terblance 2.qxd The vegetation of three localities of the threatened butterfly species Chrysoritis aureus (Lepidoptera: Lycaenidae) R.F. TERBLANCHE, T.L. MORGENTHAL and S.S. CILLIERS Terblanche, R.F., T.L. Morgenthal and S.S. Cilliers. 2003. The vegetation of three local- ities of the threatened butterfly species Chrysoritis aureus (Lepidoptera: Lycaenidae). Koedoe 46(1): 73–90. Pretoria. ISSN 0075-6458. The vegetation and habitat characteristics of three localities of Chrysoritis aureus at the Alice Glockner Nature Reserve, Suikerbosrand Nature Reserve and Malanskraal farm near Heidelberg in South Africa, were compared. A numerical classification technique, TWINSPAN, was used and refined by using Braun Blanquet procedures to classify the vegetation at the different localities. A DCA ordination was applied to confirm the results of the classification. Although the general vegetation structure at the three habi- tats of Chrysoritis aureus were found to be similar, marked differences in the floristic composition were evidenced. A different sub-community, compared to the vegetation at Suikerbosrand and Alice Glockner Nature Reserve, was recorded at the Malanskraal habitat of Chrysoritis aureus. These differences in floristic composition, but with simi- larities in vegetation structure, indicate the possible importance of fire for the ultimate survival of these butterflies in the Rocky Highveld Grassland. The host plant of Chrysoritis aureus, Clutia pulchella, collected at Malanskraal differed markedly and consistently in their morphology, compared to the individuals from the habitats at Suikerbosrand and Alice Glockner Nature Reserve. These differences in the floristic composition of one of the habitats compared to the others, raise research questions con- cerning the butterfly metapopulation structure, since the subpopulations seem to be adapted to slightly different habitat conditions. The variation in the habitat suggests that the “last remaining locality scenario” for other localised butterflies in South Africa, such as Orachrysops niobe, needs to be redressed. Management strategies are addressed while the importance of conserving rare, localised ecosystems are highlighted by the phytosociological study. Key words: myrmecophilous, Lycaenidae, Chrysoritis, Clutia, Crematogaster, habitat, phytosociology, conservation R.F. Terblanche (drkrft@puknet.puk.ac.za), T.L. Morgenthal & S.S. Cilliers, School of Environmental Sciences and Development, Private Bag X6001, Potchefstroom Univer- sity for CHE, Potchefstroom, 2520 Republic of South Africa. ISSN 0075-6458 73 Koedoe 46/1 (2003) Introduction The butterfly Chrysoritis aureus—previously known as Poecilmitis aureus before the new combination was designated by Heath (1997) —is considered to be Rare in the Red Data book of butterflies (Henning & Henning 1989) and the updated list of the Red Data book of butterfly species (Henning & Hen- ning 1995). C.R. Barrett and F. Coetzee dis- covered Chrysoritis aureus on Christmas Day 1959 on a rocky slope below the water tower near the Natal road of Heidelberg in Gauteng (Pringle et al. 1994). During 1983, Chrysoritis aureus was added to the protect- ed wild animal list of the former Transvaal under Ordinance 12, since the type locality had been the only known area where the but- terfly ocurred (De Wet 1987). G.A. Henning found another locality near Greylingstad (Pringle et al. 1994). Members of the Lepi- dopterists’ Society of Africa found another locality in Suikerbosrand Nature Reserve (Henning & Henning 1989) after teams of terblance 2.qxd 2005/12/09 10:31 Page 73 lepidopterists carried out organised exten- sive exploration in this reserve (Fig. 1). The colony at Suikerbosrand seems to be limited in numbers (Henning & Henning 1989; Roos & Henning 2000) though the numbers tend to remain stable (The word ‘colony’ refers to the subpopulations of these butterflies, that is where they actually breed. The adults usu- ally also fly at or close to these restricted areas). In the Suikerbosrand Nature Reserve, the only recorded habitat of the butterfly is limited to the north eastern extreme of the reserve, even though the Lepidopterists’s Society of Africa organised surveys that cov- ered many different habitats in the reserve (G.A. Henning 1986a, 1986b, 1987). After decades of exploring by a number of lepi- dopterists in the hilly areas of Gauteng and the adjacent Mpumalanga (Table 1), only five confirmed localities of Chrysoritis aureus were reported. The distribution of the butterfly can be described as patchy and Koedoe 46/1 (2003) 74 ISSN 0075-6458 Table 1 The locality, altitude and geological system present at the confirmed localities of Chrysoritis aureus Locality Province Locality * Altitude (m) Geological Supergroup Alice Glockner NR 8 km S Heidelberg Gauteng 26º34'S, 28º22'E(2628CB) 1690 Witwatersrand Hill N of Greylingstad Mpumalanga 26º44'S, 28º44'E(2628DA) 1810 Ventersdorp Malanskraal locality 14 km W Balfour Mpumalanga 26º39'S, 28º26'E(2628CB) 1780 Ventersdorp Suikerbosrand NR Gauteng 26º29'S, 28º18'E(2628AD) 1720 Witwatersrand Type locality at Heidelberg Gauteng 26º29'S, 28º21'E(2628AD) 1600 Witwatersrand * The grid reference reading is from the north western corner of each minute by minute square. **The Malanskraal locality is close to the borderline between the Ventersdorp supergroup and the Witwatersrand Supergroup. Fig. 1 The location of the habitats of Chrysoritis aureus that were studied. terblance 2.qxd 2005/12/09 10:31 Page 74 Chrysoritis aureus is very localised at each of these localities. The three habitats studied each cover approximately 200 m² – 400 m². All the recorded localities of Chrysoritis aureus are situated at sites where formal or informal urban development pose an ever- increasing