1.1 The Cape Floristic Region's exceptional biodiversity

The CFR is a Mediterranean-type region, and with an extent of about 90 000 km², it is the smallest floristic region in the world (Figure 2). It is high in local plant endemism, and of the 9 000 or so plant species, 70% are endemic to the region (Goldblatt and Manning, 2002; Goldblatt, 1997). The CFR is also listed as one of the world's biodiversity hotspots, implying very high biodiversity levels while facing severe human impacts (Mittermeier et al., 2004). This region is composed of various vegetation types, but dominated by fynbos (sclerophyllous shrubland), renosterveld (grassy shrubland), strandveld (coastal shrubland), with pockets of Afrotemperate forest. As with all Mediterranean-type biomes globally, biodiversity is disproportionately sensitive to land-use and climate change (Newbold et al., 2020). However, the exact level of arthropod diversity is not known except for a few conspicuous groups, owing to the seemingly extremely high numbers of species, especially in less conspicuous groups. Furthermore, new species are being discovered on a regular basis, and following molecular analysis, many cryptic species complexes are coming to light (Samways et al., 2024).

Threats in the CFR are mostly from landscape transformation, habitat loss and fragmentation, and invasion from alien plant species (Rouget et al., 2014). Nevertheless, considerable progress has been made in conserving arthropod biodiversity in BRs of the region. The core areas and buffer zones of the BRs are identified through negotiation between all stakeholders. Once the biosphere reserve is established, priorities can be re-assessed in an iterative manner while critically appraising what conservation action has been effective relative to existing and emerging threats (Heijnis et al., 1999). This approach also requires effective management decisions.

1.2 Aims and structure of the review

We summarise recent evidence on the effectiveness of the four current BRs in the CFR for conserving arthropods in this megadiverse region. We performed a literature search on the Web of Science online database, searching article titles, abstracts and keywords to identify articles for inclusions in this review (refer to Supplementary File 1 for the list of keywords used in the search). Research articles that did not focus on arthropods specifically, those that investigated marine arthropods, and those that were outside the geographic scope of the CFR, were excluded from this review. References within relevant articles, and several articles that were missed by the keyword search but gathered through additional expert searches, were also included.

We summarise the value of BR core areas and evaluate the conservation impact of agroecological approaches to enhance arthropod diversity in buffer and transition zones. We then summarize findings on other management approaches that enhance arthropod conservation across whole BR areas. While the Table Mountain National Park on the Cape Peninsula is not a proclaimed BR, it effectively interacts with the surrounding landscape as one. Against the background of BR design and management, we highlight the exceptional biodiversity of Table Mountain and surrounding areas. Finally, we provide management recommendations for future directions, highlighting the relevance of BRs in arthropod conservation.

2.1 Kogelberg Biosphere Reserve

The Kogelberg BR (KBR) is at the heart of the CFR and was South Africa's first biosphere reserve, designated in 1998 (Figure 3). This BR has core terrestrial areas in the formally protected Kogelberg and Groenlandberg Nature Reserves, mostly restricted to mountainous areas. This core zone has extraordinarily high levels of biodiversity and endemism (Johns and Johns, 2001). The nature reserves are bordered by a buffer zone that is mostly in a natural and near-natural state, whereas the transition zone consists of specialized conventional deciduous fruit production and commercial timber plantations.

2.2 Cape West Coast Biosphere Reserve

The Cape West Coast Biosphere Reserve (CWCBR) is located on West Coast of South Africa, designated in 2000 (https://www.unesco.org/en/mab) (Figure 3). The core area of the BR is centred around West Coast National Park and is known for its mass flowering displays in spring and for protecting West Coast renosterveld, sand plain fynbos, and coastal strandveld vegetation types, all being unique to this region. Surrounding the BR core are game farms and some mixed agriculture. Agriculture dominates the landscape outside the protected area, with wheat and canola production along with cattle farming being most prominent.

