Deciphering complex relationships between apparently unrelated species



Franck Courchamp, Ecologie, Systématique & Evolution, UMR CNRS 8079, Univ. Paris Sud, Orsay 91405, France


Rabbits are considered by biologists to be among the worst invasive species. Their impact on invaded ecosystems, epitomized in Australia and on islands worldwide (e.g. North, Bullock & Dulloo, 1994; Williams et al., 1995; Burbidge & Moris, 2001; Garzon-Machado et al., 2010), has earned them a well-deserved place in the infamous Invasive Species Specialist Group's ‘100 of the Worst Invasive Alien Species’ list. Rabbits are capable of profoundly altering communities, ecosystems and landscapes by sapping biodiversity and biomass from the basic trophic levels of often fragile ecosystems. However, they also have more complex, and hitherto seldom documented effects on higher trophic levels (Zedler & Black, 1992; Priddel, Carlile & Wheeler, 2000; Oliver, Luque-Larena & Lambin, 2009; Denyer, Hartley & John, 2010).

In the context of a few, but loud voices attempting to question the importance of alien invasive species (but see Lambertini et al., 2011; Lockwood, Hoopes & Marchetti, 2011), the paper by Brodier et al. (2011) is well timed. It reminds us that restoration ecology and the fight against biodiversity loss are possible, in particular on islands, as they generally have many endemic and threatened species. This paper also highlights two very important, yet previously overlooked points. The first is the importance of long-term follow-up after restoration programmes. The second is the need to consider the indirect (and consequently often hidden) effects that alien invasive species can have on seemingly unconnected species.

By monitoring the populations, several petrel species and their avian predator (the brown skua, Catharacta skua) following the removal of rabbits, Brodier et al. (2011) show that rabbits compete for space with blue petrels, Halobaena caerulea. Following rabbit eradication, the competition for burrows was removed and the blue petrel population increased up to eightfold in a few generations. Brown skua also increased their survivorship in response to the increase in its main prey, the blue petrel. Interestingly, the population of another petrel, the Antarctic prion, Pachyptila desolata, was also impacted by the rabbits. Following the eradication, the Antarctic prion population declined fourfold (to exclusion) in areas of deep soil and remained stable where the soil was too shallow for blue petrel (and rabbits). The resulting system, illustrated in Fig. 1, shows that one unexpected, and up to now unreported, effect of rabbits was to erode the niche partitioning of nesting seabirds on the island. Removing the rabbits led to both a competitor release (of the blue petrel, freed from the competition with rabbits) and to an apparent competitor release (of the shared predator, the brown skua, boosted by increased blue petrel availability). The combination of direct and indirect interactions restricted Antarctic prions to shallow soil areas, less favoured by the blue petrel.

Figure 1.

Schematization of the main relationships among the major vertebrate components of an insular ecosystem with (a) and without (b) introduced rabbits. Blue arrows are for competition, red arrows for predation and purple arrows for apparent competition. The size of the arrow is an indication of the strength of the interaction.

This neat system shows how alien invasive species can have effects beyond the trophic level constituting their resources. The contrast between the spectacular recovery of the blue petrel, the increased survival but stable number of skua breeding pairs and the decrease of the Antarctic prion also illustrates well the difficulties to forecast outcomes that conservation program may face even in relatively simple trophic webs. This is especially so in the current context of global climate change, which already affects subantarctic islands (Chapuis, Frenot & Lebouvier, 2004; Dowding et al., 2009), and which could have played a role in the system, for example, in the delayed recovery of the skua.

Monitoring population densities, survival and predator diet, every year, on this remote, far reaching island, for long after the rabbit eradication might have been seen as a risky bet. As it turned out, the bet paid off, in the best currency ever: knowledge and understanding. Deciphering the indirect effects a terrestrial herbivore can have on a community of seabirds with which they have few interactions if any, is crucial in our battle against alien invasive species (Bergstrom et al., 2009; Genovesi, 2011). It demonstrates very well that if we were to ‘judge alien species only on their effects’, we would miss all but the tip of the iceberg and would allow important but unapparent, indirect effects of alien invasive species on native ecosystems.

One regret from this study is that the monitoring effort conducted after rabbit eradication was not also implemented for a few years before the eradication. This would have allowed a robust comparison of the situation prior to and following rabbit eradication, giving important information on the complexity of their effect on the invaded community. For example, knowing quantitatively the composition of skua diet would likely have strengthen the demonstration that rabbits did not boost the skua population by providing a surplus of prey (e.g. hyperpredation; Roemer, Donlan & Courchamp, 2002). Such pre-eradication studies are key to success, especially as most islands are generally large, more complex and multi-invaded, with a guild of introduced competitors, mesopredators and top predators. Assessing the precise place of the alien species in the invaded trophic web, although long advocated (Zavaleta, Hobbs & Mooney, 2001; Courchamp, Chapuis & Pascal, 2003), is unfortunately seldom a luxury that funding agencies will grant, or that invaded ecosystem can afford (Simberloff, 2003). Yet, it is an approach that may help implement the most efficient restoration programs, prevent surprise effects (Caut, Angulo & Courchamp, 2009; Bonnaud et al., 2010) and generate further understanding on trophic web dynamics, and beyond that on ecosystem functioning.