threat to the habitat. Habitat destruction across the world as the main cause for the extinction of populations of localised butterflies, is well recognised, thereby emphasising the primary importance of habitat conservation (S.F. Henning 1987; Larsen 1995; Kudrna 1995; Munguira 1995; New 1995; Oates 1995; Opler 1995; Pullin et al. 1995; Thomas 1995; Warren 1995). An ongoing comparative study of the Chrysori- tis aureus habitats (vegetation) is therefore imperative to gain knowledge for develop- ing strategies for the conservation of these habitats. A description of the vegetation of the Chrysoritis aureus habitats is needed if management practices such as translocation of butterflies or habitat restoration is consid- ered. Vegetation studies of the habitats of the threatened Golden Sun Moth in Australia (O’Dwyer & Attiwill 2000), provided essen- tial information for the restoration of its habitat. A vegetation analysis of the then only known habitat of a threatened butterfly, Aloeides dentatis subsp. dentatis, in South Africa, (Deutschländer & Bredenkamp 1999) was very applicable for its conserva- tion management. For this study, the vegeta- tion of three habitats were compared to iden- tify possible key factors and to detect possi- ble differences between these habitats. This study also forms part of an ongoing investigation into the reasons why the sub- populations of the butterfly are confined to such small areas. In South Africa, studies on myrmecophilous butterflies, such as Chrysoritis aureus, are limited to life histo- ry descriptions or reporting of information on the host plant and host ant. Some clues for the confinement of Chrysoritis aureus to its habitats can be found in its life history (Henning 1983). The host plant of the larvae is the small shrub Clutia pulchella L. Sens. lat. (Euphorbiaceae) which is widely distrib- uted in South Africa (Arnold & De Wet ISSN 0075-6458 75 Koedoe 46/1 (2003) Fig. 2. Clutia pulchella ("Heidelberg morpho- type"), the host plant of Chrysoritis aureus. 1993; Retief & Herman 1997) (Fig. 2). The larvae of Chrysoritis aureus are also con- stantly tended by ants of the genus Cremato- gaster (Henning 1983). During the day, the larvae shelter for protection and only appear at night to feed on their host plants (Henning 1987a). Well developed dorsal nectary organs (Cottrell 1984) or honey glands (Hen- ning 1983; 1987a) are present on the larvae and the ants are attracted to feed on the secre- tions of these organs (Henning 1983, 1987a). In the absence of the ants, a fungal infection occurs in the dorsal nectary organ and the larvae often die after a few days (Henning 1983) (Fig. 3). Since both the ant and the host plant are needed for survival of Chrysoritis aureus at any locality, the habitat requirements are complicated. These include specific local vegetation communities. We therefore investigated the vegetation, to con- terblance 2.qxd 2005/12/09 10:31 Page 75 tribute to a niche/habitat profile of Chrysori- tis aureus. Study area Three localities inhabited by Chrysoritis aureus were studied: one at the Suiker- bosrand Nature Reserve, one at the Alice Glockner Nature Reserve near Heidelberg, Gauteng and one at the farm Malanskraal (Fig. 1). The habitats of Chrysoritis aureus are situated on the south-facing slopes of rocky ridges and hills. The first two habitats studied are situated on the quartzites of the Wiwatersrand Supergroup while the habitat at Malanskraal habitat is situated on the lavas of the Ventersdorp Supergroup. The landscape is rugged and broken in appear- ance, since the Witwatersrand System was subjected to considerable folding, faulting and erosion in the past (Du Toit 1954). The study sites are in the summer rainfall area, with temperate summers and cold Koedoe 46/1 (2003) 76 ISSN 0075-6458 Fig. 3. The larva of Chrysoritis aureus. The two white structures at the anterior (head) area, are tubercle organs. Note the fungal growth at the posterior end at the dorsal nectary organ. Photo: R.F. Terblanche. frosty winters. According to information obtained from the South African Weather Service, the long-term average annual rain- fall data from nearby weather stations vary considerably between 620-750 mm per annum (South African Weather Service, Pri- vate Bag X097, Pretoria, 0001). Variation in topography, the rock cover and vegetation structure results in different meso- and microclimates. The sites that were investigated are situated in the Grassland Biome (Rutherford & Westfall 1994; Low & Rebelo 1996), the Bankenveld (Acocks 1988) or Rocky High- veld Grassland (Bredenkamp & Van Rooyen 1996). Materials and methods Sample plots (5 m x 5 m) were placed on the habi- tats of Chrysoritis aureus. These plots fitted well on the restricted habitats of Chrysoritis aureus at the Alice Glockner Nature Reserve and Malanskraal, while four sample plots were adequate for the habi- tat at Suikerbosrand. Since the range of occurence of Chrysoritis aureus is so small, the samples are thought to be representative. A transect of sample plots over the habitat from the summit to the mid- slopes and down to a valley bottom, was surveyed at the Alice Glockner Nature Reserve to compare the vegetation on the slopes close to the habitats with the habitats of Chrysoritis aureus (Fig. 4). The floristic composition and cover abundance accord- ing to the Braun-Blanquet scale (Mueller-Dombois & Ellenberg 1974) were noted in each plot during surveys on 12 and 13 February 1999. The number of Clutia pulchella individuals that occurred in each sample plot were determined. Plant specimens were deposited in the A.P. Goossens Herbarium of the Potchefstroom University for C.H.E. The height of trees (Von Breitenbach 1990; Von Breitenbach et al. 2001) was noted. These