Divaricate plant species account for more than 10% of New Zealand/Aotearoa's woody flora, a higher proportion for that life form than anywhere else on earth. Two main hypotheses have been proposed to account for the prevalence of this phenotype. The first suggests herbivorous birds, particularly large flightless moa, exerted selective pressure on many plants to adopt the form. The second proposes climate as the main driver since insular New Zealand has many exposed habitats and, historically, experienced stressing conditions during the Pleistocene. Our study investigated two questions to shed further light on the evolution of divaricates, in particular related to the potential influence of browsing by large avian herbivores. First, divaricate plants have been posited to have a higher tensile strength than non-divaricates as a defence mechanism against moa browsing. We tested the great majority of New Zealand plant genera within which divaricates occur, contrasting the stem tensile strength of these plants against their closest non-divaricate counterparts using accurate testing technology. The results indicate divaricate species have tensile strength that approaches twice that of non-divaricates across a broad and disparate phylogenetic range that includes gymnosperms and many clades of angiosperms. Second, we tested the species level distributions of widely dispersed woody plant genera across the New Zealand archipelago on islands where moa had been present or always absent. We found that no endemic divaricate species occurred on any islands from which moa had been permanently absent. In contrast, their non-divaricate counterparts were commonly endemic to those same islands.

Predicting species’ potential distributions and niches requires multi-scale data encompassing the past and present. Increasingly, researchers have advocated using historical contexts to inform ecological niche models (ENMs). Two key sources of past distributions are fossils and historical records. Fossils are subject to sampling and taphonomy biases but can offer insights into the temporal dynamics over millennia. Historical records are filtered by human perceptions and biases and have a shorter temporal range but compared to fossils provide different contextual information from a broader range of habitats. New Zealand has a relatively short history of human occupation and rich fossil archives and historical literature. Approximately 25% of the world’s seabirds, nearly half of which are endemic, breed in New Zealand. Since human arrival in New Zealand, many seabird populations have declined in numbers and breeding ranges, primarily due to introduced mammalian predators. Here, we explored the ecological niche space for breeding colonies of three size groups of burrowing procellariiforms using four admixtures of locational records (fossil bones, fossil bones + historical observations, historical observations, and post-1990 observational records). We fitted ENMs using the maximum entropy algorithm and calculated niche metrics. For all groups, the breeding niche space captured separately by the fossils and historical data had low overlap with each other and reflected different environmental aspects. The combined fossil + historic datasets predicted a niche that overlapped the post-1990 observed niche. Moreover, the combination of the fossil and historic datasets demonstrated that breeding grounds, now restricted mainly to predator-free settings, were once more widespread across New Zealand. We show that historical and fossil datasets complement each other mitigating biases unique to either dataset. Together, such records can provide critical insights into the historical drivers of species range contractions, contextualising current ecosystems.