An ant–plant byproduct mutualism is robust to sel

Authors: Tom M Fayle David P Edwards William A Foster Kalsum M Yusah Edgar C Turner

Publish Date: 2015/01/10

Volume: 178, Issue: 2, Pages: 441-450

PDF Link

Abstract

Anthropogenic disturbance and the spread of nonnative species disrupt natural communities but also create novel interactions between species Byproduct mutualisms in which benefits accrue as side effects of partner behaviour or morphology are often nonspecific and hence may persist in novel ecosystems We tested this hypothesis for a twoway byproduct mutualism between epiphytic ferns and their ant inhabitants in the Bornean rain forest in which ants gain housing in rootmasses while ferns gain protection from herbivores Specifically we assessed how the specificity overlap between fern and grounddwelling ants and the benefits of this interaction are altered by selective logging and conversion to an oil palm plantation habitat We found that despite the high turnover of ant species ant protection against herbivores persisted in modified habitats However in ferns growing in the oil palm plantation ant occupancy abundance and species richness declined potentially due to the harsher microclimate The specificity of the fern–ant interactions was also lower in the oil palm plantation habitat than in the forest habitats We found no correlations between colony size and fern size in modified habitats and hence no evidence for partner fidelity feedbacks in which ants are incentivised to protect fern hosts Per species nonnative ant species in the oil palm plantation habitat 18  of occurrences were as important as native ones in terms of fern protection and contributed to an increase in ant abundance and species richness with fern size We conclude that this byproduct mutualism persists in logged forest and oil palm plantation habitats with no detectable shift in partner benefits Such persistence of generalist interactions in novel ecosystems may be important for driving ecosystem functioningWhen humans severely alter the environment ecologically stable novel ecosystems can be created Hobbs et al 2009 These ecosystems are characterised by the extinction of some species Tilman et al 2001 invasion by others Mack et al 2000 and shifts in the abiotic environment Ewers and BanksLeite 2013 As a consequence the network of interactions between species can be significantly altered Tylianakis et al 2010 with the loss of individual species leading either to coextinctions of species reliant on each other Koh et al 2004 or increases in abundance of species whose competitors or natural enemies disappear Ritchie and Johnson 2009 These processes are expected to change the adaptive landscape in newly formed ecosystems with shifts in the costs and benefits of associations between specific partner species Aslan et al 2013Mutualistic interactions between species are widespread in both natural and novel ecosystems Some mutualisms are highly specific with each partner having evolved to invest reciprocally in the other’s fitness Foster and Wenseleers 2006 However it is now clear that generalist mutualisms including those defined as “byproduct mutualisms” Leimar and Connor 2003 are also widespread Bronstein et al 2006 A byproduct mutualism occurs when each of the species involved acts to optimise its own fitness and the byproduct of doing this is an increase in the fitness of the partner Critically there is no evolved investment in partner fitness and these relationships tend to remain diffuse and nonspecific Leimar and Connor 2003 Byproduct mutualisms are expected to persist in novel ecosystems in which many native species are no longer present because the species involved are generalists and not reliant on a single partner Hence the interaction could persist under two scenarios 1 a subset of the remaining partners continue to interact or 2 invading nonnatives fill the roles vacated by locally extinct partnersGlobally many species with no recent shared evolutionary history are being brought together by the creation of novel ecosystems through anthropogenic habitat degradation and the transport of species around the world Across the tropics a major driver of such changes is selective logging of forests Edwards et al 2014b and subsequent conversion to agriculture Tilman et al 2001 Selectively logged forest often retains similar numbers of species to primary forest across a range of taxa Gibson et al 2011 but has altered species composition Edwards et al 2014a Conversion to agriculture typically has more extreme impacts Edwards et al 2014a Gibson et al 2011 with loss of forest species and invasion by nonnatives Mack et al 2000—and hence increased likelihood of the formation of novel species interactions Although these nonnative species are often considered in a negative context they have the potential to be functionally important in novel ecosystems Davis et al 2011Mutualistic interactions between ants and plants are widespread with plants trading housing or food for antprovided protection from herbivores seed dispersal or food collection Bronstein et al 2006 Mutualistic ant–plant networks are known to be fairly robust to habitat degradation Bruna et al 2005 Passmore et al 2012 but the effects of degradation on the benefits for partners at least for mutualisms where plants provide housing are less wellknown Wetterer 1997 In the study reported here we focussed on an ant–epiphyte interaction that persists across a disturbance gradient from primary forest to selectively logged forest and oil palm agriculture Fayle et al 2010 Turner and Foster 2009 in Malaysian Borneo Southeast Asia This region is a biodiversity hotspot Myers 2000 which is suffering from an accelerating rate of deforestation Hansen et al 2013 driven mainly by expansion of oil palm plantations Koh and Wilcove 2008 Bird’s nest ferns Asplenium spp are littertrapping epiphytes that form a twoway byproduct mutualism in primary forest with their ant inhabitants with ferns providing multiple colonies of ants with a nest site in their root mass and the ants protecting ferns from herbivores Fayle et al 2012 However since the root mass is not enclosed or partitioned being adapted for the capture and breakdown of leaf litter ferns are unable to direct housing benefits to more beneficial ant partners Larger ferns thus support more ant colonies of different species rather than larger colonies of cooperating species which in turn means that ant species that are more beneficial are unlikely to benefit greatly from helping ferns to grow ie they are less likely to receive partner fidelity feedbacks Fayle et al 2012 This is in contrast to plants that house ant partners in adapted enclosed structures termed domatia and are consequently able to direct benefits by rewarding or punishing specific ant colonies depending on their behaviour Edwards et al 2006Logging and conversion to oil palm plantation might affect the bird’s nest fern–ant interaction in a number of ways Increasing homogenisation of ant faunas with disturbance in particular compression of communities in response to vertical collapse of microclimate regimes Fayle et al 2009 Scheffers et al 2013 is expected to reduce the specificity of the interaction but see Beaulieu et al 2010 Since bird’s nest ferns support substantial amounts of leaf litter we would expect this collapse to result in increasing similarity between ferndwelling ant communities and ant communities inhabiting leaf litter on the forest floor Predation rates by ants are known to increase with disturbance Tvardikova and Novotny 2012 and nonnative species which commonly invade disturbed areas are often better protectors of plants than native species Ness and Bronstein 2004 The relationship between fern size and the number and size of resident colonies has implications for the benefits to ants to invest in fern growth and hence on the potential for partner fidelity feedbacks Weyl et al 2010 There are two possible disturbancelinked changes to these partner fidelity feedbacks either resident ants expand colony sizes as fern size increases and hence reap benefits from investment in fern fitness or extra living space is coopted by new ant colonies and hence residents would not benefit from investment as is the case in primary forest see above Nonnative ant species which can play key roles in mutualisms Ness and Bronstein 2004 occur in bird’s nest ferns in oil palm plantations Fayle et al 2010 but their role in this byproduct mutualism is not known A species pool of functionally equivalent ferndwelling ant species both natives and nonnatives including those robust to habitat disturbance would allow persistence of the byproduct mutualism in this altered habitat

Keywords: