7 Discussion

We have shown that over the past decades a broad range of different research approaches has emerged in order to produce systems, target and transformation knowledge for invasive species management. We can ultimately not judge if current invasion research produces effective knowledge for problem-solving, because this has to be assessed through deliberations among the experts and stakeholders involved in the issue (boundary management). However, we summarize in the following how invasion research has evolved in the past and point out potential future transformations of the research field. Then, we discuss the forms of boundary management that were (not) in place to frame adequate research questions about processes, values and practices for effective problem-solving. We conclude by suggesting a number of achievements and limitations of how problem-orientation evolved in invasion research, and relate these to the forms of boundary management that were in place or missing.

7.1 The evolution of invasion research: The search for an adequate framing of problem-oriented research on biotic invasions

The evolution of research on systems knowledge

Since the emergence of the research field of invasion biology in the late 1950s, a main motivation behind the research of most involved scientists was to contribute to the solving of a perceived societal problem (Davis, 2006*). In particular, the invasion classic by Elton (1958) and the first SCOPE research program (Drake et al., 1989*) had a clear conservation focus. Initially, most research was focused on producing systems knowledge. Thereby the assumption was that the causal dynamics of biotic invasions could not be fully understood by simply applying established ecological principles. Rather than initiating, for instance, landscape ecological research on invasion spread, novel guiding research questions were defined that were thought to be adequate for problem-solving. In the beginning these questions focused on the traits of problematic invasive species (invasiveness), and the characteristics of habitats that are vulnerable to invasions (invasibility) (“classical model”, Table 1). With time it proved that the focus of these initial guiding questions was too narrow, and new aspects had to be addressed, that were integrated into the established framework of invasion research (e.g. “phase transition models”, Table 1). Currently, concepts such as invasive species, invasiveness, invasibility, propagule pressure and invasion phases provide a theoretical framework that is rigid enough to sustain biological invasion research as an independent research field, i.e., research papers on biotic invasions address primarily such invasion specific concepts, rather than, for instance, general ecological principles and their application to biotic invasions.

The future development of the research field will reveal whether research on biotic invasions will be able to maintain a separate niche besides general ecological research or other related research fields such as weed science, restoration ecology, global change biology or conservation biology, which increasingly also study biotic invasions from their perspectives. For instance, it has recurrently been argued that invasion research is artifically dissociated from general ecology, and the need for a separate conceptual framework has been questioned (Davis, 2006*; Davis et al., 2001). In fact, general ecological research on biotic invasions has recently gained momentum (“natural experiments”, Table 1). It is not evident whether invasions of alien species will in the future still be considered a separate issue from other colonisation and migration processes that occur with global change. Thereby the fate of the research field will on the one hand depend on how well biological invasion research will be able to synthesise existing knowledge and integrate neglected processes, such as spatial processes (compare “landscale ecology”, Table 1), into an invasion specific theoretical framework. On the other hand, it will depend on the assessment by the relevant experts and stakeholders of the adequacy of current biological invasion research for problem-solving. Whether biotic invasions are studied as unique processes, or based on the frameworks of general ecology, weed science, restoration ecology or conservation biology, respectively, will have important ramifications on how the problem of invasive species is valued and managed.

In the early 2000s, social sciences research revisited earlier research (see Kowarik, 2003*) on the role of human agency in biotic invasions (“vector science” and “land use science”, Table 1), and this has further dynamised the framing of research questions on the causal dynamics of biotic invasions. It proved that human agency is a critical factor in biotic invasions, which challenged the legitimacy of purely biological invasion research. The future of invasion research will show to what extent the integration of social sciences research will transform the field. It may be that existing research fields at the boundary of ecological and social sciences such as urban ecology (cf. Davis, 2006), ethnoecology (Clayton, 2003; Drake and Hunt, 2007; McDowall, 1994), or research at the boundary between epidemiology and ecology (compare Wilcox and Kueffer, 2008) may contribute to an integrative framework. It has been suggested that social sciences research on land-use change may have the potential to interpret many ecological global change phenomena including biotic invasions from a social science perspective (Jay and Morad, 2006; Robbins, 2004; Schneider and Geoghegan, 2006). Again, such transformations of the research field would not only have consequences for research but also for problem-solving.

