How high is your hummock? The importance of Triodia height as a habitat predictor for an endangered marsupial in a fire‐prone environment. (22nd January 2016)

• Record Type: Journal Article
• Authors: Moseby, Katherine; Read, John; McLean, Amanda; Ward, Matthew; Rogers, Daniel J.
• Is Part Of: Austral ecology. Volume 41:Number 4(2016)
• Page Start: 382
• Page End: 395
• Publication Date: 2016-01-22
• Subjects: arid -- desert -- fire -- habitat preference -- sandhill dunnart -- vegetation structure; Ecology -- Southern Hemisphere -- Periodicals; Ecology -- Australia -- Periodicals; 557
• DOI: 10.1111/aec.12323
• Languages: English
• ISSNs: 1442-9985

Abstract

Abstract: Fire and seral vegetation succession are known to influence the distribution and abundance of a wide range of arid and semi‐arid mammal species. In T riodia hummock grasslands, the gradual increase in T riodia cover after fire is a significant factor influencing mammal distribution and abundance. However, the height of fire‐adapted hummock grass species is often ignored during habitat preference studies despite the fact that taller hummocks are likely to have improved insulation properties and greater protection from predators. We tested the relative importance of a range of T riodia habitat characteristics in determining the distribution and abundance of a fire‐adapted mammal inhabiting T riodia mallee dune fields in semi‐arid Australia. We sampled 77 sites and collected information on habitat attributes including T riodia cover, height and time since fire. The 90th percentile T riodia height (>400 mm) was the most reliable predictor of sandhill dunnart abundance, and breeding, inferred through the presence of subadults. The presence of adult sandhill dunnarts was best explained by T riodia cover, increasing when cover exceeded 25%. We suggest that while Triodia cover may be an important variable for predicting the presence of adult sandhill dunnarts, the height of T riodia is important for breeding, when tall Triodia are possibly sought for nesting sites. These T riodia height and cover requirements were not recorded at sites until at least 10 years post fire but the relationship between fire and T riodia cover and height was inconsistent; after 20 years T riodia cover declined with increasing fire age while the 90th percentile T riodia height remained relatively constant. This incongruence may explain why the presence of sandhill dunnarts appears more constrained by a minimum rather than maximum time since fire and could help explain patterns of post‐fire distribution in other arid zone mammals. Importantly, the 90th percentile T riodia height highlighted the possible significance of scattered, tall T riodia hummocks for arid zone fossorial mammals.