial nature reserve in the Limpopo province of South Africa (Figure 1). The reserve lies just south of the Wolkberg Wilderness Area near Haenertsburg, Hoedspruit and Tzaneen in the Wolkberg Mountain range, and it forms part of the Limpopo Drakensberg. It is characterised by kloofs filled with indigenous forests, interspersed with grassland hilltops (Figure 2). Lekgalameetse means ‘Place of Water’, and the reserve is the catchment area for the Letaba River that flows east and supports several Limpopo PA near the Kruger National Park. Two perennial east-west flowing rivers, the Selati and Makutsi, and a number of smaller streams dissect the reserve.

The LNR consists of a relatively low-lying (750 m – 900 m) eastern side that contains Tzaneen Sour Bushveld; a central mountain ridge (reaching 1800 m at the highest point) encompasses the Northern Escarpment Quartzite Sourveld, interspersed with patches of Northern Mistbelt Forest at altitudes from 900 m – 1650 m; and the western plateau of the Park (‘The Downs’), which consists of Wolkberg Dolomite Grassland (Mucina & Rutherford 2006; Williams & Altenroxel 2014).

As a team effort, SANSA has overcome some of the problems associated with invertebrate monitoring, such as sorting, identifying large samples and using a standardised sampling protocol (Haddad & Dippenaar-Schoeman 2015). The SANSA sampling protocol requires that four habitat types representative of the area be selected in the degree-square grid being sampled. All the habitat layers are sampled, including pitfall traps, litter sifting, Winkler traps (all targeting ground dwellers), sweeping and beating (plant dwellers) and hand and night collecting (ground- and plant dwellers). Five surveys were undertaken at the LNR to sample arachnids (Table 1) and the following areas were sampled: The Downs (-24.14, 30.31), Farm Balloon (-24.20, 30.34), Farm Malta (-24.17, 30.25), Farm Haffenden Heights (-23.82, 30.16) (Figure 1) and Church Forest (-24.16, 30.25) (Figure 2c).

Species determinations were performed by three of the authors. Voucher specimens are deposited in the National Collection of Arachnida housed at the Agricultural Research Council – Plant Protection Research Institute, Pretoria (NCA). The taxonomy of some families, for example, Agelenidae, Amaurobiidae, Dictynidae, Liocranidae and Theridiidae, is still unresolved and some specimens could not be identified to species level. Generic names were included in the checklist when immature specimens were sampled and could not be identified to species level, and when these genera were not represented by other identified species.

Because most spiders live in a defined environment with limitations set by both physical conditions and biological factors, species can be grouped into guilds or functional groups based on available information on their habitat preferences and predatory methods. Functional groups provide additional insight on habitat functioning and utilisation by spiders, and contribute to our knowledge of ecosystem functioning. This is because spiders have diverse lifestyles and have developed diverse methods of capturing prey.

Spiders are broadly grouped as web builders (WB) and free-living wanderers (W) (Foord et al. 2011a). The wanderers can be further divided into plant wanderers (PW) and ground wanderers (GW), and the GW into the free-living ground dwellers (FGW) and burrow dwellers (BGW). The web-building spiders are subdivided into different guilds based on the web types they construct: funnel-web builders (FWB), gumfoot-web builders (GWB), orb-web builders (OWB), modified orb-web builders (MOWB), retreat-web builders (RWB), sheet-web builders (SWB) and space-web builders (SPWB).

The conservation status of the species is important for conservation agencies and in preparing the first Red Data List of spiders. As part of the first Spider Atlas, an endemicity index was provided for each species (Dippenaar-Schoeman et al. 2010). It was calculated based on currently known distribution, which included six endemicity categories: 6 = endemic, known only from the type locality, 5 = known from one province only, wider than the type locality, 4 = known from two adjoining provinces only, 3 = South Africa, known from more than two provinces or two provinces that are not adjoining, 2 = southern Africa (south of Zambezi and Kunene Rivers), 1 = Afrotropical region and 0 = Africa and wider. In this paper, the following categories are recognised: LC = least concern, species with a wide distribution falling in categories 0–4, LE = Limpopo endemics (5), RE = reserve endemics (6) and DDT = data deficient for taxonomic reasons. This could be because of the unresolved taxonomy of the group or the species being undescribed.

As part of SANSA, a photographic Virtual Museum was developed to access photographs of arachnid species (Dippenaar-Schoeman et al. 2015). Spiders sampled during three of the surveys at LNR were photographed by the last author. A photo gallery of the spiders of LNR will be made available on the SANSA website. Images can also be viewed at http://www.arc.agric.za:8080/.

