Assessing the invasive status of cryptostegia grandiflora and C. madagascariensis in southern Africa and predicting potential risk areas

Abstract

Biological invasions are a major threat to global biodiversity. The complexity and high cost associated with management of invasions emphasizes the need for a strategic approach that can predict and prioritise potential invasive species. Risk assessments are commonly used tools that evaluate a species potential to become invasive. While risk assessments can be based on a variety of species characteristics, accurate estimates of a species contemporary distribution and the species potential future range, provide key information to prioritise management decisions. Understanding the factors driving the distribution of invasive species at different spatial scales can provide insight into the full range of environmental variables that influence its distribution. At broad-scales climate variables are expected to be the dominant factors determining where a species occurs and at finer scales, biotic interactions, disturbance, soil type and microclimatic conditions are predicted to become more important. In consequence, coarser scales are usually more suitable for directing management strategies (i.e. where to focus monitoring or control efforts), while finer scales are better to examine environmental features that influence distribution patterns within a single habitat. In this study, a multi-scale approached was used to assess the risk and predict the potential distribution of two Cryptostegia species identified as emerging invaders in southern Africa. Occurrence surveys were conducted across Botswana, Namibia and South Africa to determine the current extent and invasive status of Cryptostegia grandiflora and Cryptostegia madagascariensis. The potential broad-scale distribution of both species was predicted using species distribution models. By incorporating climate data and the current global distribution of both species habitat suitability maps were created using maximum entropy (MaxEnt; Chapter 2). At local scale, an invasive C. grandiflora population occurring along the Mogalakwena River in northern South Africa was used to determine the variables influencing fine-scale spatial variation in occurrence and cover (Chapter 3). Occurrence surveys confirmed both species are present in southern Africa. Cryptostegia grandiflora is widespread occurring as cultivated, naturalised and invasive populations in many semi-arid regions of southern Africa. In contrast, few C. madagascariensis occurrences were observed and none found outside of cultivation. Habitat suitability maps predicted large areas beyond current distribution highly suitable for both species. The mean temperature during the coldest quarter is the key restricting factor and predicted areas with temperatures below 15 °C to be unsuitable. The results from fine-scale surveys revealed that the drivers influencing C. grandiflora occurrence differed from those most strongly correlated with the cover of the species. Erosion and bare soil had weaker influences on C. grandiflora cover than on species occurrence, suggesting that this species can occur in moist disturbed areas but possibly achieves optimal growth under conditions of high leaf litter, low non-woody plant cover and deep soils with high moisture to initially establish. Given the widespread distribution, existing invasive populations and large uninvaded suitable range, C. grandiflora is a high risk species with the potential to severely impact riparian areas in semi-arid regions of southern Africa. Cryptostegia grandiflora’s current extent, high seed production and large dispersal range excludes eradication as a feasible management option. Monitoring and control efforts should instead focus on high priority areas already invaded or likely to be within the suitable ranges. In particular riparian areas in or adjacent to economically and ecological valuable areas (i.e. protected areas), should be targeted to prevent spread. In contrast, C. madagascariensis’s limited cultivated occurrences and restricted suitable range shows this species is not high risk and is not an imminent threat to southern Africa and is an ideal candidate for eradication.