1. Introduction

Human activities have caused the exchange of species among different parts of the world. Over 13, 000 species of the global vascular plants have naturalized beyond their native ranges (van Kleunen et al., 2015). When naturalized non-native plant species spread widely and cause negative impacts on the environment and human societies, they become invasive (Richardson et al., 2000; Blackburn et al., 2011; Lambertini et al., 2011; Simberloff et al., 2013). The number of naturalized species that have become invasive is substantial (McGeoch et al., 2010) and is rapidly increasing (Catford et al., 2009; Qian et al., 2022a). Invasive species may pose significant threats to biodiversity and ecosystem structure and functioning (Levine et al., 2003; Rejmánek et al., 2013; Gaertner et al., 2014; Kerns et al., 2020). Identifying potentially invasive species and preventing their expansion are of critical importance in invasion ecology (Qian et al., 2022a). Because species with shared ancestry are ecologically more similar to each other than to distantly related species (Cavender-Bares et al., 2009), and previous studies have shown that many functional traits are phylogenetically conserved (Donoghue, 2008; Ackerly, 2009), phylogenetic relatedness among species in a regional flora could be used as a proxy of similarity in functional traits among the species (Webb, 2000; Zhang et al., 2011; Krishna et al., 2021). Phylogenetic relationships of naturalized (particularly naturalized invasive) species with native species, and phylogenetic relationships of naturalized invasive species with naturalized non-invasive species, have become a focal study component in ecology (e.g. Park et al., 2020; Omer et al., 2022; Qian et al., 2022a; 2022b).

Previous studies on the phylogenetic relatedness of plants at the transition from naturalization to invasion have shown mixed results. For example, Zhang et al. (2021a); Qian et al. (2022a), and Qian (2023) showed that invasive angiosperm plants are a phylogenetically clustered subset of overall naturalized angiosperm plants in China. In contrast, Omer et al. (2022) found that the probability of becoming invasive is higher for naturalized species distantly related to the native flora. These opposing conclusions may result from the use of different methods in different studies. For example, the aforementioned three studies on the China's flora compared invasive species in a region with naturalized species or all (i.e. native plus naturalized) species within the region. In contrast, Omer et al.'s (2022) finding is based on an analysis comparing 524 naturalized species in the country of South Africa, 310 of which are considered invasive, with angiosperm species worldwide. Phylogenetic relatedness of each non-native species in Omer et al.'s study was calculated with respect to the species pool including global angiosperms. Because many, or most, of the angiosperm species of the world that are non-native to South Africa may not become naturalized in South Africa due to ecological constraints even if they are introduced into the country (e.g. a large number of plant species restricted to tropical rain forests), Omer et al.'s analysis may not be appropriate to address the question whether the probability of becoming invasive is higher for naturalized species closely or distantly related to the native flora. Furthermore, because Omer et al. did not compare the naturalized and invasive species of each region with the overall (i.e. native plus naturalized) species or native species within the region based on the regional species pool, which is an appropriate and straightforward approach to determine the phylogenetic relationship between non-native invasive species and native species, their conclusion needs to be tested.

Here, I use the same methods to investigate phylogenetic relatedness of invasive species and overall naturalized angiosperm species in South Africa and China. First, using the same set of naturalized and invasive angiosperm species in South Africa as in Omer et al.'s analysis, I compare phylogenetic relatedness of naturalized species with that of invasive species; in this analysis, phylogenetic relatedness for each region is quantified using each regional species pool including native and naturalized species. Second, I determine phylogenetic relatedness of invasive species with respect to each regional species pool including native and naturalized invasive species; this analysis allows determining whether invasive species are closely or distantly related to native species. Last, I compare phylogenetic relatedness of invasive species with that of native species in each regional angiosperm flora.

Similarly, I conduct a set of analyses for regional angiosperm floras in China. I include China in this study partly because both native and naturalized plants in China have been well documented and studied (e.g. Ma and Li, 2018; Qian et al., 2019, 2022a, 2022b; Hao and Ma, 2023), and partly because China covers long climatic gradients, which allow investigation of phylogenetic relationships between native and non-native species in different climatic regions (from warm and humid climates to cold and arid climates).

2. Materials and methods

As in Omer et al. (2022), South Africa was divided into nine regions. For each of the nine regions, I extracted the presence-absence data for the 524 non-native angiosperm species used in Omer et al. from the GloNAF database (van Kleunen et al., 2019) and the presence-absence data of the native angiosperm species from Germishuizen and Meyer (2003). I standardized botanical nomenclature according to The Plant Lists (v.1.1; http://www.theplantlist.org/), using the package U.Taxonstand (Zhang and Qian, 2023), and combined the occurrences of an infraspecific taxon with those of its species. As a result, 17, 850 species from South Africa were included in my analysis.