2.3 Cape Winelands Biosphere Reserve

The Cape Winelands Biosphere Reserve (CWBR) is in the centre of the CFR wine growing region and was established in 2007 (Figure 3). The core area of this BR is in the Hottentots-Holland Nature Reserve which protects the mountains and the water catchments for much of the area. Surrounding the core and at lower elevations are conservancies and low impact viticulture. Much of the CWBR transition zone is highly productive wine farms and urban areas.

2.4 Garden Route Biosphere Reserve

Established in 2017, the Garden Route Biosphere Reserve (GBR) is on the eastern limit of the CFR, incorporating both fynbos and naturally patchy forest (Figure 3). Core areas include the Garden Route National Park, Tsitsikamma, Goukamma and Robberg Nature Reserves. Much of the surrounding area has been converted to plantation forestry and dairy pastures. In the buffer zone, sustainable timber harvesting is permitted.

4.1 Edge effects between zones

Species turnover of surface-active arthropods in the KBR is related to mesoclimate, fire history, and geology. In this BR, the buffer zone has important complementary value and increases representativeness across the landscape, especially in areas of increased intra-annual temperature variability. As deciduous fruit orchards influence arthropod diversity in adjacent natural areas by <1 km, the buffer zone is important for protecting the core zone, for example, by reducing the impact of too frequent fires and any pesticide drift. Across all KBR zones, habitat quality in the form of high plant diversity, low human disturbance and being clear of alien invasive plants, is a key determinant of katydid diversity (Thompson et al., 2020). This highlights that good management of the buffer zone is essential for long-term maintenance of the arthropods in the core, as well as promoting biodiversity within the whole BR (van Schalkwyk et al., 2019).

At the smaller spatial scale of the edges between deciduous fruit orchards and natural habitat in the KBR, it is important that there are buffers of >85 m wide where management (such as fire protection, control of invasive alien plants, and avoidance of spillover of fertilizers) protects the natural habitat to sustain natural arthropod diversity (van Schalkwyk et al., 2020a). This approach is enhanced where there is improved functional connectivity at this smaller spatial scale within buffer and transition zones (van Schalkwyk et al., 2020b). As spillover from natural vegetation to orchards is not enhanced by reduced edge contrast or reduced crop management intensity, an important way in which this connectivity could be achieved is through corridors and stepping-stone habitats of natural vegetation (van Schalkwyk et al., 2020b), although this has yet to be tested in the region.

In the GBR, edge effects, as measured by arthropod compositional changes, were only 20 m between the fynbos and Afrotemperate forest, which is far smaller than when forests are adjacent to the structurally similar pine plantations (30 m), and clear felled areas (50 m) (Swart et al., 2018). Interestingly, hiking paths show edge effects into Afrotemperate forests of around 5 m, narrow 4 × 4 tracks about 20 m, and gravel roads >50 m, indicating how sensitive this vegetation type is to fragmentation (Swart et al., 2019).

4.2 Agroecological farming in the buffer and transition zones

With agricultural intensification threatening biodiversity, there has been a shift in the CFR to agroecological farming especially in vineyards of the CWBR. These approaches focus on reduced management intensity within crop fields and increasing heterogeneity throughout the farmland mosaic. This shift requires identifying locally specific management practices that are practical and can effectively enhance biodiversity, guiding farmland conservation efforts.

In CWBR vineyards, overall arthropod species richness, and that of predators, detritivores, and herbivores is highest in natural sites, followed by organic vineyards and then by integrated vineyards, partially due to higher non-crop vegetation complexity and less intense management in the organic vineyards (Gaigher and Samways, 2010). For species composition, arthropod assemblages are similar in organic and integrated vineyards, both of which are complementary to natural sites. Furthermore, variation among arthropod assemblages associated with natural vegetation is also much greater than among arthropod assemblages of cultivated sites. While organic vineyard management makes an important contribution to arthropod conservation in the CFR at the field scale, at the landscape scale, natural habitat supports a much more diverse set of species. This emphasizes the importance of conserving natural fragments in the vineyard landscape as an effective measure to maintain biodiversity of production landscapes in the CFR (Gaigher and Samways, 2010).

