Research into the Brenton Blue Butterfly

The earlier years

The Brenton Blue’s scientific name is Orachrysops niobe (Trimen) (Lepidoptera: Lycaenidae), and Roland Trimen discovered it in 1858 at an unspecified locality in Knysna (Trimen 1862; 1887).

No further records were made in the Knysna vicinity, although butterflies thought to be conspecific with O. niobe were found over a wide area of South Africa as summarised by Swanepoel (1953). He was also the first one to draw attention to the occurrence of dwarfism in species of this group.

The first attempt to establish the life history of the closely related Orachrysops lacrimosa was by Clark & Dickson (1971), and since they were unable to raise the larvae past the 2nd instar, it was assumed that they were not phytophagous in the later stages and were in all probability carnivorous on ant brood, similar to other Lepidochrysops species. At this stage of our knowledge niobe continued to be placed in the Lepidochrysops genus.

In 1977 Jonathan Ball discovered a population of O niobe at Nature’s Valley near Plettenberg Bay (Ball 1996). On comparing the specimens with other supposed niobe from the Eastern Cape, he made the remarkable observation that the Nature’s Valley butterflies were different from the ones found further east. This aroused the interest of taxonomists and Dr Lajos Vári of the Transvaal Museum began a detailed study of the group, resulting in the erection of the new genus Orachrysops, which contained initially five species including niobe (Vári & Kroon 1986).

When the first South African Red Data Book was published, O. niobe was recognised as one of South Africa’s most endangered butterflies and was classified as Vulnerable (Henning & Henning 1989).

The colony at Brenton-on-Sea (1991-1996)

During Jonathan Ball’s searches for O. niobe the only other place where he came across the species was near the top of a hill at Brenton-on-Sea. Consequently, following up this lead, Ernest Pringle was driving up the road to this hilltop when he saw a dark butterfly flit across the road (Pringle 1991). This turned out to be O. niobe, and in good numbers, so clearly a breeding colony was nearby.

The 2nd Edition of Pennington’s Butterflies published in 1994 included the definitive work on the Orachrysops genus by Henning & Henning (Pringle et al. 1994). They recognised seven species and two sub species.

In October 1993, Dave Edge moved to Knysna and began intensive investigations into the colony at Brenton-on-Sea (Edge & Edge 1994). The flight path of the males and the size of the colony were established, the flight period was confirmed to be October/November and February/March, and males and females were observed indulging in mating rituals.

During November 1995 two more butterfly experts – Dr Martin Kruger, head of Entomology at the Transvaal Museum and Dr Mark Williams, an acknowledged expert on butterfly-ant relationships, arrived in Knysna to further intensify the research. Together with Edge they were able to establish the full range and extent of the colony, observe mating rituals and females ovipositing (Williams 1996). The foodplant was identified as Indigofera porrecta (sic). Fifteen eggs were collected and feeding of O. niobe larvae began in plastic containers. In order to explore the ant relationship a colony of ants from the site was established in a formicarium, and larvae were introduced to the ants.

From the eggs collected Edge managed to get four of them through to adult butterflies (Edge & Pringle 1996). Unexpectedly, no ant interaction was necessary, and the larvae were entirely phytophagous. It was now clear that an abundance of foodplant was even more important than previously thought, so the assistance of Jan Vlok, a botanist from Cape Nature Conservation was sought. He carried out an investigation into the botanical community at the site and suggested ways in which growth of Indigofera porrecta (sic) could be promoted (Vlok 1996).

The EWT research initiative (1996-1997)

In November 1996 The Endangered Wildlife Trust (EWT) commissioned a number of scientific experts to conduct research at the Brenton-on-Sea site. Professor Roy Lubke of Rhodes University Department of Botany, and two colleagues, conducted an investigation into the vegetation and floristics of the site, identifying 150 plant species, two main plant communities, and the characteristics of the microhabitats where the O. niobe females were ovipositing. The food plant was re-identified as Indigofera erecta (Thunb.). Impacts of the housing development on the habitat and microclimate were assessed and recommendations for the conservation of the butterfly were made, which included complete cessation of building activities (Lubke et al. 1997).