China was divided into 28 regions, as in Qian et al. (2022a). Species lists of native, naturalized, and invasive non-native plants for the regions that were used in this study were compiled by previous studies (Qian et al., 2019, 2022a; Qian and Qian, 2022) based on various sources, including, but not being limited to, China's flora (eFlora for China, http://www.efloras.org/flora_page.aspx?flora_id=2), Ma (2013); Ma and Li (2018), and Hao and Ma (2023). The definition of invasive plants in China is similar to that of Zengeya and Wilson (2021) for invasive plants in South Africa; that is, invasive species are those non-native (alien) species that sustain self-replacing populations over several life cycles, produce reproductive offspring commonly in very large numbers far from the site of introduction, and have the potential to spread over long distances.

I generated a phylogenetic tree for the angiosperm species in South Africa and China based on the time-calibrated phylogeny of Smith and Brown (2018), using the package U.PhyloMaker (build.nodes.1, Scenario 3; Jin and Qian, 2022, 2023). Previous studies have shown that plant phylogenies generated based on Smith and Brown's time-calibrated phylogeny are robust for studies of community phylogenetics (Qian and Jin, 2021). Phylogenetic trees generated by U.PhyloMaker or its sister packages (Qian and Jin, 2016; Jin and Qian, 2019, 2022) have been commonly used in studies on phylogenetic diversity and structure in regional and global floras (e.g. Qian et al., 2019; Yue and Li, 2021; Zhang et al., 2021a; Zhang et al., 2021b; Qian, 2023; Qian and Deng, 2023; Zhou et al., 2023). The phylogenetic trees used in Omer et al.'s (2022) and my studies were both built based on the time-calibrated megatree generated by Smith and Brown (2018). About 73, 200 angiosperm species in 10, 364 genera were included in the time-calibrated megatree generated by Smith and Brown (2018) when plant names were standardized according to The Plant List; thus, about 23% and 75% of species and genera, respectively, were resolved in the phylogenetic trees used in Omer et al. (2022). In contrast, 39% and 86% of the 47, 605 species and 4509 genera, respectively, in the phylogenetic tree used in the present study were included in the time-calibrated megatree generated by Smith and Brown (2018). Therefore, the phylogenetic tree used in my study was better resolved at both species and genus levels than that used in Omer et al. (2022).

I calculated two commonly used metrics of phylogenetic relatedness, net relatedness index (NRI) and nearest taxon index (NTI) (Webb et al., 2002), which reflect phylogenetic relatedness at deep and shallow evolutionary history, respectively. For both NRI and NTI, a greater value indicates a greater phylogenetic relatedness (i.e. greater phylogenetic clustering). For each region in South Africa and China, NRI and NTI for both naturalized and invasive species were calculated with respect to the species pool including both naturalized and native species in the region. In addition, for each region, I also calculated NRI and NTI for invasive and native species with respect of the species pool including both native and naturalized invasive species in the region (i.e. excluding naturalized non-invasive species from the region). Species richness in a region differed by one order of magnitude between invasive and native angiosperm species (on average, 160 versus 4428 species for invasive and native angiosperms, respectively, per region). To minimize the effect of difference in species richness of invasive and native angiosperms in each region on values of phylogenetic relatedness, I used a rarefaction approach to account for difference in species richness of invasive and native angiosperms in each region. Specifically, for a particular region, I randomly selected a subset of native species from the species pool of the region angiosperm flora by setting the number of native species in the subset to be equal to the number of invasive species in the region; I repeated this 99 times; I calculated NRI and NTI for each of the 100 randomly selected assemblages of native species, and used the mean values of NRI and NTI for the randomly selected assemblages to represent NRI and NTI of the native angiosperm flora of the region.

I used the package PhyloMeasures (Tsirogiannis and Sandel, 2016) to calculate NRI and NTI. With PhyloMeasures, both indices were computed using computationally efficient algorithms described in Tsirogiannis et al. (2012, 2014). Specifically, PhyloMeasures calculates NRI and NTI based on exact solutions given a particular tree and species richness. Defining these statistical moments requires a null model and a defined species pool. The null model used in the package considers all possible combination of S species from the species pool (where S is the richness of a sample to be standardized) to be equally likely (Tsirogiannis and Sandel, 2016). With the above-described approaches, for a given phylogenetic metric (e.g. NRI), phylogenetic relatedness is quantified with a single value of the phylogenetic metric derived from a null model. The above-described approaches have been commonly used in studies on phylogenetic relatedness (e.g. Webb, 2000; Webb et al., 2002; Lu et al., 2018; Yue and Li, 2021; Zhang et al., 2021b), including those for non-native naturalized and invasive plants (e.g. Sandel and Tsirogiannis, 2016; Park et al., 2020; Zhang et al., 2021a; Qian et al., 2022a, 2022b; Qian and Sandel, 2022; Qian and Deng, 2023).

3. Results and discussion

My analysis showed that when NTI was considered, phylogenetic relatedness was greater for invasive species than for naturalized species in three of the nine regions in South Africa, consistent to Omer et al.'s finding. However, phylogenetic relatedness was smaller for invasive species than for naturalized species in five of the nine regions in South Africa when NRI was considered (Fig. 1), suggesting that the probability of becoming invasive is higher for naturalized species closely related to the native flora, contrary to Omer et al.'s finding. Because NTI represents phylogenetic relatedness at the tip branches of the phylogenetic tree of a regional species pool whereas NRI represents phylogenetic relatedness across the whole phylogenetic tree of the regional species pool, and thus phylogenetic relatedness based on all evolutionary depths of the phylogenetic tree, NRI is more appropriate for quantifying overall phylogenetic relatedness of a community, compared to NTI. The present study found that NRI and NTI of invasive species in regional floras in South Africa were significantly greater than zero when they were derived with respect to regional species pools including native and invasive species (Fig. 2). The results of the present study for South Africa suggest that the probability of becoming invasive is higher for naturalized species closely related to the native flora.