While ground-living spiders are diverse in vineyards of the CWBR, especially in organic vineyards, it is natural habitat remnants that contribute significantly to overall spider diversity (Theron et al., 2020). Natural vegetation remnants support rare and range restricted species, and many species absent from the agricultural matrix. Both patch- and landscape-scale variables influence spider diversity and assemblage structure. Remnant patch size and its vegetation, as well as topographic variety, positively influence overall spider diversity and assemblage structure of all guilds. Spider diversity is maintained by remnants of good quality natural vegetation, but also influenced by the complexity of the landscape. Conservation of remnant vegetation patches, regardless of size, across privately-owned farm conservancies maintains spider diversity, an encouraging sign (Theron et al., 2020). Other environmental variables relating to plant diversity, management intensity, and distance to nearest natural habitat influence spider abundance and species richness in vineyards (Gaigher and Samways, 2014).

Vineyard management practices in the CWBR are being improved, but advances require verification. In a study of arthropod biodiversity in vineyards under varied management intensities, the most influential factor is percentage herbaceous vegetation cover in vineyards, which has a positive effect on arthropod species richness across several taxa (spiders, beetles and true bugs), microhabitats (ground level, on cover crops, and on vine foliage), and feeding guilds (herbivores, omnivores and predators) (Geldenhuys et al., 2021). Other vegetation-related variables, such as volume of plant litter, plant species richness, and plant height also positively influence the biodiversity assemblages present. This multi-taxon approach did not find any negative effect of agrochemicals used in an integrated approach, or any strong effect of overall farming approach (organic vs. integrated) on the species richness or evenness of any of the focal taxa, sampled positions, or guilds. But in this area, vineyards are under the auspices of the Integrated Production of Wine Programme, which applies pesticides with much discretion. Overall, indications are that the maintenance of dense and diverse cover crops, along with sensitive management supports integration of biodiversity conservation with farming. It also highlights that minimal adjustment to management practices can greatly benefit farmland biodiversity conservation and is in keeping with the ethos of the BR concept (Geldenhuys et al., 2021).

In vineyards, herbaceous vegetation cover promotes ubiquitous arthropod species richness, although neither vegetation cover, nor any other in-vineyard measures promote spillover from fynbos into vineyards (Geldenhuys et al., 2022a). However, assemblages on the soil surface display lower nestedness differences, indicating greater spillover by these ground-level arthropods. This emphasizes again the importance of conserving natural fynbos patches alongside sensitively managed vineyards for harmonizing biodiversity conservation and viticulture (Geldenhuys et al., 2022a).

Furthermore, these assemblages in integrated vineyards are similar in composition irrespective of landscape context, whereas they are different among organic vineyards in landscapes with different levels of natural vegetation. This indicates a relatively lower homogenizing effect of organic farming compared to nonorganic farming on landscape-scale biodiversity. The point here is that within this farming and conservation mosaic, it is important to consider the interplay between local management and the surrounding landscape in promoting biodiversity (Geldenhuys et al., 2022b) (Figure 4). Further assessment of how on-farm measures can promote species turnover across the buffer and transition zones will be important to prevent regional homogenization in the BRs.

Vineyards support a relatively high taxonomic diversity of arthropods, and high species turnover, although compositionally, assemblages differ greatly from those in nearby natural habitats in the CWBR. Vineyards also support lower functional diversity of spiders and beetles (Geldenhuys et al., 2022c). Assemblage composition among the two taxa differs according to system relative to arthropod traits. Plant-dwelling spiders and small-bodied beetles are strongly positively associated with fynbos, and predatory beetles within vineyards. Common spiders and predatory beetles are strongly negatively associated with fynbos. Overall, results indicate that well-manged vineyards can support high arthropod diversity (Geldenhuys et al., 2022c), while also recognizing that inter-rows with high plant species diversity can improve food networks in the case of web-building spiders (Arvidsson et al., 2020). However, remnant patches conserve irreplaceable biodiversity and functional diversity, making their continued conservation in farmland a priority (Figure 5).