The second research team (Dave Britton and Dr Tish Silberbauer) investigated the life history and ecology of the butterfly. The earlier work by Edge and Pringle (1996) was confirmed, and an accurate record was made of the oviposition sites. In the absence of any observations of butterfly – ant interactions, emphasis was placed on the importance of maintaining an optimal density of the foodplant, and the urgency of establishing appropriate management techniques (Britton & Silberbauer 1997; Silberbauer & Britton 1999). Concurrently with this investigation an ant expert from the South African Museum, Dr Hamish Robertson, did a census of ant species at the site and the surrounding areas. He identified those ants that were candidates for an O. niobe association and assessed the potential threat from the Argentine ant Linepithema humile to the ant communities on the site. No ant – butterfly larvae interactions were observed but the need for further research was emphasised (Robertson 1997).

The publication by Alan Heath of the definitive description and taxonomic diagnosis of the Kammanassie Blue Orachrysops brinkmani in March 1997 laid to rest speculation as to the status of this insect and its relationship to O. niobe, which is a closely related but separate species (Heath 1997). This also reconfirmed the critical importance of the Brenton-on-Sea colony as the only known viable breeding colony of O. niobe.

Robertson (1998; 2000) followed up his previous work and did further research in November 1997, 1998 and 1999. Slight encroachment towards the reserve site of the Argentine ant was observed, as was some deterioration of the I. erecta plants due to overgrowth of bracken. He provided colonies of Camponotus niveosetosus and C. baynei to butterfly breeder Jill Reid of Oudtshoorn who was able to raise several larvae on cut food plant to third instar stage from eggs supplied by Pringle. Some indication was obtained from ant-larvae interactions that C. niveosetosus was the likely ant host. Reportedly larvae were taken from the foodplant by the ants and returned several days later significantly larger (Pringle 1999).

Research conducted by Dave Edge (2000-2005)

Dave Edge commenced a research programme in October 2000, as part of his studies towards an MSc in Environmental Science through North West University. The research was funded in part by the Green Trust.

A preliminary study looked at some of the key ecological factors influencing the breeding success of the Brenton Blue – host plant abundance and condition, nectar sources, climate/ microclimate, and vegetation management techniques. The adult butterfly population was monitored over an entire breeding season; host plants were identified and individually monitored; and egg counts were done. This enabled the effects of a number of different management techniques to be evaluated (burning, cutting, physical removal of invasive elements, and combinations thereof). A fivefold increase in the population of O. niobe was observed over the breeding season. This increase was positively correlated to a similar increase in host plant abundance in the areas where cutting and physical removal of invasive elements was practiced. Burning by contrast appeared to have a negative impact on host plant and butterfly abundance over the same period. Impacts of other factors such as climate, nectar sources and the natural strength of the second brood are discussed. A hypothesis of megaherbivore activity as the principal historical disturbance mechanism promoting locally favourable conditions for O. niobe to establish and maintain colonies is proposed. Recommendations for reserve management and future research were made. (Edge 2002a).

In November 2005 Edge published his doctoral thesis entitled “Ecological factors influencing the survival of the Brenton blue butterfly” (Edge 2005b). This research investigated the life history and ecological interactions of O. niobe and recommended management techniques for the BBBR. Adult nectar sources and female oviposition behaviour were described. The first two larval instars feed on the leaflets of the host plant Indigofera erecta Thunb., and the 3rd and 4th instar larvae feed on this plant’s woody rootstock, attended by ants Camponotus baynei Arnold. Cannibalism takes place in the early larval stages. Adults emerge from late October – early December, from late January to early March, and occasionally there is a third brood in April. Morphological and behavioural features of the larvae were described, and larval growth patterns were compared to other polyommatines. The nature of the myrmecophilous behaviour is assessed and the phylogenetic relationships between Orachrysops, Euchrysops, Lepidochrysops, and other polyommatine genera were discussed.

Ant assemblages at the BBBR were described from pitfall trap sampling and field observations of interactions between O. niobe larvae and ants. The ant assemblages at the BBBR, Nature’s Valley Fynbos Reserve (NVFR) and Uitzicht 216/ 40 were compared. The potential impact of the Argentine ant at the BBBR was discussed.