Forty-five families represented by 168 genera and 268 species have been collected from LNR from 2009 to 2016 over a total of only 19 sampling days (Table 2 and Online Appendix 1). The number of species collected compares well with spider diversity recorded from other Limpopo reserves sampled over periods of 1–12 months: Blouberg Nature Reserve (186 spp., Muelelwa et al. 2010); Nylsvley Nature Reserve (175 spp., Dippenaar-Schoeman et al. 2009); Western Soutpansberg Conservancy (222 spp., Muelelwa et al. 2010); Makelali Private Game Reserve (286 spp., Whitmore et al. 2001) and Polokwane Nature Reserve (275 spp., Dippenaar et al. 2008). This averages 228 species per reserve, ranging from 175 to 286 species.

Of the 268 species sampled, 33 spp. (12.3%) were data deficient and could not be identified to species level, 13 of them possibly being new to science (Table 3). However, these putative new species are representatives of species-rich families, and only after revisionary studies would it be possible to tell whether they are indeed new to science. Only the endemic species falling into categories 3–6 (75 spp., 29.0%) need to be evaluated using the International Union for Conservation of Nature (IUCN) criteria. The majority of the species sampled have a wide distribution in southern Africa or the Afrotropical Region and can be listed as LC (160 spp., 59.7%), while only 75 spp. (28.0%) are South African endemics.

Two species are so far unknown outside the reserve, Hasarinella distincta Haddad & Wesołowska, 2013 (Salticidae) and Ballomma legala Jocqué & Henrard, 2015 (Zodariidae), while five Limpopo endemic species are protected in the reserve: Hortipes coccinatus Bosselaers & Jocqué, 2000 (Corinnidae), Xevioso lichmadina Griswold, 1990 (Phyxelididae), Euophrys limpopo Wesołowska, Azarkina & Russell-Smith, 2014 (Salticidae), Rumburak tuberatus Wesołowska, Azarkina & Russell-Smith, 2014 (Salticidae) and Palfuria spirembolis Szuts & Jocqué, 2001 (Zodariidae).

During this study, new distribution and life history data on several species have been collected. Two species have been recorded from South Africa for the first time: Ctenus caligineus des Arts, 1912 (Ctenidae), a species previously only known from D.R. Congo and Burundi, and Belippo calcarata (Roewer, 1942) (Salticidae), a species previously known only from Angola and D.R. Congo. Many described species are known only from one sex, and during this survey the unknown sex of three species was sampled but has not yet been described: the females of Hogna adjacens Roewer, 1959 (Lycosidae), Anyphops leleupi Benoit, 1972 (Selenopidae) and Tmarus natalensis Lessert, 1925 (Thomisidae).

Only 5 species, H. distincta, Heliophanus ndumoensis Wesołowska & Haddad, 2013, Langona hirsuta Haddad & Wesołowska, 2011, Thyenula sempiterna Wesołowska, 1999 and B. legala were previously recorded from the LNR (Haddad & Wesołowska 2013; Wesołowska et al. 2014; Jocqué & Henrard 2015) and 263 spp. are recorded from the reserve for the first time. Presently, 2220 spider species are known from South Africa, thus 12.2% of South African species are protected in this reserve.

Results from the Savanna Biome indicated that four spider families consistently dominate assemblages in terms of species richness (Foord et al. 2011b): Araneidae, Gnaphosidae, Salticidae and Thomisidae. In this study, three families correspond to this pattern, with the Salticidae (41 spp.), Araneidae (38 spp.) and Thomisidae (32 spp.) being the most species rich. The only deviation was the high diversity in the Theridiidae (20 spp.), while the Gnaphosidae was represented by only 9 spp. (Table 2 and Online Appendix 1). The low diversity of Gnaphosidae could be explained by their generally higher diversity in more arid savanna habitats in South Africa, while most of the sites in Lekgalameetse are mesic in character.

Salticidae: The Salticidae are free-living spiders found on tree trunks, soil, rocks and vegetation. They build small silk nests attached to various substrates, in which they moult, oviposit and sometimes mate, or which they occupy during periods of inactivity (Dippenaar-Schoeman 2014). Of the 41 spp. sampled, 37 are new records for the LNR, 1 is a LNR endemic (H. distincta), Belippo calcarata is recorded from South Africa for the first time and the Rhene sp. has been identified as possibly new to science (data deficient). Three other species are LE (7.3%), a further 9 species are South African endemics (21.9%), 10 species are southern African endemics (24.4%), while the rest (18 spp., 43.9%) are more widely distributed in Africa (Online Appendix 1).

Araneidae: The Araneidae are WB and produce typical orb-webs (OWB) and modified orb-webs (MOWB). The taxonomy of many genera in Africa is still unresolved and some of the species listed here might even be new. The family is very diverse, and most genera and species have a wide African distribution. Of the 38 species, 5 spp. (13.2%) are data deficient and the rest (33 spp., 86.8%) are of LC because of their wide distribution throughout Africa. Only three species, Nemoscolus elongatus Lawrence, 1947, Ideocaira transversa Simon, 1903 and Larinia natalensis (Grasshoff, 1971), are South African endemics, but they all have a wide distribution in the country. Members of two genera, Acanthepeira and Larinioides, were reported for the first time from South Africa (Dippenaar-Schoeman 2014). Species of four genera (Acanthepeira, Araneus, Larinioides and Gea) have been identified as possibly new, while one species could not be placed in a genus.