species will, therefore, be referred to as trees, even if they are physionomical- ly not regarded as trees. The height classes as described by Edwards (1983) were used. Habitat features such as slope, aspect and rockiness were noted, though not quantitatively measured. Notes on the presence of Chrysoritis aureus and other butter- flies were continuously made by means of direct observations on 7 and 13 November, 1998 as well as 12 and 13 February 1999. The local distribution of Chrysoritis aureus was confirmed by P. Roos and G.A. Henning who have visited the localities for a terblance 2.qxd 2005/12/09 10:31 Page 76 number of years (Roos & Henning 2000). The same procedure was also followed for the breeding sites, i.e., where pupae, larvae, and oviposition by the female were evidenced. The ants associated with Chrysoritis aureus have been identified by Dr. H.G. Robertson, South African Museum, Cape Town, 8000. The TWINSPAN classification algorithm (Hill 1979a) was used to analyse the floristic data and the result was refined by using Braun Blanquet proce- dures by means of the BBPC-suite (Bezuidenhout et al. 1996). A DCA ordination was applied (Hill 1979b) to confirm the results of the classification. Results and discussion The host plant Mr. I. Bampton discovered the host plant of Chrysoritis aureus in January 1976 (Owen- Johnston 1991). It is sometimes reported to be Clutia galpinii (Owen-Johnston 1991; Heath 2001). Clutia galpinii is considered to be a synonym of Clutia pulchella (Retief & Herman 1997). The entity Clutia galpinii refers to a relatively smaller plant of 91 cm – 152 cm, whereas Clutia pulchella reaches heights up to 244 cm (Prain 1925). Clutia pulchella individuals that have been mea- sured near or at the habitats of Chrysoritis aureus, ranged from 13 cm to more than 200 cm tall (in more dense vegetation) and fits the descriptions of both entities overviewed by Prain (1925). The plant specimens from the localities of Chrysoritis aureus were compared to the material at the National Herbarium. It does not fit Clutia monticola of which the leaves are sessile (Retief & Herman 1997). The host plant of Chrysoritis aureus was confirmed to be only Clutia pulchella at all the localities. Clutia pulchella constantly showed specific variation from different localities. Individu- als from Malanskraal have more lanceolate leaves, are of a less bluish-green colour and the upper surface of the leaves are covered with velvety hairs. This is in contrast to the more ovate, more glaucous and much less hairy leaves of individuals from the Alice Glockner Nature Reserve and Suikerbosrand Nature Reserve. The reasons for these differ- ences could be due to the different geologi- cal systems where these plants occur. These morphotypes may be the subject of more detailed study on the taxonomic status of the host plant. Larvae of the different localities ISSN 0075-6458 77 Koedoe 46/1 (2003) Fig. 4. Position of the 5m x 5m sample plots from the summit (AC 14) to the lower valley bot- tom (AC 8, AC 9) of the hill (southern aspect) at Alice Glockner Nature Reserve. The seven plots (AC 1–AC 7) that are inhabited by Chrysoritis aureus were put out horizontally next to each other along the rocky ridge. terblance 2.qxd 2005/12/09 10:31 Page 77 seems to be adapted to feed on these differ- ent morphotypes of Clutia pulchella. The number of Clutia pulchella individuals at the sample plots where Chrysoritis aureus is present varies (2-14 indiv/25m²) with an average of 7.8 individuals/25m² for all the habitats combined—that means an average of 0.31 Clutia pulchella indiv/m². The local- ity at the Alice Glockner Nature Reserve showed the lowest average of Clutia pul- chella individuals namely 6.7 indiv/25m². Though Chrysoritis aureus was not present at sample plots AC 10, AC 11 and AC 14, host plant individuals were recorded at these sites. The number of host plants at sample plots AC 10 and AC 11 (Table 2) seems to be adequate to serve as possible breeding areas. All the sites where Clutia pulchella individ- Koedoe 46/1 (2003) 78 ISSN 0075-6458 Table 2 The number of Clutia pulchella counted as well as the status of Chrysoritis aureus at the plots (5m x 5m) Plot Number of Status of number Clutia pulchella C. aureus individuals 1 = present 0 = absent AC 1 4 1 AC 2 3 1 AC 3 4 1 AC 4 7 1 AC 5 13 1 AC 6 14 1 AC 7 2 1 AC 8 0 0 AC 9 0 0 AC 10 5 0 AC 11 4 0 AC 12 0 0 AC 13 0 0 AC 14 2 0 MK 15 7 1 MK 16 8 1 MK 17 10 1 SB 18 6 1 SB 19 10 1 SB 20 8 1 SB 21 13 1 uals were present are characterised by the presence of large rocks and this seems to be one of the key habitat requirements of the host plant. The factors that can play a role in excluding Chrysoritis aureus from establish- ing at sample plots AC 10 and AC 11, are not known. However, both sample plots AC 10 and AC 11 are situated on the lower mid- slopes compared to the other sites where Clutia pulchella occurs at higher midslopes in the Alice Glockner Nature Reserve (Fig. 4). Another subject for further research is the local distribution and vitality of Clutia pul- chella in the vicinity of Chrysoritis aureus habitats. It was observed that Clutia pulchel- la may have higher densities and seemed to have better vitality (higher plants) in some of the surrounding areas, where no plots were put out, as Chrysoritis aureus was not pre- sent here. A quantitative survey may shed more light on the above observation. Vegetation classification The floristic composition of the plant com- munities identified is represented in a phyto- sociological table (Table 3) and the results of the DCA-ordination (Fig. 5). The phytosoci- ological analysis resulted in the following two communities with their sub-communi- ties and variants: 1. Melinis nerviglumis-Aristida trans- vaalensis Community 