Catalog Home Page

Landscape genetics and the effects of climate change on the population viability of declining avifauna in fragmented eucalypt woodlands of the West Australian wheatbelt

Angel, Antonia Sara (2015) Landscape genetics and the effects of climate change on the population viability of declining avifauna in fragmented eucalypt woodlands of the West Australian wheatbelt. PhD thesis, Murdoch University.

[img]
Preview
PDF - Whole Thesis
Download (3MB) | Preview

Abstract

The Rufous Treecreeper (Climacteris rufa), Yellow-plumed honeyeater (Lichenostomus ornatus) and the Western Yellow Robin (Eopsaltria griseogularis) are focal species and were investigated to assess the impacts of climate change and severe habitat fragmentation on the genetics and viability of remaining populations. This study was located within the west Australian wheatbelt where 93% of the native vegetation, including 97% of the York gum, wandoo and salmon gum woodlands have been cleared for agriculture (Saunders, et al., 1989) and where climate modelling predicts hotter and dryer weather conditions (CSIRO, 2005, IOCI, 2002). The Dryandra woodlands contains the largest native vegetation remnants in the central wheatbelt with a combined area of 28 066 ha and provides habitat for a diverse assemblage of flora and fauna many of which are in Decline, Threatened or Specially Protected (NWC, 1991).

The effects of habitat loss and fragmentation on the gene flow and population structure on the Rufous Treecreeper, was assessed within the Dryandra woodlands and across a range of fragmented habitat spanning approximately 100 km. Microsatellite and mitochondrial DNA data was applied to a spatial genetic and phylogeographic analysis. AMOVA shows genetic variation to be higher within populations (78%) than among populations (22%) and populations did not conform to Hardy Weinberg Equilibrium. This infers gene flow exceeds genetic drift across the region and the presence of migration between remnant habitats. Isolation by Distance was not found within Dryandra or across the region and infers the effective dispersal distance of the Rufous Treecreeper exceeds the geographical distance of sampling sites. However a Mantel’s Test found a correlation (r=0.316, p=0.004) with a distance of 28kms, within the Dryandra woodlands. A Spatial Autocorrelation of microsatellite DNA found a genetic structure of up to approximately 25kms (V=0.55) and beyond the Dryandra woodlands, shows genetic discontinuities where dispersal is more likely to occur. Landscape interpolation of genetic distance shows high genetic differentiation within the Dryandra woodlands and decreasing in an easterly direction where habitat size decreases and the distance between habitat increases. The Maximum Difference Delaunay Triangulation shows population boundaries of 12 populations within the woodlands including 3 central populations that are 1.3 km apart. A Bayesian Computation of microsatellites found a Continent-Island pattern of population structure across a distance of 85 km. Ritland’s Kinship Coefficient found dispersal patterns amongst populations within the Dryandra woodlands and a genetic neighbourhood size of about 1.7 km. Loiselle’s Kinship Coefficient found a unidirectional pattern of migration from the woodlands to smaller, isolated habitats with a maximum dispersal distance of 48 km. A Landscape Interpolation of male and female Rufous Treecreepers show a female bias in dispersal from Dryandra, with higher genetic divergence patterns in isolated remnants where habitat and nesting hollows are limiting.

Rufous Treecreeper mitochondrial DNA (partial cytochrome b gene) data was applied to the Mantel’s Test and found no correlation in Dryandra or the surrounding area but did show a positive correlation at a distance of 500kms and infers at least 2 different bioregions within this distance for this species. Results from the Interpolation and Principal Component Analysis show genetic variation decreasing with increasing distance from Dryandra in an easterly and southerly direction. The highest divergence patterns were found in Dryandra, North Yilliminning, Wickepin and Commondine Reserve. Genetic patterns with high similarity were found in Dongolocking and Highbury sites south- east of Dryandra and are most likely remnant populations that once belonged to a larger, continuous population or gene pool. A geographical distribution of shared mitochondrial haplotypes found a historical range prior to land clearing of approximately 85kms. A genealogy study based on coalescence found the earliest ancestral haplotypes belonged to Dryandra, North Yilliminning and Wickepin populations and should be prioritised for long term conservation purposes. Also, novel sequences of partial cytochrome b gene for the Yellow-plumed Honeyeater and Control Region for the Western Yellow Robin was resolved for further research.

The ecological niche and distribution of the Rufous Treecreeper was assessed using a distance based Redundancy Analysis (db-RDA) and a Habitat Suitability study. The db-RA found slope and aspect explained 29.16% (p= 0.04) of the genetic variation (phi) of mitochondrial DNA, which infers a relationship between landscape features and historical divergence patterns. Since old growth Eucalyptus wandoo trees are a critical habitat requirement for nesting hollows (Rose, 1993) a georeferenced (GIS) habitat suitability map was constructed from a vegetation survey (Coates, 1995) to show the distribution of E.wandoo and Rufous Treecreepers within Dryandra. Also using demographic information of the Rufous Treecreeper from a previous study (Luck, 2001) and RAMAS GIS (Akcakaya, 2002), it was estimated that the Dryandra contained enough suitable habitat for a maximum of 158 populations or 1 106 individuals.

The impact of climate change on the Dryandra woodlands and the Rufous Treecreeper was measured by annual rainfall measurements (BOM, 2011), satellite imagery of tree foliage cover of each sampling site and mist net capture recapture data. This study found a declining trend in rainfall patterns and in 2010, the annual rainfall (277.4mm) fell below the minimum climatic range (350mm) of E.wandoo forests. Based on climate modelling (CSIRO, 2005) the predicted reduction rainfall will eventually will negatively impact these forests by inducing a permanent state of drought. A critical threshold of 7.73% foliage cover was found, where foliage cover does not appear to recover foliage cover beyond 11.53% after a reduction to 7.73% in 2003. This indicates a critical threshold of percentage tree canopy cover for the E. wandoo in Dryandra. A linear regression found a significant relationship (p = 0.036) between previous year’s rainfall and percentage foliage cover. This delayed response to rainfall is explained by the defence mechanisms of E.wandoo that provide this species with drought tolerance (Veneklaas & Manning, 2007). A logistic regression (GLM) found foliage cover within the same year to be a significant predictor (p = 0.039) of Rufous Treecreeper captures. Therefore declining rainfall patterns and tree canopy cover have a direct impact on the abundance of Rufous Treecreepers.

The apparent survival rate estimate for the Rufous Treecreeper was 0.65 (SE 0.13) and 0.303 (SE 0.08) for the Yellow-plumed Honeyeater. Alternate modelling is required for the Yellow-plumed honeyeaters to account for their varied seasonal dispersal patterns and the Western Yellow Robin data could not be used for this demographic study because of small sample size. During 1997 and 1999 adult survival rates for Rufous Treecreepers within Dryandra was 0.76 (Luck (2001) and show the Rufous Treecreepers within the Dryandra woodlands are continuing to decline. A comparison of the two survival rates shows there is a reduction of 0.11 within an 8 year period (a single generation), which coincided with a 5.16% decrease in mean foliage cover during sampling times. This study concludes that climate change is negatively impacting E.wandoo forests and that tree foliage cover is not only a significant predictor in determining the presence of Rufous Treecreepers within the Dryandra woodlands, but also effects the short term survival and long term viability of this focal species.

Publication Type: Thesis (PhD)
Murdoch Affiliation: School of Veterinary and Life Sciences
Supervisor: Bradley, Stuart
URI: http://researchrepository.murdoch.edu.au/id/eprint/26420
Item Control Page Item Control Page

Downloads

Downloads per month over past year