Conservation and restoration genomics of Syzygium maire, a nationally critically threatened tree species of Aotearoa New Zealand

Genetic degradation in forest trees driven by habitat shifts and fragmentation is likely to lead to substantial maladaptation and species loss. Endemic island species are uniquely threatened, due to lack of buffer populations outside of affected ranges, and already low or localised adaptive diversity confined to a limited number of specialised habitats. In Aotearoa New Zealand, approximately 80% of vascular plants are endemic, while 70% of land cover is affected by human development. Swamplands, considered a taonga (treasure) to Māori, the Indigenous peoples of Aotearoa, are some of the most severely threatened, with loss upwards of 98% in some regions. Their restoration is currently a priority, but restoration decisions are being made on the basis of generalised recommendations, with most studies in Aotearoa focussed on commercially important plants or charismatic animal species.

Syzygium maire is an endemic, critically threatened, canopy tree species of swamp forests in Aotearoa. Remnant stands are often small, isolated and surrounded by pasture, restricting movement of its primary insect and bird dispersers. The species is able to self-fertilise and seeds do not persist in the seedbank. Additionally, S. maire is highly susceptible to myrtle rust (Austropuccinia psidii), which threatens reproductive viability through destruction of flowers, fruit and seedlings. These factors place S. maire in significant danger of further population decline and genetic degradation. While many swampland restoration programmes include S. maire on their species lists, little is known about the genetic structure and adaptive potential of the species. These projects may therefore be compounding threats through further restricting diversity and perpetuating fragmentation.

My thesis aims to build genomic resources for S. maire and describe extant patterns of genetic diversity at both range and local scales to inform conservation and restoration of the species. Here I 1) construct a reference genome for S. maire and identify genes involved in pathogen defence, 2) assess the species' population genetic structure and diversity across its range, and 3) evaluate the impact of habitat fragmentation on genetic diversity of the species in remnant stands.

Using long-read, high-accuracy sequencing, I produce a reference-quality genome for S. maire. The individual tree and its genome have been named Ngā Hua o te Ia Whenua by Rangitāne o Manawatū, the local Māori tribe acting as kaitiaki (guardians) over it. I detect large scale synteny between three Syzygium species and Eucalyptus grandis, providing support for the stability of Syzygium and Myrtaceae genomes. I annotate several hundred nucleotide-binding oligomerization domain-like receptor, leucine-rich repeat (NLR) genes implicated in pathogen defence, identifying putative homologs of myrtle rust resistance genes. This resource will support conservation efforts by facilitating studies aimed at understanding the distribution of genetic diversity, adaptive traits, and pathogen resistance mechanisms.

Population genetic analyses of 269 wild trees based on whole-genome resequencing revealed five distinct regional clusters with local structure and admixture. Diversity patterns suggest more recent radiations in the south and stable populations in the north. Similar diversity throughout the range, however, indicates against north-south radiation since the last glacial maximum. I suggest that these patterns are a result of glaciation related ecological inversion leading to cycles of refuge-radiation across the species range. Finally, I discover adaptive variation associated with three climactic and two soil variables. I mirror the recommendations of previous work on native trees in that seed sourcing should be broader than currently practised to maximise diversity. However, I describe an additional eastern cluster and caution that more local scale evolutionary signatures do exist. These findings will contribute to national level prioritisation of S. maire populations and genetic material for conservation and restoration.

Finally, I assess genetic effects of natural and artificial habitat fragmentation in the Greater Wellington Region. I identify evidence of pre-fragmentation genetic homogeneity in extensive swampland with structure relating to isolation by distance. Spatial analyses and close kin relationships in habitat patches, however, reveal potential for genetic isolation among trees separated by just a few kilometres. Concurrently, small but significant differences in genetic diversity between adults and seedlings indicate that fragmentation is already degrading the species' genetic integrity. Restoration should aim to improve population connectivity and conserve as many genotypes as possible within regional clusters to protect adaptive potential.

This thesis presents the first genomics analysis of an exclusively animal dispersed tree species in Aotearoa New Zealand. The reference quality genome sequence creates opportunity for future research, including further exploration of putative myrtle rust resistance genes which may inform breeding for resistance. Discovery of broad genetic clusters and centres of diversity inform conservation management and restoration decisions for the species, and contribute to the understanding of the unique island biogeography of Aotearoa. Finally, identification of effects of fragmentation and fine scale spatial genetic structure on remnant stands of trees provides practical and direct goals for species restoration at the local scale. The findings of all chapters will contribute to the restoration and conservation of S. maire and swamp forest ecosystems in the country.