1.1. Melinis nerviglumis-Aristida trans- vaalensis–Eragrostis curvula Sub- community 1.2 Melinis nerviglumis-Aristida trans- vaalensis-Loudetia simplex Sub-com- munity 1.2.1. Anthospermum hispidulum Variant 1.2.2 Digitaria monodactyla Variant 1.3. Melinis nerviglumis-Aristida trans- vaalensis-Englerophytum magalis- montanum Sub-community 2. Elionurus muticus-Eragrostis trichopho- ra Community terblance 2.qxd 2005/12/09 10:31 Page 78 Community description 1. Melinis nerviglumis - Aristida transvaalensis Community This community is found on the south-facing slopes and on the terraces within the slopes, and includes a representation of all the habi- tats of Chrysoritis aureus that were studied. Apart from the terraces, large rocks (often exceeding one metre in length) are conspicu- ous on all the sites. The community is char- acterised and dominated by the grasses Meli- nis nerviglumis, Aristida transvaalensis and Cymbopogon plurinodis (Species group A, Table 3). Three subcommunities are recog- nised: 1.1 Melinis nerviglumis-Aristida trans- vaalensis -Eragrostis curvula Sub- community This sub-community occurs on the south- facing slopes at Malanskraal on the Venters- dorp Supergroup (Table 1). It is charac- terised by the absence of Loudetia simplex and the constant high cover-abundance values of Eragrostis curvula, Helichrysum nudifolium and the constant presence of Garuleum pinnatifidum and Tephrosia capensis (Species group B, Table 3). This sub-community is further characterised by a number of plant species with a low constan- cy which are exclusive to this subcommuni- ty. These include the forbs Calpurnia glabrata, Pelargonium sidoides and the dwarf shrub Rhus discolor. Three tree species, namely Leucosidea sericea, Buddle- ja salviifolia and Gymnosporia buxifolia were recorded in this sub-community with a low cover-abundance and constancy. None of these tree species exceeded 2 m in height and therefore the woody layer is represented only by shrubs lower than 2 m. The Malans- kraal habitat of Chrysoritis aureus is also the exclusive habitat to this sub-community. ISSN 0075-6458 79 Koedoe 46/1 (2003) Fig. 5. DCA-ordination of the sample plots at the Alice Glockner Nature Reserve (AC), Malanskraal (MK) and Suikerbosrand Nature Reserve (SBR). terblance 2.qxd 2005/12/09 10:31 Page 79 Koedoe 46/1 (2003) 80 ISSN 0075-6458 Table 3 A phytosociological table of the surveys done at the Alice Glockner Nature Reserve, Malanskraal and Suikerbosrand terblance 2.qxd 2005/12/09 10:31 Page 80 ISSN 0075-6458 81 Koedoe 46/1 (2003) Table 3 (continued) terblance 2.qxd 2005/12/09 10:31 Page 81 Koedoe 46/1 (2003) 82 ISSN 0075-6458 Table 3 (continued) terblance 2.qxd 2005/12/09 10:31 Page 82 ISSN 0075-6458 83 Koedoe 46/1 (2003) Table 3 (continued) 1.2. Melinis nerviglumis-Aristida trans- vaalensis-Loudetia simplex Sub-com- munity This sub-community was recorded on the quartzite slopes of Alice Glockner and Suikerbosrand on the Witwatersrand Super- group (Table 1). The diagnostic species include the grasses Loudetia simplex, Diheteropogon amplectens, Tristachya leu- cothrix and the forbs Vernonia staehelinoides and Senecio venosus. (Species group C, Table 3). It is interesting to note that the forbs Ver- nonia staehelinoides and Senecio venosus (Species group C, Table 3) are absent from the sites of all the other communities, but are found scattered throughout this community. This community can be divided into two variants: 1.2.1 Anthospermum hispidulum Variant This variant is recorded from the south- facing rocky slopes of the Alice Glockner Nature Reserve and the Suikerbosrand Nature Reserve. The constant presence of Anthospermum hispidulum characterises this variant while Clutia pulchella occurs abun- dantly on most of the sites that were surveyed in this community. The bracken fern Pteridi- um aquilinum was only found at the Suiker- terblance 2.qxd 2005/12/09 10:31 Page 83 Koedoe 46/1 (2003) 84 ISSN 0075-6458 bosrand habitat (Table 3) which at first glance gave the Suikerbosrand habitat a dif- ferent appearance. The similarities with the habitats at Alice Glockner Nature Reserve were only revealed after the vegetation clas- sification. Almost no tree species were found at all these sites. The woody layer is repre- sented by Canthium gilfillanii, Myrsine africana and Diospyros lyciodes subsp. guerkei (Table 3). These woody species occurred sporadically, had low cover-abun- dance values, and none were taller than 1 m. Chrysoritis aureus occurs on most of these sites, excluding sites AC 10 and AC 11 that are present on the lower midslopes (Fig..4). Of interest is that Chrysoritis lycegenes, a closely related species of Chrysoritis aureus, uses Myrsine africana as host plant (Hen- ning 1983). Chrysoritis lycegenes, however, is mainly restricted to central Kwa-Zulu Natal and the eastern Free State. 1.2.2 Digitaria monodactyla Variant This variant was observed on two flat ter- races on the slopes at Alice Glockner Nature Reserve (Fig. 4). The diagnostic species include the grasses Aristida aequiglumis, Digitaria monodatyla and Schizachyrium sanguineum (Table 3, Species Group E). The terraces are narrow and form a mini-plateau, whereas the huge exposed rocks that occur elsewhere on the slopes are absent. Chrysoritis aureus does not colonise these flat terraces, and this coincides with the absence of the host plant Clutia pulchella (Table 2, Table 3). 