BREEDING SYSTEMS, POLLEN FLOW, AND CONTINUING EVOLUTION IN AVON GORGE SORBUS (WHITEBEAMS, ROWANS, AND SERVICE TREES)

The Avon Gorge is one of the richest botanical sites in Britain, and is internationally important for its Sorbus diversity, with 3 endemic species and at least 3 endemic hybrids amongst the 19 taxa recorded (Rich et al. 2008). Four species are currently on the IUCN Red List of endangered species: S. bristoliensis, S. wilmottiana, S. anglica and S. eminens (IUCN, 2007). Recent research by Hiscock and CASE partner, Rich, has shown that Sorbus diversity in the gorge is generated primarily by hybridisation between sexual diploids S. aucuparia, S. torminalis, and (especially) S. aria, and the polyploid facultative apomict S. porrigentiformis and its derivatives, S. eminens and S. bristoliensis (Robertson et al. in preparation). The type of mating system employed by the various taxa is pivotal to this continuing evolutionary process. The 3 sexual taxa are obligate outcrossers whereas the hybrid polyploids are either obligate apomicts (producing asexual seed), or facultative apomicts (reproducing by a combination of apomixis and sexual reproduction). In the polyploids, apomixis is via pseudogamy with pollination being required to induce asexual seed development and produce a sexual endosperm. Importantly, however, our recent studies have shown that the rare apomictic polyploids, S. bristoliensis and S. wilmottiana, cannot utilize self-pollen or pollen from other con-specific individuals (clones) because of self-incompatibility. This necessitates that they receive pollen from other Sorbus species, in particular S. aria, in order to produce seeds via pseudogamy (Robertson et al. in preparation). Thus pollination is essential for all reproductive processes among the Sorbus taxa and consequently for their population dynamics and ongoing evolution. This research will aim to quantify pollen movement and pollen limitation among Sorbus taxa in the Avon Gorge and relate this to population dynamics and ongoing evolution of new Sorbus taxa. These data will then be used to inform our collaborators, Bristol City Council, Natural England, and The National Trust, on the most appropriate conservation strategies to preserve rare endemic Sorbus species and to maintain the process responsible for creating Sorbus diversity within the Avon Gorge. Our preliminary findings (Robertson et al. in preparation) indicate that S. aria is an essential source of pollen needed for seed production by the rare apomictic Sorbus species suggesting that S. aria will need to be considered in all Sorbus conservation strategies. The research proposed here has two main aims: 1) to characterize the mating system(s) of the 5 new Sorbus taxa and 2) to quantify flowering time, pollen movement and pollen limitation among the different Sorbus taxa in the Avon Gorge and relate this to reproductive output, population dynamics, and ongoing evolution of new Sorbus taxa. In particular, this work will confirm whether S.aria pollen is indeed driving reproduction and ongoing diversification of Sorbus in the Avon Gorge, and provide estimates of the number and density of S. aria needed to maintain these processes and guaratee successful reproduction of the rare endemic species. This will be accomplished through delivery of the following objectives: 1. Establish modes of reproduction (mating system) in newly described Sorbus taxa not already tested. 2. Determine annual and seasonal variation in flowering time for all Sorbus taxa. 3. Determine the primary pollinators of Sorbus pollinators. 4. Measure pollen limitation among sexual and apomictic Sorbus taxa. 5. Estimate distances of pollen flow from individual Sorbus trees (particularly S. aria) using fluorescent dust and polymorphic microsatellite markers. 6. Calculate pollen neighbourhoods and predict densities of S. aria required to effect seed production in rare pseudogamous apomicts.