Resource partitioning for phosphorus (P) in a P-limited plant community: preference for different soil P sources among co-occurring species

The mechanisms that allow many species to co-exist have long interested scientists. Resource partitioning is one such mechanism that allows species to share resources and so reduces competition between them and hence promotes co-existence. In plants, this may occur where species share a nutrient that is in limited supply by having preference for different forms of that nutrient. However, despite phosphorus (P) being a limiting nutrient in approximately half the world's plant communities, interspecific differences in preference for different P forms has never been demonstrated and so the existence of P partitioning remains unknown. Moreover, the potential role of P partitioning in driving plant community structure has bever been investigated. This is a major oversight, given (i) the prevalence of P limitation in plant communities, (ii) that P limited systems often have high floristic diversity but it is not known whether P partitioning contributes to that diversity, and (iii) that P partitioning is likely to occur given the high diversity of different forms of P in soil and the many adaptations that plants have to access soil P. Proving the existence of P partitioning and establishing its mechanistic basis would provide a fundamental advance in our understanding of the importance of resource partitioning by demonstrating its operation through one of the most commonly limiting nutrients. This will be achieved using a P limited grassland as a model system. Different 33P radioisotope labelled P forms representative of some major soil P pools will be supplied to microcosms of co-occurring plant species to determine which species show preference for which P forms. Since P partitioning is proposed to operate through interspecific differences in plants' abilities to access pools of different levels of bioavailability, we will use P sources ranging from highly bioavailable to refractory forms. We will therefore supply the P sources of (i) orthophosphate - which is directly taken up by plant roots but occurs in very small quantities in soil, (ii) DNA, relatively less bioavailable than phosphate but more bioavailable than (iii) inositol phosphate, an organic source of high abundance but low bioavailability. Importantly, uptake of P from different forms will be combined with analyses of concentrations of those forms naturally occurring in soil, so that total uptake from the naturally occurring and labelled pools can be quantified and allowing true preference for different forms to be properly determined. The plants for which uptake from these P sources will be quantified include (i) a sedge with specialist root adaptations that may enhance access to refractory P, (ii) a grass with mycorrhizal symbioses and root secretions that that may enhance access to some refractory P, (iii) a non-legume forb with mycorrhizal symbioses but coarse root systems that may show less ability to access refractory P and some preference for more bioavailable forms, (iv) a legume, that may show preference for very bioavailable P given their high P demand, and (v) a forb that is non-mycorrhizal and so may show the greatest preference for very bioavailable P. Overall, we therefore predict these species to show preference for different P forms and therefore provide the first evidence for resource partitioning of P. This research will provide a springboard for more detailed work on the importance of P partitioning (detailed in 'academic beneficiaries'). However, we believe such more extensive (and expensive) work first needs proof-of-concept for the preferential use of different P forms by co-occurring species, hence our focussed small grant application.