A further consideration is soil compaction which is higher in agricultural areas than in natural vegetation in the CFR. While degree of soil compaction negatively correlates with arthropod species richness but not with abundance, neither soil moisture nor leaf litter depth on their own significantly affect arthropod species richness or abundance. Furthermore, alien trees have a strong negative effect on both arthropod species richness and abundance, and much more so than vineyards. However, the situation is reversible, with removal of the alien trees leading to a rapid recovery of soil structure and of arthropod assemblages (Magoba et al., 2015).

Grasshopper abundance is highest in CWBR vineyards compared to either fynbos or deciduous fruit orchards. However, species richness, diversity, and evenness are highest in fynbos, followed by vineyards then orchards. In turn, orchards have no unique species, while vineyards have few unique species, and assemblages are overall similar between vineyards and orchards. Fynbos has several unique species, mostly made up of flightless endemics, with open vineyards offering more opportunities than canopied orchards. Maintaining patches of natural good quality fynbos among transformed land is essential for securing endemic species (Adu-Acheampong et al., 2016).

As fynbos transformation progresses, there is replacement of endemic grasshopper species with more widespread generalist species (Adu-Acheampong and Samways, 2019a). Besides land-use type and species traits, seasonality also plays a major role in their spatial dispersion patterns. Highly mobile, generalist feeders are abundant and widely dispersed across the landscape, especially late season when they can take advantage of high plant productivity in the vineyards. In contrast, low mobility, specialist feeders only occupy natural fynbos vegetation in both seasons. Furthermore, highly mobile generalists benefit in two ways: being able to occupy transformed areas, while receiving a population boost late season. In the case of low-mobility specialists, they are doubly disadvantaged: not able to move far, while lacking their specific host plants in the transformed areas. This means that the conservation emphasis for CFR endemic grasshoppers needs to be on improving functional connectivity in natural fynbos (Adu-Acheampong and Samways, 2019b). Among species in the CFR endemic grasshopper genus Euloryma, two species, E. umoja and E. ottei can tolerate some transformation away from fully natural fynbos. However, the traits of E. larsenorum and E. lapollai are such that they can only survive in fully natural fynbos (Adu-Acheampong et al., 2017).

In turn, there are only weak differences between grasshopper assemblages in the riparian zone of CFR rivers compared to the terrestrial zone, apparently driven by deep history, complex geomorphology, stressful environmental conditions, a diverse vegetation, land mosaics, and possibly high predator pressure. The large-sized, well-flighted, geographically widespread generalists are more abundant in the riparian than terrestrial zone, while the small, flightless or poorly flighted vegetation specialists that are narrow-range endemics are adapted to both riparian and terrestrial zones but are most abundant in the terrestrial zone. Overall, indications are that narrow-range CFR grasshoppers are best conserved in the terrestrial zone (Pronk et al., 2017).

In the case of parasitoids, fynbos fragments are higher in abundance, species richness, and diversity than vineyards, with little spillover between the two systems, and very different assemblages in both. While the remnant fynbos patches are isolated, they are important reservoirs of natural parasitoid diversity. Improvements to the agricultural matrix through encouraging functional connectivity would likely enhance long-term persistence of parasitoids, especially during adverse weather events (Gaigher et al., 2015).

Abandoned vineyards represent an intermediate level of transformation between natural vegetation and productive vineyards. They are closest to natural vegetation in their parasitoid and predator arthropod diversity, while also having some unique species in these two guilds, with overall parasitoid and predator diversity benefiting from high plant diversity and arthropod prey abundance in the abandoned vineyards. Parasitoids have greater habitat specificity than the predators, with the parasitoids likely to benefit most from an improved mosaic of natural vegetation and abandoned fields plus better functional connectivity (Gaigher et al., 2016) (Figure 5).

6.1 Managing for fire events

CFR vegetation is fire-prone and fire-adapted, with fire being a principal driving force in the ecological dynamics of fynbos and renosterveld ecosystems, and both inevitable and necessary. However, with an increasing human population, fire events have become more frequent than in the past (Van Wilgen, 2013). For arthropod conservation, better understanding of historically complex fire survival by various taxa and guilds is needed.

