The total number of naturalized species can be a reliable predictor of the number of alien pest species


In her recent paper, Shawna J. Dark (2004) concluded that the distributions of ‘invasive alien plants’ and ‘non-invasive alien plants’ in California were significantly correlated.

She found the highest numbers of both ‘invasive’ and ‘non-invasive’ species in regions with low elevation, high road density, and high native species richness. Dark actually uses two definitions of ‘invasive’ in the paper; we discuss the implications of this on her conclusions. We point out that the most important message of Dark's paper is that the total number of naturalized species can be, at least at some spatial scales, a reliable predictor of the numbers of exotic pest species. We also provide another example to support this conclusion. We follow the definitions of native, non-native, naturalized, casual, pest, and weed in Pysek et al. (2004) in this note unless otherwise specified.

It is necessary to examine how Dark defined ‘invasive’ in her paper because the data she used include only a small number of plant species (78) classified as ‘invasive’ while the remaining 1097 naturalized species listed in the most recent statewide flora and online updates (total 1175; Hickman, 1993; were classified as ‘non-invasive alien plants’. Dark first defines ‘invasive alien plants’ in the paper's second sentence as ‘alien plants that produce offspring in very large numbers and at large distances from parent plants and thus have potential to spread over a considerable area (Richardson et al., 2000)’. Later, however, we learn that species actually classified as ‘invasive’ for data analysis in this study are only those rated as ‘A’ and ‘B’ on the California Exotic Pest Plant Council's (1999) list of ‘Exotic Pest Plants of Greatest Ecological Concern’ (Dark, 2004; text under Fig. 1; CalEPPC, 1999). CalEPPC (1999) in turn defines its A and B categories as covering species that displace native species and disrupt habitats. Dark notes that this list is a work in progress and not yet complete, particularly for grasses but, using only this list to define invasive species, she states that ‘in the state of California there are over 1000 documented alien plant species, and an estimated 6% of these are considered to be invasive (Hickman 1993)’. Hickman (1993), however, does not list such a percentage and in fact does not use the term ‘invasive’. It is worth noting that far more alien plant species have been documented in California but that most of them grow only in cultivation. Hickman covers only alien plants that are naturalized plus some casual alien species.

Figure 1.

Relationship between the number of ‘noxious’ non-native weedy species and the number of remaining non-native naturalized or casual plant species in the 50 US states (P < 0.001). Note that the regression function starts to be meaningful only for X-values greater than 200. As both variables are negatively correlated with the area of individual states (see Rejmánek (2003) for explanation), the positive correlation between Y and X is not driven by species-area effect. Data: Kartesz & Meacham (1999).

The 78 species from the CalEPPC list used in the paper include many serious exotic pests in natural areas and rangelands. However, they represent only a subset of what are sometimes called ‘environmental weeds’. These 78 species are undoubtedly invasive (i.e. they are spreading in California), but this list also does not include many alien (and invasive) agricultural, garden and yard weeds and many others invasive (spreading alien) species in California.

Recently, Rejmánek (2000) estimated that about 650 naturalized plant species were invasive (spreading) in California. Even this number may be an underestimate since the most recent manuals of Californian weeds that explicitly cover all species regarded as pests in agriculture, forestry, rangelands, natural areas, natural and artificial waterways, yards and other urban settings in the state (DiTomaso & Healy, 2003, 2005) cover 701 non-native species in the main text (J. DiTomaso, pers. comm.). Even if not all alien weeds are invasive (spreading), most of them are. On the other hand, some invasive (= spreading non-native) species do not cause any detectable environmental or economic damage and therefore are not called weeds or pests (Richardson et al., 2000). To sum up, it would be safe to say that among the 1000–1300 naturalized plant species in California (Hickman, 1993; Rejmánek & Randall, 1994; Hrusa et al., 2002), at least 60% are invasive as defined in the second sentence of Dark's paper, and less than 40% can be classified as non-invasive.

Unfortunately, terminological confusions are common in and between research, management and public policy publications and discussions regarding invasive species. These terminological differences have the potential to lead to confused or incorrect data interpretation, poor communication between different groups of researchers, and inappropriate management and public policy decisions (Pysek et al., 2004).

Despite the problems outlined above regarding the exclusion of many invasive species in Dark's analysis, her results are very interesting—and for reasons in addition to those advanced in her paper. Dark's study indicates that the total number of naturalized species can be, at least at some spatial scales, a reliable predictor of the numbers of exotic pest species. Therefore, total numbers of naturalized species, as we are getting them from recently published manuals and checklists, seem to be good indicators of present or future problems with exotic pests.

Dark's results probably have more general validity. Here we present another example from North America. We used non-native species that are listed as federal and/or state-level (listed in at least one state) ‘noxious weeds’ as a surrogate for exotic plant pest species in general. We found that the number of species in this group is significantly positively correlated with the number of remaining non-native naturalized and casual alien species in the US states (Fig. 1). The category non-native ‘noxious weeds’ (Kartesz & Meacham, 1999) includes 304 exotic pests that are difficult to manage (e.g. Alternathera philoxeroides, five Carduus species, 12 Centaurea species, Cortaderia jubata, eight Cuscuta species, Cyperus rotundus, Cytisus scoparius, eight Euphorbia species, Heracleum mantegazzianum, Hydrilla verticilata, Lepidium latifolium, Lonicera japonica, Melaleuca quinquenervia, five Pennisetum species, Pueraria montana (P. lobata), four Salsola species, Salvinia molesta, Silybum marianum, eight Solanum species, two Striga species, Taeniatherum caput-medusae, seven Tamarix species, Triadica sebifera (Sapium sebiferum), Ulex europaeus, Viscum album). However, this list of species is somewhat biased toward agricultural weeds and it does not include some important exotic pests of natural areas (e.g. Ailanthus altissima, Arundo donax, Foeniculum vulgare, Lonicera maackii, Lygodium spp., Microstegium vimineum, Rubus armeniacus). On the other hand, the fact that a species is declared as ‘noxious’ in one or more states does not mean that that species is a serious pest in all states where it is present. Therefore, ‘noxious weeds’ in Kartesz & Meacham (1999) is still not an ideal category for this analysis, but we believe that it includes, at least, a representative sample of exotic pests. Using this data and the regression equation in Fig. 1, we can conclude that 15–30% of naturalized species can be classified as ‘pests’ in the US states. Obviously, this is substantially larger proportion than 10%, the number proposed by Williamson (1996).

The main mechanism behind correlations described above is likely just a simple sampling effect. With increasing numbers of introduced species, increasing numbers of species become naturalized, and, with increasing number of naturalized species, there is an increasing probability that a large number will cause some environmental or economic damage. Most naturalized species are incorporated into native floras without any serious consequences. They are concentrated mainly in disturbed areas and very often are just symptoms of human-created disturbances rather than causes of environmental changes. In general, however, a high numbers of naturalized species represents a red flag: problems with at least 15% of those species.