Geographically contrasting biodiversity reductions in a widespread New Zealand seabird.

Unravelling prehistoric anthropogenic impacts on biodiversity represents a key challenge for biologists and archaeologists. New Zealand's endemic Stewart Island Shag (Leucocarbo chalconotus) comprises two distinct phylogeographic lineages, currently restricted to the country's south and southeast. However, fossil and archaeological remains suggest a far more widespread distribution at the time of Polynesian settlement ca. 1280 AD, encompassing much of coastal South Island. We used modern and ancient DNA, radiocarbon dating, and Bayesian modelling, to assess the impacts of human arrival on this taxon. Our analyses show that the southeast South Island (Otago) lineage was formerly widespread across coastal South Island, but experienced dramatic population extinctions, range retraction and lineage loss soon after human arrival. By comparison, the southernmost (Foveaux Strait) lineage has experienced a relatively stable demographic and biogeographic history since human arrival, retaining much of its mitochondrial diversity. Archaeological data suggest that these contrasting demographic histories (retraction vs. stability) reflect differential human impacts in mainland South Island vs. Foveaux Strait, highlighting the importance of testing for temporal and spatial variation in human-driven faunal declines.

Demographic rates of northern royal albatross at Taiaroa Head, New Zealand.

Demographic rates, such as annual survival rate, are generally difficult to estimate for long-lived seabirds, because of the length of time required for this kind of study and the remoteness of colonies. However, a small colony of northern royal albatross (Diomedea sanfordi) established itself on the mainland of New Zealand at Taiaroa Head, making possible regular banding and monitoring of its individuals since the first chick fledged, in 1938. Data on the presence/absence of birds, as well as on breeding outcomes, were available for the period from 1989-90 to 2011-12, and included 2128 annual resightings of 355 banded individuals of known age. The main goal of the present study was to estimate the annual survival rate of juveniles, pre-breeders, and adults at Taiaroa Head. These rates were estimated simultaneously in a single Bayesian multi-state capture-recapture model. Several models were fitted to the data, with different levels of complexity. From the most parsimonious model, the overall annual adult survival rate was estimated as 0.950 (95% CI [0.941-0.959]). In this model, adult survival declined with age, from 0.976 (95% CI [0.963-0.988]) at 6 years, the minimum age at first breeding, to 0.915 (95% CI [0.879-0.946]) at 40 years. Mean annual survival of pre-breeders was 0.966 (95% CI [0.950-0.980]), and 0.933 (95% CI [0.908-0.966]) for juveniles. There was no discernible difference in survival between males and females, and there was no apparent trend in survival over time. Estimates of other demographic rates were also obtained during the estimation process. The mean age at first return of juveniles to the colony was estimated as 4.8 years (95% CI [4.6-5.1]), and the mean age at first breeding as 8.9 years (95% CI [8.5-9.3]). Because all the birds of the colony were banded, it was possible to estimate the total population size. The number of northern royal albatross present annually at the Taiaroa Head colony has doubled since 1989-90, and the current total population size was estimated to be over 200 individuals. The ratio of the total population size to the number of annual breeding pairs varied from 5 to 12 among years, with an overall mean of 7.65 (95% CI [7.56-7.78]), and this high variability highlights the need for a sufficient number of surveys of seabird breeding populations before reliable conclusions on population trends can be made. Although long-term data allowed estimates of demographic rates of northern royal albatross at Taiaroa Head, the location of the colony and the ongoing management by staff mean that the population dynamics may differ from those of the main population on the Chatham Islands.

Strong phylogeographic structure in a sedentary seabird, the Stewart Island Shag (Leucocarbo chalconotus).

New Zealand's endemic Stewart Island Shag (Leucocarbo chalconotus) comprises two regional groups (Otago and Foveaux Strait) that show consistent differentiation in relative frequencies of pied versus dark-bronze morphotypes, the extent of facial carunculation, body size and breeding time. We used modern and ancient DNA (mitochondrial DNA control region one), and morphometric approaches to investigate the phylogeography and taxonomy of L. chalconotus and its closely related sister species, the endemic Chatham Island Shag (L. onslowi). Our analysis shows Leucocarbo shags in southern New Zealand comprise two well-supported clades, each containing both pied and dark-bronze morphs. However, the combined monophyly of these populations is not supported, with the L. chalconotus Otago lineage sister to L. onslowi. Morphometric analysis indicates that Leucocarbo shags from Otago are larger on average than those from Foveaux Strait. Principal co-ordinate analysis of morphometric data showed substantial morphological differentiation between the Otago and Foveaux Strait clades, and L. onslowi. The phylogeographic partitioning detected within L. chalconotus is marked, and such strong structure is rare for phalacrocoracid species. Our phylogenetic results, together with consistent differences in relative proportions of plumage morphs and facial carunculation, and concordant differentiation in body size and breeding time, suggest several alternative evolutionary hypotheses that require further investigation to determine the level of taxonomic distinctiveness that best represents the L. chalconotus Otago and Foveaux Strait clades.