Some groups, such as ants, are highly resilient to fire, while others, like pollinators, are less resilient at a local scale. As fire also allows new species to enter burned sites, it can promote local biodiversity, so long as fire events are not too frequent (Pryke and Samways, 2012a). This differential resilience to fire events emphasizes the importance of sampling a wide range of groups to represent overall responses of compositional biodiversity. This means using a variety of groups, such as butterflies, ants, and scarab beetles, for monitoring arthropod recovery following fire, while also having adequate control sites, as there is much annual variation in arthropod diversity and abundance in unburned areas (Pryke and Samways, 2012b).

Furthermore, the effect of fire on species diversity and abundance varies greatly among taxa. This effect emphasizes the importance of using species compositional diversity measures, especially as different time-since-fire categories support distinctive assemblages, and that some indicator species are present across all temporal categories while others are present only at one time-since-fire category (Yekwayo et al., 2018).

Even within guilds there can be different responses to fire. In the case of spiders, while there are no differences among web-builders relative to time since fire, wandering spiders take longer to recover than web builders. In practical terms, this means that wanderers are good candidates for monitoring postfire recovery (Yekwayo et al., 2019).

Species composition of flowering plants and anthophiles, especially bees and flies, differ significantly across postfire categories. Flower and bee abundance are largely congruent. For the other anthophiles, other resources such as logs for breeding flies are essential (Adedoja et al., 2019a). Importantly for conservation is that fire refuges, in the form of ravines, rocky outcrops, screes, and areas that by chance escape the fire front, play a major role in providing source areas from which to recolonize previously burned and recovering areas (Adedoja et al., 2019b). It is important that these refuges are present across the various zones of BRs, so mitigating the regional impacts of landscape transformation (Figure 7).

As there is high beta diversity of plants and arthropods in the CFR, fire management, especially in BRs, must be carried out carefully. This is emphasized by the insect borers on shrubby plants in the region (e.g., Protea spp.). There are distinct differences in frequency of occurrence of insect taxa associated with the various plant species as a result of physical host-plant characteristics, such as infructescence and seed set variables, and also there are local climatic factors. As plant diversity is largely maintained by burning, management burns should not be applied too frequently or over too large areas where there might be extirpation of borer species. As the borers are effective umbrella species for many other insect species, their conservation has applicability to CFR insect diversity in general (Wright and Samways, 1999).

6.2 Managing invasive alien species

Invasive alien plants, especially trees and shrubs, are highly impactful on CFR biodiversity and can aggravate fire events (Van Wilgen et al., 2016) (Figure 8). In terrestrial areas, alien trees in the genus Pinus are also interactive with the alien ant Linepithema humile by variously impacting indigenous ant species. Of particular concern is that formicine ants in the genus Camponotus are affected most, with these ants playing an important role in the burial of seeds of many indigenous fynbos plants, while also having obligate interactions with localized and endemic butterfly species (Schoeman and Samways, 2011, 2013). In the case of the small forest specialist butterfly Charaxes xiphares occidentalis, alien trees (Acacia mearnsii) are adversely synergistic with agriculturally induced landscape fragmentation, affecting its behaviour and ecology, and hence its survival (Crous et al., 2015).

Comparison of surface-active arthropod species in vineyards of the CWBR, alien tree-invaded areas, and other areas cleared of the alien plants, indicate that species richness in cleared areas is higher than in vineyards and more similar to that in natural fynbos, while alien trees have lowest species richness (Magoba and Samways, 2012). Overall abundance is highest in cleared areas, closely followed by fynbos, then vineyards, and lowest under alien trees. Several species are restricted to each vegetation type, including alien tree infested areas. However, assemblage composition differs greatly among the vegetation types, with fynbos and vineyards grouping together. When rare species are excluded, vineyards and cleared sites group together, indicating some recovery but only involving those species that are abundant and habitat tolerant. In general, vineyards retain a greater complement of indigenous species than alien tree infested areas, but clearing of these aliens soon encourages establishment of indigenous species, with soil moisture and litter depth a major driver (Magoba and Samways, 2012).