The evolution of research on target knowledge

The valuation of impacts of invasive species was initially based only on biological research on ecological effects of invasive species on ecosystem properties and native biota (“biological impact research”, Table 2). This was problematic because normative statements were based solely on empirical studies without explicitly discussing the underlying normative assumptions. It was basically assumed that any strong effect or even simply the presence of an alien species is per se problematic. Biological impact research was only partly able to differentiate whether invasive species are a driver or result of ecological change (Didham et al., 2005), rarely considered potential positive effects, and was not in a position to develop a framework that allowed weighting positive and negative impacts. More recently, concepts from other research fields have been applied to the issue that allow for an explicit valuation of impacts. These inputs came from four different backgrounds, namely philosophy (“native/alien debate”, Table 2), risk research (“risk assessment”, Table 2), (ecological) economics (“economic valuation”, Table 2), and social sciences research (“socioeconomic valuation”, Table 2). However, these imported concepts have been applied to the issue without considering the particular challenges involved in valuating invasive species, but valuation concepts need to be tailored to the context of a particular issue (compare e.g. Brun and Hirsch Hadorn, 2008). Conceptual debates about the valuation of invasive species (“native/alien debate”, Table 2) have not led to a new conceptual understanding of valuation beyond criticising current normative thinking. In contrast, social sciences research in the field of restoration ecology has worked towards a new normative thinking about attempts to restore or design nature (e.g. Gobster and Hull, 2000; Higgs, 2003). Invasion research on target knowledge awaits new integrative concepts that synthesise ecological knowledge and normative thinking.

The evolution of research on transformation knowledge

The production of transformation knowledge is characterised by a tension between on the one hand producing technical knowledge and “tools” that are tailored to the existing options of actors, and on the other hand identifying strategies that allow to fundamentally transform the management context. In the first case, problem-oriented research is typically applied research in the sense that research questions are based on an established understanding of the management issue and on the existing institutional framework. To date, much of the invasion research on transformation knowledge is such applied research. Based on a conceptual model that structures management into the three phases of prevention, early detection and eradication, and control, management tasks are divided up among existing agencies and these in turn define their immediate research needs based on existing invasion research and expertise (Table 3). For instance, prevention depends on effective risk assessment systems and techniques to detect new potentially invasive alien species at borders or minimize secondary releases of already introduced invasive species in the region, or credible target knowledge to inform public awareness campaigns.

In contrast, research on transformation knowledge may also target fundamental changes of the current framing and institutional setup of invasion species management. However, it seems that invasion research has so far not explicitly investigated if current management approaches are appropriate or if they need to be reconsidered fundamentally. Although it has recurrently been emphasised that current institutional setups are not appropriate to deal with invasions that happen on multiple spatial scales (Lodge et al., 2006*), there has been no research on how to fundamentally adapt governance of invasive species issues in contrast to other environmental issues (e.g. Folke et al., 2005; Hirsch Hadorn et al., 2002). Invasion research on transformation knowledge may profit from transferring institutional innovations from other management fields to the issue of invasive species. For instance, rapid response teams that have been in place for a long time to respond to emergencies such as wild fires, chemical, biological or radioactive spills, or medical emergencies have been discussed as a model for invasive species eradication programs (Curt Daehler, pers. comm.).

The co-production of systems, target and transformation knowledge

We have discussed research on systems, target and transformation knowledge separately, and we think that this decomposition of research questions is useful because the three forms of knowledge are categorically different. However, we have also emphasised that the three forms of knowledge are interrelated. Innovation in one realm has consequences for the other knowledge forms. For instance, a move towards landscape-scale research on systems knowledge may initiate a shift towards large-scale management approaches. Alternatively, innovations in research on target knowledge will influence stakeholders’ valuation of invasive species, which will feed back on research on systems knowledge that attempts to understand human motivation in order to explain invasion processes.

In a number of cases, close affinities between research on systems, target and transformation knowledge have evolved (compare Tables 1, 2, and 3). For instance, research on plant invasiveness is closely linked to the development of risk assessment systems. Biological impact research, economic valuation and awareness building form another cluster. Bioeconomic modelling builds on (socio)economic valuation, landscape ecology of invasion spread and transformation knowledge related to control techniques. Finally, a particularly interesting case is the emergence of vector science because it represents at once a fundamental transformation of research on systems and transformation knowledge.

Contextualisation – A fundamental process for enhancing problem-orientation

Across all three types of knowledge – systems, target and transformation knowledge – research has with time become more specific to particular contexts in order to reduce the scientific and management complexities. Research on systems knowledge has moved from identifying general traits of invasive species and invaded habitats (“classical model”, Table 1) to study particular invasion phases (“phase transition models”, Table 1), transport pathways (“vector science”, Table 1), or invasion cases (“multifactorial case studies”, Table 1). Research on target knowledge has shifted from a general to a context-specific assessment of the impacts of an invasive species. For instance, risk assessments became more specific to particular transport pathways, regions or habitats. Economic valuation has focused increasingly on supporting particular management decisions. The general discussion about the validity of the alien/native dichotomy has gained clarity through research on the socioeconomic valuation of particular management cases. In the case of transformation knowledge, prevention shifted from general to pathway specific approaches, and control efforts are designed for particular species and habitats.

