Aim: Biogeographic boundaries can act as either weak or strong barriers to the spread of species undergoing distributional change. Once a novel species spreads across a boundary, it can have a substantial impact on the ecosystem, for instance by competing with local species, and, over the long-term, re-engineer the ecosystem. Marine biogeographic regions are clearly defined on the coast of southern Africa and we tested the influence of their boundaries on the spread and spatial structure of an alien ecosystem engineer.Location: Southern Africa.Taxon: An invasive mussel, Mytilus galloprovincialis. Methods: Records of M. galloprovincialis were compiled into a database to determine its decadal patterns of spread across multiple bioregions. Distribution and abundance (estimated using density and semi-quantitative abundance scale [ACFOR]) of this mussel were surveyed to determine patterns in spatial structure across bioregions. In addition, we compared the size structure of populations at the eastern margins of its range with those of larger populations nearer the range centre.Results: Initial breaching of biogeographic boundaries was associated with rapid spread, but other boundaries encountered decades later acted as barriers to further spread. Across >2,800 km of coast, spatial autocorrelation was observed in densities (low and mid shore levels) and in ACFOR abundances. Repeating spatial patterns in densities and ACFOR abundance were detected at scales of 120-160 km and of 400-990 km. Considerable effort was required to detect populations at the absolute eastern limits of its distribution.Main conclusion: Detection of spatial structures at multiple scales suggests that ecologically determined processes regulate abundance at both intra-bioregional and inter-bioregional scales, which may help tease apart the historic and contemporary consequences of interspecific interactions on the structure of rocky shore communities. This study demonstrates the influence of biogeography in driving temporal patterns of spread and spatial structure on the distribution and abundance of an invasive species.

No abstract

No abstract

We hypothesized congruence in the spatial structure of abundance data sampled across multiple scales for an ecological guild of consumers that exploit similar nutritional and habitat resources. We tested this hypothesis on the spatial organization of abundance of an herbivorous guild of sea urchins. We also examined whether the amount of local along‐shore rocky habitat can explain the observed spatial patterns of abundance. Standardized estimates of abundance of four intertidal sea urchins—Diadema cf. savignyi, Echinometra mathaei, Parechinus angulosus, and Stomopneustes variolaris—were determined by six observers at 105 sites across 2,850 km of coast of South Africa. For each species and observer, wavelet analysis was used on abundance estimates, after controlling for potential biases, to examine their spatial structure. The relationship between local sea urchin abundance and the amount of upstream and downstream rocky habitat, as defined by the prevailing ocean current, was also investigated. All species exhibited robust structure at scales of 75–220 km, despite variability among observers. Less robust structure in the abundances of three species was detected at larger scales of 430–898 km. Abundance estimates of sympatric populations of two species (D. cf. savignyi and E. mathaei) were positively correlated with the amount of rocky habitat upstream of the site, suggesting that upstream populations act as larval sources across a wide range of scales. No relationship between abundance and habitat size was found for P. angulosus or S. variolaris. Within the range of scales examined, we found robust congruence in spatial structure in abundance at the lower, but not the larger, range of scales for all four species. The relationship between abundance and upstream habitat availability in two species suggests that larval supply from upstream populations was probably the mechanism linking habitat size and abundance.