1.3 Melinis nerviglumis-Aristida trans- vaalensis-Englerophytum magalis- montanum Sub-community The summit of the ridge at the Alice Glock- ner Nature Reserve is the habitat of this sub- community which is dominated by shrubs rather than grasses and forbs. A small patch of vegetation, measuring only approximately 10 m x 10 m, which differs physiognomical- ly from the surrounding areas, is present. The diagnostic species is the shrub Englero- phytum magalismontanum. Other dominant shrubs present are Myrsine africana, Ancy- lobotrys capensis and Rhus magalismon- tana. No Chrysoritis aureus occur at this site since the habitat is not suitable. A num- ber of other butterflies namely Coeliades forestan, Junonia archesia, Junonia oenone, Cigaritis natalensis and Anthene butleri subsp. livida showed typical hilltopping behaviour at this site. Hilltopping indicates a type of territorial behaviour where males of butterfly species gather at tops of hills, often at midday—individuals will choose a site which they will defend by flying after “intruders”. No hilltopping of Chrysoritis aureus was observed. 2. Elionurus muticus – Eragrostis trichophora Community. This community is found in the valley bot- tom on the southern hillside below the rocky ridges at the Alice Glockner Nature Reserve. It is characterised by the diagnostic species namely the grasses Elionurus muticus, Era- grostis trichophora, Digitaria tri- cholaenoides and Heteropogon contortus as well as the forbs Indigofera comosa and Ver- nonia oligocephala (Species Group G, Table 3). Very little exposed rocks occur in this area and the plant species cover abun- dance is relatively higher than in all the other sample plots. Chrysoritis aureus is absent in this community. DCA-ordination The results of the DCA-ordination (Fig. 4) support the classification of the communities identified from the phytosociological table. The difference between the floristic compo- sition of the habitats of Chrysoritis aureus at Malanskraal and those at Suikerbosrand Nature Reserve and Alice Glockner Nature Reserve, is confirmed by the results of the DCA ordination – a topographical gradient from the valley bottom to the upper slopes are reflected by the configuration of the ordi- nation. The position of the sample plots on ordination axis two can be related to the geology, with habitats on the Ventersdorp Lava Supergroup at the top and habitats on terblance 2.qxd 2005/12/09 10:31 Page 84 ISSN 0075-6458 85 Koedoe 46/1 (2003) the Witwatersrand Supergroup in the middle and at the bottom. The discontinuous posi- tion of AC 14 on the scatter diagram may partly be due to its location on the summit of a ridge in contrast with all the other sample plots that are on south-facing slopes. The Elionurus muticus-Eragrostis tri- chophora community is at the left extreme along ordination axis 1 of the scatter diagram and the Melinis nerviglumis-Aristida trans- vaalensis community (Community 1.1., 1.2 and 1.3) to the middle and right of the scat- ter diagram. The Melinis nerviglumis-Aristi- da transvaalensis-Eragrostis curvula sub- community (Community 2) that occurs at Malanskraal is situated at the top right end of the DCA scatter diagram and the Melinis nerviglumis-Aristida transvaalensis-Loude- tia simplex sub-community in the middle. Most of the sample plots, of which the plant species composition, geology and topogra- phy favour the presence of Chrysoritis aureus, are to the middle-right and top end of the scatter diagram. This confirms the impor- tance of open vegetation on south-facing rocky upper slopes for the survival of the butterfly. Discussion The habitat of Chrysoritis aureus Table 4 provides a matrix in which some of the important environmental habitat charac- teristics that became apparent in this study, are listed. The altitude at the investigated habitats varies between 1690 m–1780 m, Table 4 A table with specified habitat factors/ observations that apply to a relevé (= 1) or not (= 0). If a num- ber 1 is in bold it indicates that the presence of Chrysoritis aureus coincides with the factor. The pres- ence of Chrysoritis aureus is indicated in the first row of the matrix terblance 2.qxd 2005/12/09 10:31 Page 85 Koedoe 46/1 (2003) 86 ISSN 0075-6458 indicating that the butterfly only occurs in a certain climate regime (Table 2). The butter- fly colonies are only found at locations that are covered with large rocks. These rocks seem to be a habitat requirement for the host plant. In addition, steep upper midslopes seem to be essential even on the most local scale (the butterfly does not breed on the ter- races). Only the cool southern-facing slopes are chosen by the colonies of the butterflies for reasons which are not yet clear. This small-scale study showed that Chrysoritis aureus is indeed specific on a very local scale with regard to its choice of habitat. Two related sub-communities with different floristic compositions, but similar vegetation structures were recorded as habitats of Chrysoritis aureus. The habitats of Chrysori- tis aureus are dominated by grass species though a high forb diversity was observed. Many of these forbs are perennials that can survive fire. These include Clutia pulchella (the host plant) as well as nectar plants for the adult butterfly, such as Tetraselago wilm- sii (Species-group D; Table 3). Despite dense stands of trees or shrubs observed to be growing on the same hill of each of the local- ities studied, the tree component was almost absent where Chrysoritis aureus is present. Since dense and relatively diverse communi- ties of trees and shrubs are particular to the Rocky Highveld Grassland (Bredenkamp & Van Rooyen 1996)—compared to other types of grassland—it is believed that the more open vegetation noted at the habitats of Chrysoritis aureus are important for conser- vation management. Fire and frost are two factors that probably play a key role in the maintenance of a suitable habitat for Chrysoritis aureus since both are needed to maintain the grasslands in these Rocky Highveld Grassland areas. Acocks (1988) interpreted Bankenveld (now included in the Rocky Highveld Grassland) as a