While C. macrocephalum (pompon weed) is extensively studied due to its status as an invasive species, existing research has predominantly focused on its potential distribution. These studies have largely relied on climatic variables and have often included considerations regarding two biological control agents (Trethowan et al. 2011). Additionally, its allelopathic potential (Goodall et al. 2010) and the influence of certain environmental variables as facilitators of invasion in South African grasslands (Goodall et al. 2011; Goodall 2016) have been investigated. However, this approach provides a somewhat limited understanding of the species’ ecological requirements. Our research represents the first study which combined ecological niche modeling with ordination techniques to analyze the niche dynamics of C. macrocephalum across both its native and invaded ranges, using climatic and edaphic variables.
Species distribution models are frequently employed to assess the potential impact of non-native species in new habitats, assuming that the climatic niche remains constant (Gallagher et al. 2010). A valuable approach to elucidating the complexity of plant invasion is to contrast global and regional niches (Taucare-Ríos et al. 2016; Zhang et al. 2021). Reciprocal modeling from the invaded range to the native range can assist in identifying the potential origin of the biotypes that invaded a country (Suárez-Mota et al. 2016). The results obtained from the ENMs and RNMs carried out in this study demonstrate that this is an adequate strategy to predict the current potential distribution of C. macrocephalum, since all models tested accurately predicted the presence of the species in the areas where it currently occurs. This implies that environmental variables considered effectively explain the species' geographical distributions. In addition, the models targeted additional suitable areas, mainly in the invaded area. The occurrences predicted by our ENMs largely coincide with the predictions of the ENMs obtained by Trethowan et al. (2011) in the invaded area. Furthermore, the studies agreed on using BIO2, BIO12 and BIO15 as predictor variables for the construction of the models. However, it is important to note that Trethowan et al. (2011) did not perform a correlation of the variables using the climatic data of the occurrence points or based on the background area used for modeling. Their criteria for the selection of variables were based on the study by Beaumont et al. (2009), which was carried out on Hieracium species, native to Europe and native to Australia, New Zealand and North America. In this study, we incorporated edaphic and topographic variables, which contributed a significant percentage to the construction of all models. This allowed us to obtain maps with more accurate predictions. Previous studies have demonstrated the importance of using edaphic variables in the construction of ENM, as they can help explain the ecological preferences of species (Almiron et al. 2022; Xian et al. 2022; Santamarina et al. 2023). However, the majority of the studies on invasive plant species have not yet incorporated edaphic nor topographic variables into their analyses. Conversely, the RNMs have been able to predict the distribution of the species in both the native and invaded areas, as evidenced by the maps generated with the overlapping predictions. This supports the assumption that the environmental conditions occupied by C. macrocephalum in its native range are similar to those occupied in the invasive range. The analysis of the invasion status revealed that the majority of C. macrocephalum populations are in a state of stability. This finding is corroborated by the results of the analysis of niche dynamics, which demonstrated a high proportion of stability.
Ordination techniques such as principal component analysis are adequate for the quantification of niche dynamics (Guisan et al. 2014). These analyses, in conjunction with metrics that estimate niche emptiness, stability, and expansion, plus equivalence and similarity tests, provide an effective tool for predicting the invasive potential of alien species, their capacity to thrive and expand in novel habitats and, consequently, to identify environments or geographic areas most at risk for future invasions (Gallagher et al. 2010; Dinis et al. 2020; Zhang et al. 2021; Aravind et al. 2022). The results of the comparisons of environmental variables and the environmental PCA indicate that the native and invaded ranges of C. macrocephalum evidence similar ecological tolerances. Although the D overlap value between both niches was limited, the individual analysis of the variables showed higher overlap values (low to moderate). The niches demonstrated high conservation along the edaphic variables (PH, BLDFIE and CRFVOL). This type of univariate analysis of niche dynamics has been previously studied on alien herpetofaunal invasions to identify spatially and temporally conserved realized climatic niche components (Liu et al. 2017), It has also proven to be useful in the study of invasive plant species as well (Khuroo et al. 2019). The approach is useful for identifying the factors that limit species’ geographical ranges and for developing a robust predictive framework to investigate rapid shifts of the geographical ranges of both native and non-native species under climate change.
Our findings indicate that invasive C. macrocephalum populations exhibit niche conservatism with respect to native populations, as found for most terrestrial plants (Petitpierre et al. 2012; Aravind et al. 2022). Additionally, our results suggest that the species is not currently occupying all the suitable areas within the invaded range. Although the degree of overlap was relatively low, it is within the range of values found for other invasive plant species (Ahmad et al. 2019; Bello et al. 2020; Aravind et al. 2022; Xian et al. 2022). The low degree of overlap may be attributed to the fact that the area currently occupied by C. macrocephalum is relatively small in comparison to the background we used, which also encompasses a considerable diversity of environmental conditions. The observed environmental differences could be attributed to the unfilling of native niches in the invaded distribution area. However, since unfilling may reflect a future colonization in the introduced range after a certain time has passed, it is accurate to characterize the magnitude of niche shifts focusing on the expansion results (Petitpierre et al. 2012; Strubbe et al. 2013; Li et al. 2014).
Campuloclinium macrocephalum populations having large niche unfilling may be attributed to the short time since introduction in this region (approximately 60 years ago; Henderson and Piwowar. 2006) and also to its so far restricted distribution in South Africa, perhaps due to a single introduction event plus the local biological control programs that are currently being carried out for the species, which would successfully prevent its expansion to new areas. Niche unfilling seems to be larger for species introduced recently and into a small number of locations, compared with those with ancient colonization history and introduced in several points in space (Strubbe et al. 2015). On the other hand, South Africa is identified as invasion hotspots and is one of the few countries that has regulations in place on biological invasions (van Wilgen et al. 2020). Although the results demonstrate that the C. macrocephalum niche is conserved and that the current populations are in a state of equilibrium and little expansion is evident, the large proportion of niche unfilling could indicate an ongoing colonization process with potential for invasion by the species towards new geographical regions. This was evident also in ENM and RNM maps which pointed out suitable areas northwards from its current distribution in South Africa, regions to which special attention should be paid.
The invasion process is dynamic and influenced by environmental factors and agents that facilitate or restrict the invasion process, which collectively determine the likelihood and extent of either conservation or shift of niche. Nevertheless, a great part of the invasive success of a species is attributable to its own characteristics. In this sense, “pompom weed” has evolved adaptations to survive fires and frost via its perennial underground structures, to resist herbivory and diseases (Farco and Dematteis 2014; Goodall 2016); it has the capacity to revert to a dormant state during drought, and it is tolerant to a broad spectrum of growing conditions, among other factors (see Henderson et al. 2006 for a list of the characteristics that make C. macrocephalum succeed as a weed). What is more, studies conducted in the grasslands of Gauteng Province (South Africa), where C. macrocephalum is most abundant, showed its preference for environments disturbed by heavy grazing, abandonment of agricultural lands, and modification through wetland drainage (Goodall 2016). Also, the plant is unpalatable to livestock, leading to increased grazing pressure on the remaining vegetation, and accelerating veld degradation, which enables its establishment and spreading in grasslands disturbed by different agents that reduce the basal grass cover and exacerbate soil erosion. Considering this, the main invasive strategies of C. macrocephalum entail its biological traits which, combined with environmental conditions, anthropogenic disturbances, and the lack of natural enemies in the invaded range, have allowed the species to thrive and succeed.