Evidence of invader decline
To determine the prevalence of novel pathogen accumulation on invasive plants, we compiled examples of plant invader disease from the literature. We searched Web of Science with the following terms: (”plant inva*” OR ”invasive plant*”) AND (dieback* OR disease* OR pathogen* OR infect*) on March 16, 2020. This search produced 661 results. References cited in articles found through this search were also analyzed. The following conditions were required for inclusion: (1) the plant species must be considered non-native invasive (based on the CABI database or local invasive plant lists); (2) disease must have been observed on the invader in its non-native range; (3) the cause of the disease must have been identified by inoculating the invader with the pathogen and observing disease symptoms; and (4) the disease agent must be native to the non-native range of the plant invader or globally distributed. Because information on the natural ranges of plant diseases can be scarce, we included plant diseases where the available information suggests they are likely native, newly identified taxa on invaders in their non-native ranges, as well as globally distributed pathogens of unknown origin. While globally distributed pathogens may have encountered invasive plants in their native range, they are not subject to stochastic introduction processes as they are already widely distributed.
We found 20 invasive plant-pathogen combinations with 16 different invasive plant species experiencing native/novel pathogen effects from 17 unique pathogens (Table 2). The species of plants came from a wide variety of functional groups including grasses (3), forbs (6), shrubs/vines (4), leguminous shrubs (2), and a tree (1). Of these records, 13 were from North America, 2 from South Africa, 3 from Europe, and 2 from Australia. Seven of these examples had globally distributed pathogens of unknown origin and 13 were native or newly described pathogens. We recorded the approximate year of introduction of the plant invader and the year the pathogen was first observed on the invader. The amount of time for disease to develop on the invaders ranged from ~60 years to over 200 years, with the average approximately 120 years. This range is consistent with evidence of reduced enemy release after similar residence time (Hawkes 2007). It is possible that potential invaders that are impacted by native diseases after short residence time while still in the lag phase do not become problematic invaders.
Some of the pathogens in our survey had major impacts on invader populations in the field such as Pseudolagarobasidium acaciicolaon Acacia cyclops , while others showed high levels of pathogenicity in laboratory trials but had low to moderate effects in the field such as Erysiphe cruciferarum on Alliaria petiolata (Cipollini et al. 2020). There are a variety of reasons that effects in the field could be lower than in laboratory trials. For example, pathogen inoculum density in the field may be too low to cause severe disease symptoms (De la Cruz et al.2018), the pathogen may have poor dispersal ability in the field preventing widespread impacts on the invader, the pathogen may only cause symptoms in combination with certain environmental conditions making its impact variable either spatially or temporally (Aghighi et al.2014), or there may be differences in population level susceptibility of the invader (Meyeret al. 2001). While we focus primarily on interspecific differences in invader traits, it is also important to recognize that differences in population level susceptibility are also likely related to plant traits (Cipolliniet al. 2020). In some of the cases we identified, the pathogen is thought to be vectored by insects. For example, rose rosette disease is vectored by mites (Pemberton et al.2018) and Pseudomonas syringae pv. syringae disease on C.stoebe ssp. micranthos and Botryodiplodia theobromae onMimosa pigra require wounding to infect plants and are thought to be associated with introduced biocontrol insects (Wilson & Pitkethley 1992; Kearing 1996; Kearing et al. 1997). Plant traits related to defense against herbivores could also influence the interactions between invaders and insect vectors.
On the whole, we found relatively few examples of novel pathogen accumulation compared with the total number of invasive plant species, but our review also highlights areas where further research is needed. Additional cases of novel pathogen accumulation may exist that are not recorded in peer-reviewed literature, or are excluded from our review because, for example, it is difficult to identify the cause of decline or because of the limited information on the natural ranges of microbial pathogens (Aghighi et al. 2014). There are also strong geographic biases in the study of invasive plants and there could be additional cases of novel pathogen interactions in understudied regions (Pyšek et al. 2008). With these limitations in mind, how can we better allocate resources to predict what invasive species may be more susceptible to decline?