Design a site like this with
Get started

A Global Test of the Passenger vs Driver Models of Invasion

The proliferation of alien invasive species poses a significant threat to biodiversity, ecosystem stability, and ecosystem services (Charles & Dukes 2007).  The interactions that lead to invasive species success are widely considered context-dependent.  While general rules link invasive species together in theory (rapid growth, lack of natural enemies, etc…), invasive species come from phylogenetically disparate sources, exhibit widely variable morphologies, and appear to thrive under a multitude of local contexts.  Furthermore, the term “invasive” implies an imprecise dichotomy where non-native species are either destructive or beneficial. 

Key Measurements

There is a problem with the passenger-driver dichotomy; both processes lead to visually similar outcomes.  An invader becomes numerically dominant and uses up more of the shared resources on the landscape regardless of whether they are true ‘passengers’ or ‘drivers’.  MacDougall and Turkington recognized this, and focused their efforts on measures that would differentiate between the two mechanisms in systems where invasive species are experimentally removed : niche overlap, dispersal limitation, and functional similarity.

Niche Overlap

‘Drivers’ are more likely to take resources that could be used by natives, rather than resources that would go unused. Productivity is therefore a zero-sum game, with increasing productivity of the invader offset by productivity declines among the native species they replace.

Dispersal Limitation

‘Passengers’ are more likely to dominate in systems where native species do not produce sufficient propagules to colonize empty resource patches efficiently.

Functional Similarity

‘Drivers’ are more likely to be functionally similar to the native species they replace.

MacDougall & Turkington (2005) proposed a theoretical contrast that is generally useful for articulating these varied responses—the driver vs passenger models of invasive species success.  The driver model suggests that invasive species, once introduced, competitively exclude native species, inhibit resource acquisition by natives and force potentially irreversible shifts in ecosystem dynamics.  Invasive species are the causes rather than symptoms of ecosystem degradation.  Indeed, the term ‘invasive’ conjures up just such an interpretation.  The passenger model suggests that invasive species are simply adept at colonizing niches in native communities that were previously made available by some progressive (or abrupt) environmental change.  In this scenario, invasive species are merely symptoms, rather than causes of ecosystem degradation.  MacDougall & Turkington (2005) tested the relative importance of these models by experimentally manipulating two invasive perennial species in a fire-suppressed oak-savanna.  They produced evidence (through a comparison of relative effects on native species via competition or dispersal limitation) that species were primarily passengers rather than drivers of community change and biodiversity loss, breaking with the traditional view that invasives are primarily causes of ecological degradation.  This evidence has often been cited (especially in reference to the passenger/driver dichotomy), but there has not been a comprehensive and systematic evaluation of the generality of either model beyond the two species tested in this oak-savanna in 2005.

In a Nutshell


3-years of data collection

Every Academic Year

Fall – Plant Survey

Spring – Invasive Species Removal

Year 2: Seed Addition Study

Total estimated material costs at startup : ~$200

Collaboration is also encouraged for PIs who do not regularly teach an instructional laboratory course suitable for this experiment.

Benefits to contributing PIs and institutions:

The proposed CDE will be conducted by PIs with an optional 2-week laboratory module suitable for courses in general ecology or environmental science (participants that use the instructional lab-module will be included as co-authors on a manuscript addressing the value/feasibility of CDEs for undergraduate instruction.  Experimental startup will involve minimal costs beyond person-hours and reliable access to natural areas (~$200, see materials list provided with protocol).  Laboratories that contribute the minimum data requirements for participation will be awarded co-authorship for both the PI and one student collaborator.  Additional student co-authors may be added for replicating the experimental protocol at multiple sites.  The 2-week laboratory module will ensure similar timing of manipulations/observations, and observations will be sufficiently documented and non-technical so as to be reliable given the incorporation of novice observers.