For the 28 regions in China, when NRI and NTI were calculated with respect to regional species pools including native and all naturalized angiosperm species, phylogenetic relatedness was greater for invasive species than for naturalized species in 13 regions when NRI was considered, and in all the 28 regions when NTI was considered (Fig. 1). This suggests that the probability of becoming invasive is higher for naturalized species closely related to the native flora in most (73%) cases when the results of NRI and NTI were considered together. When NRI and NTI were calculated with respect to regional angiosperm species pools including native and invasive species, they were much greater than zero, with mean values being 8.597 and 6.451 for NRI and NTI, respectively (Fig. 2).

For regional angiosperm floras in both South Africa and China, phylogenetic relatedness for invasive species was, on average, much greater than that of native species, as shown in Fig. 3. When NRI was considered, 36 of the 37 regions in South Africa and China had greater values of NRI for invasive species than for native species (Fig. 3). Similarly, when NTI was considered, 35 of the 37 regions in the two countries had greater values of NTI for invasive species than for native species (Fig. 3). The difference in phylogenetic relatedness between native and invasive species was greater for China than for South Africa (the average difference between invasive NRI and native NRI was 9.06 and 1.45 in China and South Africa, respectively; the average difference between invasive NTI and native NTI was 6.21 and 0.73 in China and South Africa, respectively). When the 37 regions in South Africa and China were considered together, on average, NRI and NTI of invasive species were approximately 7 and 5 times greater than those of native species, respectively (Fig. 3). These results are consistent with those of previous studies that naturalized and invasive plant species in a biological community are generally a subset of phylogenetically clustered species with respect to all (i.e. native plus naturalized) plant species in the community (e.g. Park et al., 2020; Zhang et al., 2021a; Qian et al., 2022a, b; Qian and Sandel, 2022).

Taking the results of the present study for South Africa and China together, the majority (66%) of the 74 cases (i.e. 37 regions by two phylogenetic metrics) showed that the probability of becoming invasive is higher for naturalized species closely, rather than distantly, related to the native flora when phylogenetic relatedness is quantified with species pools including native and all naturalized angiosperms. When phylogenetic relatedness is quantified with species pools including native and naturalized invasive angiosperms, invasive species are phylogenetically clustered, with respect to native species, in regional floras. Overall, these findings suggest that the probability of becoming invasive is higher for naturalized species closely related to the native flora, which is consistent with Darwin's preadaptation hypothesis (Qian and Sandel, 2022). The conclusion of the present study is consistent with those of two recent studies (Zhang et al., 2021a; Qian et al., 2022) showing that invasive angiosperm species are a phylogenetically clustered subset of naturalized angiosperm species in China.

In summary, based on both tip-weighted and basal-weighted metrics of phylogenetic relatedness, this study found that invasive species are phylogenetically more strongly clustered, compared to both naturalized species and native species, in both South Africa and China. This finding appears contrary to the finding of Omer et al. (2022), although the same data set for South Africa was used in the two studies. The reason for the contrary results to emerge is likely due to the use of different species pools in Omer et al.'s and my studies. In my study, every species included in a regional flora is either native or naturalized in the region, and a phylogenetic relatedness metric for the region was calculated based on the species of the region. This approach of calculating phylogenetic relatedness metric is commonly used in the literature, including studies on non-native species (e.g. Park et al., 2020; Zhang et al., 2021a; Qian et al., 2022a, b; Qian and Sandel, 2022). In contrast, as mentioned above, phylogenetic relatedness of each non-native species in Omer et al.'s study was calculated with respect to the species pool including angiosperms worldwide. Because many, or most, of the angiosperm species of the world that are non-native to South Africa, particularly a great number of plant species which are restricted to tropical rain forests, may not become naturalized in South Africa due to ecological constraints, let alone for them to become invasive in South Africa, even if they are introduced into the country, including those species in the species pool used in Omer et al.'s analysis might have biased the results and conclusions of their study.

The finding of the present study based on two large national floras in different continents could have significant implications in invasion ecology and conservation biology. For example, one can anticipate which non-invasive naturalized species might become invasive over time and the degree to which they might negatively impact the invaded communities, based on phylogenetic relatedness between invasive naturalized species and those naturalized species that have not become invasive yet and the degree of invasiveness of each invasive species (Qian et al., 2022a).

Acknowledgements

I am grateful to the subject editor and anonymous reviewers for their constructive comments on the manuscript.

Author contributions

H.Q. designed the study, analyzed the data, and wrote the manuscript.

Data availability

Data used in this study have been published.

Declaration of competing interest

The author declares no conflict of interest.