fire-maintained false grassland which would develop into savanna if fire was excluded. The vegetation of the Suikerbos- rand Nature Reserve that was studied by Bredenkamp & Theron (1978, 1980) pro- vides a further context within which the veg- etation studies at the Chrysoritis aureus habitat at Suikerbosrand could be seen. Bre- denkamp & Theron (1978) described a Pro- tea caffra–Helichrysum setosum savanna community that is also found on rocky steep southern slopes of the Witwatersrand Super- group in the Suikerbosrand Nature Reserve. A Protea caffra-Clutia pulchella variant was also distinguished by Bredenkamp & Theron (1978). An open tree stratum was observed in this variant consisting of a number of trees of which Protea caffra was the most con- spicuous (Bredenkamp & Theron 1978). No Chrysoritis aureus were found in the past (including surveys by members of the Lepi- dopterists’ Society) in the vegetation where Protea caffra individuals are an important component of the vegetation. The absence of Protea caffra and any tree layer at all the sites where Chrysoritis aureus occur is therefore a confirmation of the more open nature of the vegetation in the habitats of the butterfly. These open patches in the locations where Clutia pulchella occur and which are concomitant with an absence of Protea caf- fra seems to be localised, and perhaps mar- ginal. A larger study area should reveal more information about the significance of the absence of a tree stratum at the sites where Chrysoritis aureus occurs. Fire may play an important role in the regulation of the Chrysoritis aureus habitats. The timing and intensity of fires in the vegetation among the rocky ridges where Clutia pulchella grow should be continually monitored to promote the survival of Chrysoritis aureus. Frost is also mentioned in literature to play an important role in the distribution of the woody elements in the Rocky Highveld Grassland, as mentioned by Bredenkamp & Van Rooyen (1996). Grassland vegetation in the Rocky Highveld Grassland is restricted to exposed cooler sites (Bredenkamp & Van Rooyen 1996). All the habitats of Chrysori- tis aureus are at exposed sites below the summit of these hills (on the upper slopes). Chrysoritis aureus have not been found in any of the kloofs in these areas where large numbers of Clutia pulchella were observed, though unfortunately no sample plots were placed there for quantitative comparisons. terblance 2.qxd 2005/12/09 10:31 Page 86 ISSN 0075-6458 87 Koedoe 46/1 (2003) Since Chrysoritis aureus is also associated with an ant species, Crematogaster liengmei, the structure of the vegetation and the role of factors such as fire may be part of the spe- cific habitat requirements of the ants. The present ant community studies may reveal more insight into these habitat requirements, especially against the background of the veg- etation analysis provided in this study. The open nature of the vegetation at the habitats of Chrysoritis aureus has further implications for the control of invasive plants in these areas. Cover-abundances of the Pteridium aquilinum (bracken fern) stands at the Chrysoritis aureus habitat at the Suikerbosrand Nature Reserve should be monitored, since the bracken fern can be regarded as invasive. This habitat occurs at the border fence of the reserve and is regu- larly burnt as part of a fire break. The best method to control bracken, if deemed neces- sary, is by cutting (Marrs et al. 1998). Brack- en ferns possess a number of underground rhizomes and are well adapted to fire. Oinonen (1967) found that bracken estab- lished best from spores that are present on sites that had been burnt. Another consideration for the management at all the sites is to prevent bush encroach- ment and the establishment of any exotic tree species, since this will clearly destroy the localised habitats. Continued control of the invasive Acacia mearnsii, which is a declared invader (with commercial value) (Bromilow 2001; Henderson 2001) at the Alice Glockner Nature Reserve, is impera- tive since this will save the butterfly habitats from a possible catastrophe in future. Conclusion The vegetation at the three localities where Chrysoritis aureus is present, is similar with regard to general structure, but there are dif- ferences in species composition on the sub- community level. A remarkable coincidence in the Malanskraal habitats, being a different sub-community, is that the Clutia pulchella individuals collected there also differ markedly from individuals from the locali- ties at Alice Glockner Nature Reserve and the Suikerbosrand Nature Reserve. The Malanskraal habitat is also on the Venters- dorp Supergroup in contrast with the habitats at the Alice Glockner Nature Reserve and Suikerbosrand Nature Reserve. An hypothe- sis to be tested is that the genetic make-up of the Chrysoritis aureus populations at the three localities are different, since they seem to be adapted to habitats that differ to some extent. These findings would be valuable for the conservation of Chrysoritis aureus if translocation of the butterfly or restoration of habitats are considered. Securing corridors that link habitats is an aspect that deserves attention, especially in the sense that these must include adequate nectar sources. The value of a phytosociological survey to support conservation research and manage- ment strategies for a threatened butterfly are confirmed by this study and by Deutschlän- der & Bredenkamp (1999). The vegetation structure would be an important aspect to consider in the management of areas where Chrysoritis aureus occurs and could even be studied in more detail on a larger scale. Due to the open nature of the vegetation and localised distribution of the habitats of Chrysoritis aureus the application