Analyses of taxonomic diversity patterns within coastal systems has been critical in the development of the theory of biogeography. Increasing evidence, however, shows that the variety of functions that species perform in ecosystems (rather than their taxonomic identity) is a better predictor of the influence of the environment on the species. This information has been useful in predictive ecology leading to the development of trait-based approaches (TBA). Until the late 1970s, however, limited effort (particularly in marine systems) was channeled towards patterns in functional species traits and how they may be affected by changes in environmental gradients. Here, I mapped the functional biogeography of the South African coastline based on a suite of species' reproduction and development traits. Because species composition is one of the key tools used by zoogeographers to map species distribution patterns, I expected lower variability in trait composition within main biogeographic regions than in intervening transition zones based on the habitat templet theory and following the biomass ratio and limiting similarity hypotheses. In brief, the habitat templet theory proposes that “the habitat provides a templet upon which evolution forges species characteristics”, while the biomass ratio hypothesis assumes that the most abundant species traits determine ecosystem functioning. The limiting similarity hypothesis also sometimes referred to as the niche complementarity hypothesis, however, predicts that species can coexist if their niches complement one another. In light of the habitat being an evolutionary templet, abiotic and biotic habitat patterns were measured as nearshore SST and chlorophyll-a gradients, respectively. I expected the SST gradient to act as the stronger key filter of trait diversification because temperature is often considered the most influential environmental factor affecting species survival with seasonality of SST affecting the timing of spawning and along with food availability, possibly influencing fecundity. Functional trait data were thus compiled for macroinvertebrate species collected from fifty-two rocky shore sites from three main bioregions (east, south, and west) and two transition zones (south-west and south-east). Biological trait analysis and functional diversity indices were used to evaluate how traits related to species development and reproduction respond to temperature and chlorophyll -a (used as a proxy for food availability) gradients along the coastline. GLMM and hierarchical cluster analyses showed distinct patterns/shifts in SST and chlorophyll-a gradients across bioregions, with two main breaks in SST separating the east and south-east overlap (SEO) bioregions from the south, south-west overlap (SWO) and west bioregions. In contrast, chlorophyll-a exhibited three major breaks with the east, SEO–south–SWO, and west clustering independently of each other. The RLQ analysis (a type of co-inertia analysis) which simultaneously ordinates 3-matrix datasets [i.e., (environment × site[R]), (species × site[L]) and (species × traits [Q])] showed that the higher SST gradient on the east and SEO promoted higher abundance and biomass of simultaneous hermaphrodites while higher chlorophyll-a gradients on the SWO and west coasts strongly promoted reproductive maturity at larger-sizes. The combined fourth-corner analyses showed that the modalities within the development trait domain responding to chlorophyll-a gradients primarily included filter feeders, sessile and swimming species and also species living on the infratidal zone. In addition, the reproduction trait domain showed higher sensitivity and association to differences in chlorophyll-a and SST gradients than development traits. Overall, SST and chlorophyll-a gradients influenced the distribution of the most dominant traits as indicated by shifts in community-weighted mean trait values across bioregions. This suggests the importance of habitat filtering on coastal species reproduction. A separate study evaluating the influence of large-scale biogeographic effects vs the micro-scale biogenic habitat structure offered by coralline seaweeds across 24 sites revealed some notable effects of both factors on the diversity and abundance of macroalgal epifauna. There was a notable biogeographic influence on epifauna, with the SEO recording the highest epifaunal species richness and abundance, followed by the south coast, then the SWO and lastly the west coast. In addition, the total biomass gradient of the corallines followed a similar trend. The epifauna however, showed no host-specificity, illustrating that epifauna may not be species–centric as commonly assumed, and the higher diversity of epifaunal diversity may well be simply because those corallines are the available habitat within the sampled part of the coastline. Lastly, macroinvertebrate trait distribution on the South African coastline confirms that the habitat, particularly the biotic filter (in this case chl-a) provides a templet upon which evolution forges species traits. However, since temperature is a proxy for nutrient availability (cold upwelling brings nutrients), then temperature drives chlorophyll-a. Subsequently this means the abiotic component indirectly drives trait distribution by influencing the biotic environment (chl-a). For epifauna species, also, the coralline diversity and composition can also be regarded as a biotic filter influencing the epifaunal abundances and composition across different bioregions. Moreover, since temperature is regarded as a conservative trait in seaweeds, temperature tolerance defines the biogeographical boundaries of seaweeds, therefore temperature may be indirectly affecting epifauna abundances through coralline species diversity and biomass. In summary, considering the deterministic processes governing ecosystem functioning and community assemblage, the mass ratio and limiting similarity hypotheses showed complementary effects. Different bioregions provided variable support for these two hypotheses, but overall, the mass ratio hypothesis (weighted by species biomass) received stronger support and may be more meaningful to the interpretation of ecosystem functioning and persistence within rocky shore systems. Lastly, although, the SWO showed some of the characteristics of a subtraction zone based on the relatively low abundance, diversity, and biomass measures. Nonetheless, there was evidence of high functional redundancy across all other four bioregions. This suggests that in the context of development and reproduction traits, the rocky shore ecosystem along the SA coastline may be functionally stable at this stage.