Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Australie → The conservation and ecology of a rare and declining agamid lizard, the grassland earless dragon Tympanocryptis pinguicolla in the Monaro region of New South Wales

University of Canberra (2015)

The conservation and ecology of a rare and declining agamid lizard, the grassland earless dragon Tympanocryptis pinguicolla in the Monaro region of New South Wales

McGrath, Tim

Titre : The conservation and ecology of a rare and declining agamid lizard, the grassland earless dragon Tympanocryptis pinguicolla in the Monaro region of New South Wales

Auteur : McGrath, Tim

Université de soutenance : University of Canberra

Grade : Masters in Applied Science by Research 2015

Résumé partiel
The conservation of the world’s biodiversity is a challenge of paramount importance as the world changes rapidly and global land use intensifies. Today, nearly 20% of reptile species across the globe are threatened with extinction yet knowledge of diversity in this group remains poor. The problem is most critical in the world’s expanding agricultural environments where there have been extensive declines and localised extinctions of reptiles. The Monaro region of south eastern NSW in Australia is a prominent agricultural region where livestock grazing has occurred for some 180 years, largely replacing natural grassland modification processes that included wildfire, kangaroo grazing and burning by indigenous people. Much of the vertebrate fauna that once occupied this region (e.g. emus, bandicoots, rat kangaroos, wallabies and bustards) has disappeared since European settlement. Yet reptiles remain a significant component of the vertebrate fauna in this region. The grassland earless dragon Tympanocryptis pinguicolla is one such species. This cryptic lizard is one of Australia’s rarest and most endangered reptiles having declined and contracted in range dramatically. The rediscovery of this species in the Monaro region in 1993 was followed by limited surveys that suggested a relatively broad but undefined distribution within the region. Despite this, little was known about the species habitat preferences, making difficult informed decisions about its conservation management. This was concerning given the region is subject to agricultural intensification, changes to vegetation clearing laws, increasing renewable energy developments such as wind turbines and the potential sale of Travelling Stock Reserves. In this thesis, I report on three key aspects of the ecology of T.pinguicolla : (1) the effectiveness of detection techniques, (2) the extent of its distribution and habitat requirements, and (3) how land management practices in the region may affect the lizard. To address these aims, I employed a targeted, landscape wide, probabilistic sampling program which accounted for T. pinguicolla rarity through statistical inference. Specifically, I conducted repeated presence/absence winter rock turning surveys at 67 sites over four years and used a robust binomial mixture model in WINBUGS and logistic regression modelling using R to estimate occupancy and detectability and the landscape and microhabitat variables that drive its presence. I determined that T. pinguicolla has a very low detectability (0.0098) and an occupancy estimate of 0.46 suggesting the species was likely to have been missed at some sites during survey. It is therefore likely that the lizards are more widespread across the Monaro region than previously thought. I also found that the lizard’s occupancy is likely to be greatest at higher elevations in native grasslands not dominated by Themeda triandra on private lands, on both basalt and sedimentary geology between the Murrumbidgee and MacLauchlan river systems. Occupancy was not influenced by Travelling Stock Reserves or Nature Reserve indicating private grazing lands make up most of the species area of occupancy. By considering survey effort and the level of confidence in the estimate of detectability I assessed detection at 60 sites. This analysis revealed that about 2400 rocks need to be turned at an occupied site to have a 90% chance of detecting T. pinguicolla. The number of rocks increased to 5400 to raise confidence to 99%. My survey effort was sufficient to infer absence with high confidence at some sites highlighting serious concern for the species at numerous sites around Cooma in particular the Kuma Nature Reserve ; the only grassland reserve for this lizard in the Monaro region. Overall, my analysis confirms the low effectiveness and likely destructive nature of rock turning as a survey technique for this lizard and other threatened rupicoline reptiles, suggesting the need for alternative survey approaches. I also compared remote activated cameras at one site in the Monaro region with rock turning. The cameras provided clear and easily identifiable photos and video recordings of the lizard while rock turning failed to detect the species. This preliminary trial indicates that camera traps are likely to offer an effective less invasive and less labour intensive alternative approach to the detection and monitoring of T. pinguicolla. Despite the species extremely low level of detection, my surveys resulted in the discovery of new locations for T. pinguicolla on private properties and Travelling Stock Reserves, extending our understanding of the species range and habitats. Analysis

Mots clés : reptile species ; cryptic lizard ; binomial mixture model ; logistic regression model using R ; Murrumbidgee river ; MacLauchlan river ; Travelling Stock Reserves ; Kuma Nature Reserve

Présentation

Version intégrale (4,86 Mb)

Page publiée le 2 août 2017