7.2 Forms of boundary management

The adequacy of the framing of research questions for problem-solving in invasion research has been continuously debated in the field. A wealth of concept articles has been published, as well as articles that discuss the relevance of research for management and the priority needs for future research (e.g. Drake et al., 1989; Lodge et al., 2006; Mack et al., 2000; Meyerson and Mooney, 2007*; Nentwig, 2007; Simberloff et al., 2005). However, these articles were mostly written by groups of (academic) scientists, and were not based on explicit boundary management among non-academic and academic experts and stakeholders. There were also attempts to assess the scientific quality of invasion research through peer-review processes that consider the outreach to management (MEA, 2005; Roberts and Pullin, 2007; Wilson et al., 2007). But again, these were intra-scientific assessments, and primarily the credibility and to some extent the relevance of research but not its legitimacy were considered.

More generally, boundary management in invasion research has happened so far mostly through expert-driven processes, or informal and unstructured interactions among experts and stakeholders (e.g. communities of practice, trading zones; see below), while formal processes of boundary management (e.g. transdisciplinary research, participatory processes; see below) have been rare. The main informal and formal processes of boundary management in invasion research are shortly introduced in the following.

Informal processes of boundary management

A strength of invasion research are long-term informal but continuous interactions that have evolved among scientists and managers. Such interactions among scientists and practitioners with different backgrounds but a common interest in a thematically restricted domain of problem-solving issues are sometimes called communities of practice (Roux et al., 2006). The World Conservation Union (IUCN) for instance has established an Invasive Species Specialist Group that is composed of a wide range of experts both from academia and management and that interacts daily through a public mailing-list (Clout and Poorter de, 2005*). The epistemological interactions in such communities of practice are often facilitated through boundary objects (Cash et al., 2003*; Hellström and Jacob, 2003*; Pohl and Hirsch Hadorn, 2007*). A boundary object is an epistemological object that is plastic enough so that different experts can interprete it differently for their epistemological context, yet robust enough to maintain a common meaning among different expert groups (Star and Griesemer, 1989). The concept of an invasive alien species is a good example of a boundary object. Different experts have strongly differing and even contradictory understandings of an invasive alien species (e.g. Colautti and McIsaac, 2004) but are nevertheless able to fruitfully interact.

Informal collaborations that remain within a scientific context may be called trading zones (Galison, 1999). Trading zones are research contexts that facilitate the development of integrative research approaches that transgress a single scientific (sub)discipline. In the case of trading zones there are typically two or more groups of experts that depend on specific inputs from each other. For instance the development of bioeconomic models (see above) may be interpreted as a trading zone, where economists and ecologists depend on each others’ expertise. Either ecologists develop models that they publish in ecological journals based on inputs from economists (e.g. Hastings et al., 2006), or vice versa (e.g. Olson, 2007). Thereby one expert group may try to direct research in the other field of expertise according to its needs for specific expert inputs (compare e.g. Bossenbroek et al., 2005). Trading zones are a common form of interdisciplinary collaboration, and exist for instance also at the interface between ecology and global change research (Kwa, 2005).

Formal processes of boundary management

Boundary organisations, i.e., agencies that are situated at the boundary between science and management, are appropriate institutional settings to facilitate formal processes of boundary management (Cash et al., 2003; Hellström and Jacob, 2003). The most prominent example in the case of invasive species management is the Global Invasive Species Programme (GISP, www.gisp.org). The GISP has the mission to synthesis problem-oriented research (Mooney et al., 2005), raise awareness about biotic invasions, document and disseminate information, and link relevant agencies. There also exist a number of geographically and thematically restricted boundary organisations (Clout and Poorter de, 2005; Meyerson and Mooney, 2007), for instance the Mountain Invasion Research Network (MIREN, www.miren.ethz.ch) that networks mountain ecosystem and invasive species managers and invasion biologists (Dietz et al., 2006).

However, to date these boundary organisations in the context of invasive species management mainly network agencies and disseminate information, but they do not invest in establishing structured boundary management processes. For instance, to our knowledge there are to date no transdisciplinary research projects on biotic invasions. In transdisciplinary research the framing of adequate research questions is itself a topic of the research project (Pohl and Hirsch Hadorn, 2007*). Participatory methodologies that guide deliberations among stakeholders have also rarely been employed in invasion research, with the notable exceptions of a few examples, such as the use of scenario planning (Binimelis et al., 2007*; Chapman et al., 2001; Higgins et al., 1997), multi-criteria assessment (Binimelis et al., 2007; Cook and Proctor, 2007), or agent based modelling of conflicts of interests (Macpherson et al., 2006).