There is a reduction in percentage light reaching the understorey of alien pine stands which reduces understory flower abundance, with no flowers under mature pine trees. When anthophile species diversity and interactions are measured at various distances in natural fynbos away from alien pine stands, anthophile species composition differs across sampling locations with increasing distance from the mature pine understorey, an indication of fynbos plant heterogeneity (Adedoja et al., 2021). Although pine trees cause a decline in flower-visitor interaction frequency and network specialization, some bee species that are mostly tree nesters, are present in unique interactions associated with mature pines and where the alien trees provide nesting sites. While alien pine tree removal is encouraged, especially where the trees are actively invading, maintaining well-contained small stands used for threatened mammal conservation, is of value for the indigenous anthophile taxa (Adedoja et al., 2021).

Among the invasive alien insects is Polistes dominula, which has now become one of the dominant wasp species at some localities surrounding the CWBR, displacing native P. marginalis (Benadé et al., 2014). While both species prefer human infrastructure for nesting, P. dominula is significantly more abundant than P. marginalis, despite the alien species having a higher parasitoid load. However, P. dominula has a much longer activity period and higher colony productivity than P. marginalis, meaning that it can reach much greater population size than P. marginalis despite the higher parasitoid pressure (Roets et al., 2019).

The invasive alien ant Linepithema humile can be adversely synergistic with landscape transformation. When apple orchards with intensive insecticide or acaricide programmes are compared with natural fynbos, far more species of Hemiptera, Hymenoptera, and Orthoptera are present in natural vegetation than in either orchard management type. Cydnid, rove, and ground beetle families are as species rich in the sprayed as in the unsprayed orchards, whereas Diplopoda and Isopoda are more abundant in the unsprayed than sprayed orchards. However, alien L. humile is the most abundant species in both orchard types. Overall, landscape transformation and intense use of pesticides has a negative impact on arthropod diversity in the KBR (Witt and Samways, 2004).

Among invasive alien fish, Rainbow trout (Oncorhynchus mykiss) can penetrate the headwater streams where much CFR biodiversity is present. Trout change community structure, as well as the dynamics of the algae-invertebrate interaction (Shelton et al., 2015), even though they feed mainly on terrestrial invertebrates, mostly insects, which fall on the water surface (Shelton et al., 2016). More information on other alien fish, for example, Largemouth bass and Smallmouth bass, is required, as these invaders may also have impacts on aquatic arthropods and low-flying insects.

The curculionid polyphagous shot hole borer (PSHB, Euwallacea fornicatus) is an ambrosia beetle, indigenous to southeastern Asia, now a global pest and first recorded in South Africa in 2017. The PSHB vectors various fungi species, including the pathogenic Fusarium euwallaceae which grows in the xylem and along gallery walls, providing nutrition for developing beetle larvae and adults. This activity blocks the xylem, leading to twig dieback and even death of the tree. There are to date 83 known tree host species (41 indigenous and 42 alien) in which PSHB can reproduce (FABI, 2023), some of which include economically important fruit trees. Uncontrolled, the beetle has severe ecological and economic implications for the CFR and other regions of the country (Van Rooyen et al., 2021). Current control methods include removal of infected material by pruning or felling of highly infested trees. These approaches are followed by wood chipping and sun exposure of the woody material, all of which are expensive and not feasible on a large scale. As chemical control (Roberts et al., 2024) and application of commercially available broad-spectrum entomopathogenic fungi (Bacillus thuringiensis; Beauveria bassiana) (Nel et al., 2023) against PSHB are only partially effective, successful beetle management will likely depend on a combination of chemical control and other control strategies in an integrated pest management programme. This approach would need to be coupled with a strict monitoring protocol where citizen science can play a major role (Potgieter et al., 2024). Of concern in the CFR, is that this invasive beetle has been confirmed from the KBR, CWBR, and GBR. Infestation in the GBR is particularly concerning, as a high number of indigenous Afrotemperate forest trees are highly susceptible to the pest and its mutualistic fungus (Townsend et al., 2024), which could have major conservation implications for arthropod conservation in the CFR. This emphasizes the urgency of developing an effective monitoring and control programme.