of fire- management should be applied with care while bush encroachment, especially of any tree layer, should be avoided. Alien vegeta- tion in these areas should also be monitored since the establishment of exotic tree species will destroy the habitats of Chrysoritis aureus. Further research priorities that were identi- fied, include a study of ant communities locally at the habitats of Chrysoritis aureus. To study the nectar sources of Chrysoritis aureus in more detail a larger study area than the one used here, would be needed since it seems that the butterfly occasionally wan- ders of from its central breeding ground in search of nectar. Larger sample plots (100 m²) and a larger study area should therefore be included for such studies. For the studies on the optimal density of the host terblance 2.qxd 2005/12/09 10:31 Page 87 Koedoe 46/1 (2003) 88 ISSN 0075-6458 plant, aspects other than southern aspects, should also be included. It is hoped that more and diverse research on the rare and localised butterflies of Africa can be continued since it is believed that a holistic approach in the research will lead to more discoveries with regard to the intrigued habitat requirements of these organisms. It is recommended that the various sites where Chrysoritis aureus occurs, should receive priority for conservation purposes. In the light of the variation in the habitats of Chrysoritis aureus encountered during this study, the securing of one habitat for the con- servation of a local habitat-specific butterfly might be a risk. The implications for butter- flies such as Orachrysops niobe (Brenton Blue), where only one locality is known, is that such butterflies might be in a precarious situation even if the remaining habitat is con- served, since the meta-population dynamics may be limited. The threat of urbanisation to the conserva- tion of the Rocky Highveld Grassland in general is mentioned by Bredenkamp & Van Rooyen (1996) and it seems to be prophetic regarding the conservation of Chrysoritis aureus. Rocky Highveld Grassland is poorly conserved (Bredenkamp & Van Rooyen 1996) and if the number and extent of con- servation areas are considered the develop- ment and conservation of the Alice Glockner Nature Reserve will prove to be an important asset for biodiversity conservation and eco- tourism in South Africa. Acknowledgements We wish to thank Henning & Roos CC as well as Michelle Pfab and Marianne Forsyth (Gauteng Department of Agriculture, Conservation and Envi- ronment) for their professional support and assis- tance. This research would not have been possible if it had not been for the financial support of the Gaut- eng Department of Agriculture, Conservation and Environment, via Roos & Henning CC. References ACOCKS, J.P.H. 1988. Veld types of South Africa. Memoirs of the botanical Survey of South Africa No. 57. ARNOLD, T.H. & B.C. DE WET (eds). 1993. Plants of Southern Africa: names and distribution. Memoirs of the botanical Survey of South Africa No. 62. BEZUIDENHOUT, H., H.C. BIGGS & G.J. BREDENKAMP. 1996. A process supported by the utility BBPC for analysing Braun-Blanquet data on a personal computor. Koedoe 39(1): 107-112. BREDENKAMP, G.J. & G.K. THERON. 1978. A syne- cological account of the Suikerbosrand Nature Reserve I. The phytosociology of the Witwaters- rand Geological System. Bothalia 12(3): 513- 529. BREDENKAMP, G.J. & G.K. THERON. 1980. A syne- cological account of the Suikerbosrand Nature Reserve II. The phytosociology of the Venters- dorp Geological System. Bothalia 13(1&2): 199-216. BREDENKAMP, G.J. & N. VAN ROOYEN. 1996. Rocky Highveld Grassland. In: LOW, A.B. & A.G. REBELO (eds.). Vegetation of South Africa, Lesotho and Swaziland. Pretoria: Department of Environmental Affairs & Tourism. BROMILOW, C. 2001. Problem plants of South Africa. Pretoria: Briza. COTTRELL, C.B. 1984. Aphytophagy in butterflies: its relationships to myrmecophily. Zoological Journal of the Linnean Society 79: 1-57. DEUTSCHLÄNDER, M.S. & C.J. BREDENKAMP. 1999. Importance of vegetation analysis in the conser- vation management of the endangered butterfly Aloeides dentatis subsp. dentatis (Swierstra) (Lepidoptera: Lycaenidae). Koedoe 42(2): 1-12. DE WET, K. 1987. Die bewaring van die skoenlap- pers van die genus Charaxes en van Erikssonia acraeina in die Transvaal. Proceedings of the first Lepidoptera conservation Symposium, Roodepoort: Lepidopterists' Society of southern Africa: 5-7. DU TOIT, A.C. 1954. The geology of South Africa Edinburgh: Oliver & Boyd. EDWARDS, D. 1983. A broad-scale structural classi- fication of vegetation for practical purposes. Bothalia 14(3): 705-712. HEATH, A. 1997. A review of African genera of the tribe Aphnaeini (Lepidoptera: Lycaenidae). Metamorphosis: Occasional Supplement No. 2: 1-60. HEATH, A. 2001. New synonymies and taxonomic notes on the genus Chrysoritis Butler (Lepi- doptera: Lycaenidae). Metamorphosis 12 (3): 85-98. HENDERSON, L. 2001. Alien weeds and invasive plants: a complete guide to declared weeds and invaders in South Africa. Cape Town: Agricul- terblance 2.qxd 2005/12/09 10:31 Page 88 ISSN 0075-6458 89 Koedoe 46/1 (2003) tural Research Council. (Plant Protection Research Institute Handbook No. 12.) Henning, G.A. 1986a. Research at Suikerbosrand Nature Reserve, 7 September 1986. Transvaal Nature Conservation. Unpublished report. HENNING, G.A. 1986b. Research at Suikerbosrand Nature Reserve, 2 December 1986. Transvaal Nature Conservation. Unpublished report HENNING, G.A. 1987. Research at Suikerbosrand Nature Reserve, 1st quarter 1987. Transvaal Nature Conservation. Unpublished report. HENNING, G.A. & S.F. HENNING. 