Thomisidae: The thomisids are free-living spiders commonly found on grass, shrubs, flowers and trees, and only few species were sampled from the soil surface. Thomisids are easily dispersed by wind and most species have a wide distribution. In the LNR, 17 genera represented by 33 spp. were sampled. Of these, only 6 spp. (18.2%) are known South African endemics, while the rest (27 spp., 81.8%) are widely distributed throughout Africa.

Theridiidae: The theridiids are WB commonly making their webs on vegetation or close to the soil surface. Their webs are known as cob-webs or gumfoot-webs. Thirteen genera represented by 20 spp. have been sampled, but half of these species are data deficient, representing either new species or unidentifiable because of unresolved taxonomy and only 4 spp. (20.0%) are South African endemics.

Previously, general statements were made in placing species in functional groups or guilds, but with more detailed information becoming available through surveys in savanna where a wide variety of sampling methods were used, guild placement has improved considerably for some taxa (Dippenaar-Schoeman et al. 2009; Foord et al. 2008; Muelelwa et al. 2010). For the present study, two main guilds were recognised, namely wandering free-living spiders (W) and (WB), with further subdivisions based on micro-habitat and general behaviour, as observed during surveys (Online Appendix 1).

Wanderers: A total of 172 spp. (64.2%) are free-living spiders, with some species living on vegetation and others on the ground surface, and a few species occurring in both strata. The PW sampled from the grass and tree layer (Figure 3a–f) are represented by 90 spp. (33.6%). The Thomisidae (32 spp.), Salticidae (31 spp.) and Oxyopidae (10 spp.) were the most diverse plant dwellers found on grasses, shrubs and trees. Five spider species (1.9%) occur both on the ground and on vegetation and belong to Corinnidae and Salticidae (Online Appendix 1). A specimen in the genus Neotama (Hersillidae) were collected in the Church Forest and represents the first record of this genus outside the Kwa-Zulu Natal province (Foord & Dippenaar-Schoeman 2005).

The majority of ground dwellers (Figure 3g–j) are free-living soil dwellers (72 spp., 26.9%), not including those living in burrows (5 spp., 1.9%). Most of the burrow dwellers belong to the suborder Mygalomorphae and include three trapdoor spider species, Homostola vulpecula Simon, 1892 (Cyrtaucheniidae), Segregara transvaalensis (Hewitt, 1913) (Idiopidae) and Entypesa schoutedeni Benoit, 1965 (Nemesiidae), and a baboon spider species Harpactirella overdijki Gallon, 2010 (Theraphosidae). One species of wolf spider, H. adjacens, also constructs burrows. The Gnaphosidae (9 spp.) and Lycosidae (18 spp.) are the more diverse families of ground dwellers. Two of the ground dwellers are associated with fresh water, Nilus margaritatus (Pocock, 1898) and Nilus rossi (Pocock, 1902) (Pisauridae) and are commonly known as fish-eating spiders.

Web-dwellers: The web-dwellers (Figure 3k–o) are represented by 96 spp. (35.8%), with the largest number making orb-webs (57 spp., 19.0%), followed by gumfoot-webs (20 spp., 7.5%), retreat-webs (10 spp., 3.7%), sheet-webs (3 spp., 1.1%), space-webs (3 spp., 1.1%), funnel-webs (2 spp., 0.7%) and tube-webs (1 sp., 0.4%).

The physical structure of the habitat plays a role in the composition of the web-dwelling fauna, as it not only provides the necessary support for anchoring webs but also increases the availability of retreat space and modification of the microclimate, which could have an effect on spiders, as well as their prey. Several of the orb-web spiders recorded from the LNR, that is, Argiope (4 spp.), Caerostris (1 sp.), Gasteracantha (3 spp.), Isoxya (2 spp.), Nephilidae (3 spp.) and Tetragnathidae (9 spp.), make large orb-webs that they attach to trees and shrubs. Most of these species are diurnal and they are found in their webs during the day.

The LNR is water-rich, which is reflected in the number of web-dwelling tetragnathids, a group usually associated with water. This includes two species of Diphya, five species of the silver vlei spiders (Leucauge spp.) and the long-jawed water orb-web spiders (Tetragnatha boydi O.P. Cambridge, 1898). A large number (15 spp.) of OWB are associated with grasslands (Araneus, Kilima, Larinioides, Lipocrea and Neoscona). Most of these species are nocturnal, making their orb-webs at night and resting in retreats, usually constructed in grass inflorescences, during the day. Four species make modified orb-webs, ranging from the tropical tent-web (Cyrtophora) to the single line orb-webs of the Uloboridae (Miagrammopes). In the Malta Forest, the Dipluridae funnel-web spider, Thelechoris striatipes (Simon, 1889), was sampled from trees, where their webs covered the bark with the silk and were decorated with moss.