6.3 Reducing the impact of stressors on aquatic systems in the biosphere reserves

Alien invasive tree species (e.g., Acacia spp.) are a major threat to the freshwater systems, especially rivers across the CFR. However, there has been a largely successful major national initiative, the Working for Water programme, aimed at alien riparian tree removal across most rivers in the CFR (Van Wilgen et al., 2012).

Besides depriving indigenous plants of water, the alien trees have a huge impact on the local freshwater assemblages by shading out of understorey plants and through having a direct shading effect on the largely heliophilic arthropods (Remsburg et al., 2008). However, removal of the alien trees has enabled freshwater assemblages and communities to recover. For instance, the recovery process has seen the return of many dragonfly species (Samways and Sharratt, 2010) and macroinvertebrates (Samways et al., 2011), including several CFR endemic species.

The level of recovery has been quantified using dragonflies alone (Simaika and Samways, 2009), which can also stand in for other assemblages, such as the Ephemeroptera/Plecoptera/Trichoptera, the well-known and sensitive ‘EPT’ group (Kietzka et al., 2019). Furthermore, the Endangered damselfly Spesbona angusta has umbrella value for two further threatened odonate species, Endangered Proischnura polychromatica and Vulnerable Syncordulia legator, as well as some other CFR endemic odonates. Conservation of this species requires continual monitoring and removal of known invasive alien trees as a priority (Deacon and Samways, 2017). Monitoring and removal of new invading alien trees, such as Prosopis spp. already present in the northern provinces of South Africa, should also be prioritised to better protect other odonate assemblages.

Agricultural and urban landscape transformation both have a similar homogenizing effect on CFR dragonflies, with opportunities for local endemic species reduced while opening up opportunities for the widespread generalist species. Besides improved water and riparian conditions befitting the endemics, the presence of core BR zones is critical for them (Kietzka et al., 2018).

A challenge for aquatic insects in the CFR has been survival during severe droughts. Dragonflies, aquatic beetles, and aquatic bugs today have the alternative option of being able to retreat to artificial water bodies in times of severe drought, especially when these artificial ponds are similar in all abiotic and biotic factors to natural ponds (Deacon et al., 2019). While good quality natural ponds in the core zones of BRs are important, adequately managed artificial ponds, including those in buffer and transition zones, can make a substantial contribution to regional aquatic insect conservation (Figure 9).

Among beetles and bugs, the proportion of CFR endemic species in high quality artificial ponds is 36% (Apinda Legnouo et al., 2014). Encouragingly, for sandy lowland fynbos wetlands, dense urbanization has not yet had any significant impact on aquatic macroinvertebrate species composition (Blanckenberg et al., 2020). The key to the success of artificial ponds for maintaining high aquatic insect diversity is a pondscape with a range of ponds with various degrees of water level fluctuation, along with naturally diverse marginal vegetation in this drought-prone region (Jooste et al., 2020).

6.4 Impact of river impoundment in the Cape Winelands Biosphere Reserve

As water is a relatively scarce commodity in the CFR, impoundments have been put in place to secure water supply for humans. However, impoundment of a small river in the CWBR core zone resulted in macroinvertebrate species diversity below the dam dropping to only half of that in the natural area catchment area above the dam. Furthermore, Ephemeroptera, Plecoptera and Trichoptera diversity and abundance dropped to almost zero because of changing physicochemical conditions brought about by the impoundment. In contrast, abundance of Chironomidae fly larvae increased substantially below the dam (Bredenhand and Samways, 2009).

The effects of river impoundment in the other BRs are likely similar, yet more investigation is needed to determine how local freshwater arthropod assemblages respond. Importantly, ways to reconcile the needs for water storage and conservation is needed. A step forward is to simulate natural riparian vegetation composition in transformed stream areas, reduce accumulation of pollutants, and allow dam water levels to fluctuate, as would be the case under natural riverine conditions.