1995. Updating the status of South African Red Data Butterfly species. Metamorphosis 6(2): 96-98. HENNING, S.F. 1983. Biological groups within the Lycaenidae (Lepidoptera). Journal of the Ento- mological Society of southern Africa 46(1): 65- 85. HENNING, S.F. 1987a. Myrmecophilous Lycaenidae (or how ants help butterflies). South African Journal of Science 83: 8-9. HENNING, S.F. 1987b. Outline of Lepidoptera con- servation with special reference to ant associat- ed Lycaenidae. Proceedings of the first Lepi- doptera conservation Symposium, Roodepoort: Lepidopterists' Society of southern Africa: 5-7. HENNING, S.F. & G.A. HENNING. 1989. South African red data book: butterflies. Pretoria: Council for Scientific and Industrial Research, National Scien- tific Programmes Unit. (South African National Scientific Programmes report; no. 158.) HILL, M.O. 1979a. TWINSPAN. A Fortran programme for arranging multivariate data in an ordered two way table by classification of individuals and attributes. Ithaca, New York: Cornell Uni- versity. HILL, M.O. 1979b. DECORANA. A FORTRAN pro- gramme for detrended correspondence analysis and reciprocal averaging. Ithaca, New York: Cornell University. KUDRNA, O. 1995. Conservation of butterflies in central Europe. Pp. 248 - 257. In: PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. LARSEN, T.B. 1995. Butterfly biodiversity and con- servation in the Afrotropical region. Pp. 290- 303. In: PULLIN, A.S. (ed.). Ecology and conser- vation of butterflies. London: Chapman & Hall. LOW, A.B. & A.G. REBELO. (eds.). 1996. Vegetation of South Africa, Lesotho and Swaziland. Preto- ria: Department of Environmental Affairs & Tourism. MARRS, R.H., S.W. JOHNSON & M.G. LE DUC. 1998. Control of bracken and restoration of heathland VII. The response of bracken rhizomes to 18 years of continued bracken control or 6 years of control followed by recovery. Journal of Applied Ecology 35: 748-757. MUELLER-DOMBOIS, D. & H. ELLENBERG. 1974. Aims and methods of vegetation ecology. New York: Wiley. Munguira, M.L. 1995. Conservation of butterfy habitats and diversity in European Mediter- ranean countries. Pp. 277- 289. In PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. New, T.R. 1995. Butterfly conservation in Australa- sia - an emerging awareness and an increasing need. Pp. 304 - 315. In PULLIN, A.S. (ed.). Ecol- ogy and conservation of butterflies. London: Chapman & Hall. OATES, M.R. 1995. Butterfly conservation within the management of grassland habitats. Pp. 98-112. In: PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. O'DWYER, C. & P.M. ATTIWILL. 2000. Restoration of a native grassland as habitat for the Golden Sun Moth Synemon plana Walker (Lepidoptera: Castniidae) at Mount Piper, Australia. Restora- tion Ecology 8 (2):170-174. OINONEN, E. 1967. Sporal regeneration of bracken (Pteridium aquilinum) (L. Kuhn.) in Finland in the light of the dimensions and age of its clones. Acta Forestalia Fennica 83(2): 1-51. OPLER, P.A. 1995. Conservation and management of butterfly diversity in North America. Pp. 316 - 324. In: PULLIN, A.S. (ed.). Ecology and con- servation of butterflies. London: Chapman & Hall. OWEN-JOHNSTON, N.K. 1991. Thoughts and observa- tions on the genus Poecilmitis Butler (Lepi- doptera: Lycaenidae) in the Transvaal. Metamor- phosis 2(3): 49-50. PRAIN, D. 1925. Cluytia. Pp. 427-456 In: THISELTON- DYER, W.T. (ed.). Flora Capensis: Part III. Lon- don: Reeve. PRINGLE, E.L., G.A. HENNING & J.B. BALL (eds.). 1994. Pennington's Butterflies of Southern Africa. Kaapstad: Struik Winchester. PULLIN, A.S. ed. Ecology and conservation of butter- flies. London: Chapman & Hall. PULLIN, A.S., I.F.G. MCLEAN & M.R. WEBB. 1995. Ecology and conservation of Lycaena dispar: British and European perspectives. Pp. 150-164. In: PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. RETIEF, E. & P.P.J. HERMAN. 1997. Plants of the northern provinces of South Africa: keys and diagnostic characters. Pretoria: National Botan- ical Institute. ROOS, P. & G.A. HENNING. 2000. The Heidelberg Copper Butterfly. African Wildlife 54(2): 24 -27. RUTHERFORD, M.C. & R.H. WESTFALL. 1994. Bio- mes of southern Africa: an objective categoriza- tion. Memoirs of the botanical Survey of South Africa No. 63. terblance 2.qxd 2005/12/09 10:31 Page 89 THOMAS, C.D. 1995. Ecology and conservation of butterfly metapopulations in the fragmented British landscape. Pp. 46-64. In: PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. VAN OUDTSHOORN, F.P., W.S.W., TROLLOPE, D.M. SCOTNEY & P.J. MCPHEE. 1991. Guide to the grasses of South Africa. Pretoria: Briza. VON BREITENBACH, F. 1990. National list of indige- nous trees. Pretoria: Dendrological Foundation. VON BREITENBACH, J., B. DE WINTER, R. POYNTON, E. VAN DER BERG, B. VAN WYK & E. VAN WYK. 2001. Pocket list of southern African indigenous trees: including selected shrubs and woody climbers. Pretoria: Briza. WARREN, M.S. 1995. Managing local microclimates for the high brown fritillary, Argynnis adipe. Pp. 198–210. In: PULLIN, A.S. (ed.). Ecology and conservation of butterflies. London: Chapman & Hall. Koedoe 46/1 (2003) 90 ISSN 0075-6458 terblance 2.qxd 2005/12/09 10:31 Page 90 << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.4 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /ColorConversionStrategy /LeaveColorUnchanged /DoThumbnails false /EmbedAllFonts true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveEPSInfo true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org) /PDFXTrapped /Unknown /Description << /ENU (Use these settings to create PDF documents with higher image resolution for high quality pre-press printing. The PDF documents can be opened with Acrobat and Reader 5.0 and later. These settings require font embedding.) /JPN /FRA /DEU /PTB /DAN /NLD /ESP /SUO /ITA /NOR /SVE >> >> setdistillerparams << /HWResolution [2400 2400] /PageSize [612.000 792.000] >> setpagedevice