Adult and egg counts were used to study the population dynamics of O. niobe, leading to an estimate of the adult butterfly population and its fluctuations. A population dynamics model was constructed, and several factors impacting fecundity and mortality were assessed. Population studies on other polyommatines were compared and discussed.

The morphology, reproductive biology and autoecology of the papilionoid legume I. erecta were described. An explanation was offered why the larvae of O. niobe are monophagous on this plant. Microhabitat variations cause changes in its vegetative morphology and this further explains the restricted range of O. niobe. The population dynamics of I. erecta was investigated and the effects of biotic and abiotic factors (including fire) assessed.

Braun Blanquet methodology was used to sample and classify the vegetation communities at the BBBR. Ordination techniques were used to confirm the classification and to diagnose for environmental gradients. Hypotheses were generated about the ecological processes functioning at the site, and the environmental niche occupied by I. erecta. Correlations between the occurrence of vegetation types and other plant species and I. erecta were sought and a strong association with Pterocelastrus tricuspidatus (Candlewood) was demonstrated.

Three study sites were described and the ecological history of the region was reconstructed. Ecosystems are no longer fully functional because natural fire regimes and megaherbivores are absent. It is suggested that O. niobe currently inhabits an interglacial refugium under stress from the current global warming trend. The remaining suitable habitat is threatened by coastal property development and environmental degradation. An increased population of O. niobe at the BBBR, and establishment of new populations is essential to avoid extinction. Management techniques were evaluated and a management strategy for the BBBR was proposed.

Nature’s Valley (2002-2007)

A research programme was proposed by Edge (2002b), with the objectives of evaluating methods to promote the growth of a large population of O. niobe’s host plant at the Fynbos Reserve in Nature’s Valley, where O. niobe used to fly. It was suggested that part of the reserve would be burnt and that cutting would take place in the unburnt portion. The Nature’s Valley Trust (NVT) did not adopt this proposal. Instead, relying on advice from botanical experts, and with the aim to convert the reserve back into fynbos, woody elements were cut down and removed (including Candlewood trees), and the entire reserve was burnt in April 2003, using diesel and petrol to make the moist vegetation burn (Rodwell & van Hasselt 2003). After a site visit in November 2003 recommendations were made as to how growth of the Brenton Blue food plant Indigofera erecta could be encouraged (Edge 2003).

The Green Trust provided funding to attempt the re-establishment of the Brenton Blue at Nature’s Valley (Edge 2005a) and research commenced in July 2005 (Figure 6). In a December 2005 progress report (Edge 2005d), the failure to detect the host ant of the Brenton Blue, Camponotus baynei, was recorded and the poor condition of the food plants was noted. The July 2006 progress report (Edge 2006) the release of larvae at Nature’s Valley in December 2005 and February 2006 was described, and the continued failure to detect the host ant was reported on. The February 2007 progress report confirmed that none of the larvae released during April 2006 had survived to emerge as adult butterflies in November 2006. The reasons given for this failure were either that the I. erecta host plants were not healthy enough and did not have adequate rootstocks (unlikely); or that because the host ants Camponotus baynei were not present the larvae were insufficiently protected.

It was recommended that, in view of the limited ability of the BBBR O. niobe population to sustain removal of individuals for establishing a NVFR population, and because of the uncertain future of the NVFR, that no further O. niobe translocations are attempted until the  future status of the NVFR is guaranteed (the Plettenberg Bay Municipality is planning to sell it to a developer); a viable population of Camponotus baynei ants is established at the NVFR; and that the density and vigour of the I. erecta population at the NVFR is improved.

Further research and publications (2008–2017)

Monitoring of the Brenton Blue population at the BBBR is continuing, and the population has fluctuated between 60 and 190, after some very dry summers.

Brenton Blue population 2002 – 2016

Management techniques continue to be refined following the recommendations of Edge et al. (2008a) and the impact on the health of the host plants I. erecta and adult butterfly population will continue to be monitored.

The Knysna Fire of June 2017 – a new research programme