Suchergebnisse

The importance of habitat structure has been illustrated for many groups of animals. A method to objectively measure the vegetation structure of animal habitats is presented. Factors contributing to
the index are evaluated and an attempt is made to show how this index relates to other methods of measuring habitat structure.

A technique to measure crop leaf area is adapted and applied to assessing structure (the vertical distribution of vegetation) in natural vegetation.

A light meter is used to determine a vegetation index for each of several layers. The index is compared with projected cover, cover estimated with a cover board and cover repetition using pin contacts. All correlations are highly significant. In heath the dry weight of clipped vegetation contributed more to the vegetation index than did the leaf area, reflecting a greater contribution of woody stems and branches than of the small ericoid leaves. The technique has been successfully applied to the measurement of habitat requirements for small mammals and an example of such an application is included.

The Mantel test has been widely used in many areas of research in biological science since its publication in 1967 and is particularly well suited to use in dietary studies. It is a non-parametric test that has been suggested as appropriate for comparisons when sample sizes are small. The methodology is reviewed, benefits to be gained are examined, and effects of features that have considerable impact (sample-size dependence and sensitivity to inequality of sample size) are considered.

Little is known about the biology and ecology of the eastern chestnut mouse (Pseudomys
gracilicaudatus), let alone the details of its diet. In this paper, we examined the dietary
composition, diversity and variation among individuals of this species in a coastal heathland of
New South Wales, by means of microscopic examination of its faecal pellets. From 200 faecal
samples, P. gracilicaudatus was found to consume considerable amounts of fungi (20% overall)
and insects (over 10% on average). Seed was the most abundant food in its diet from August
to February, accounting for 39%, with stem contributing 28% of the diet. Stem was dominant
at other times, accounting for 34%, with seed contributing 24%. Although cohort-dependent
foraging strategies have been reported in many small mammals, age-dependent foraging only
occurred in autumn; sex-dependent foraging was not observed in P. gracilicaudatus. Analysis of
dietary diversity (H') and among-individual variation (CVp,) showed that there were significant
variations in the diet of P. gracilicaudatus in different seasons and successional stages. Negative
association between H' and CVp, for seasons and vegetation succession suggested that individuals
of P. gracilicaudatus consumed a range of food items, producing maximum dietary diversity.
These results indicate that P. gracilicaudatus is a generalist herbivore, opportunistic in foraging
and less specialised in food selection than any other Pseudomys species reported so far.

Sites at which parma wallabies, Macropus parma, were present were compared with apparently similar
sites in nearby areas from which M. parma were absent. The structure of the habitat and the availability
of different grasses as potential diet items at each site were compared using both univariate and multivariate
statistical techniques, including analysis of variance and factor analysis. Sites with M. parma
appeared to have more Blady grass and Tussock grass and fewer Herbs and Other grass than sites
without M. parma. However, it was difficult to unequivocally allocate structural parameters that
determined the presence or absence of M. parma. Several reasons for this were considered, and it
appears most likely that some of the sites from which M. parma were absent did indeed meet the
habitat requirements of this species but remained unoccupied. This situation may result from the low
population densities and relatively sparse, disjunct distribution of this rare species. However, it will not
be possible to answer these questions, and to adequately determine details of the habitat requirements
of M. parma, without the results from studies using radiotelemetry to assess habitat use, and microscopic
techniques to assess the relative importance of items in the diet.

In its diet, S. murina is similar to other small dasyurids that have been studied, being qualitatively
opportunistic in that it feeds on a wide range of the arthropod prey available to it. Quantitatively,
however, S, murina ingests significantly more Scarabaeidae, Blattodea, Coleoptera, Lepidoptera and
larvae, and fewer Formicidae, Orthoptera and Isopoda than are available in pitfall traps during
spring-summer, and so cannot be considered opportunistic in this sense. In autumn-winter it also takes
significantly more Araneida and fewer Diptera, while Lepidoptera, Orthoptera and larvae are consumed
only in the proportions in which they occur. A. stuartii consumes significantly more Coleoptera,
Araneida and larvae than expected during the spring-summer season. In autumn-winter, it consumes
significantly more Blattodea, Orthoptera and Arthropoda as well, but significantly fewer Formicidae
and Diptera.
Quantitatively, the two species differ significantly in the distribution of prey items for the
spring-summer diet, and also differ in the individual diet categories of Scarabaeidae and larva during
winter and for Arthropoda as well, in the total diet. The diet overlap is high (0.94) and there are
significant rank correlation coefficients for diet categories, indicating no difference in the order of prey
items in a qualitative sense. When the diets are analysed on the basis of habitat-season groups for each
'species' (including pitfall traps), neither habitat nor season is significant, but some of the samples are
quite small. The 'species' effect is significant but largely reflects a difference between the diet items
available (in pitfall traps) and those actually eaten by these two dasyurids. S. murina utilizes a wider
range of prey than A. stuurtii and has a greater niche breadth; however, neither species could be called a
diet specialist except in the broad sense ofinsectivores. Diet does not seem to be a significant component
of the strong, microhabitat-based, ecological separation of these species.

The New Holland mouse, Pseudomys novaehollandiae, has been reported to have a distinct 5-month breeding season, August to early January, with occasionally an extension to March. Hence it is classified as primarily a seasonal breeder with some of the flexibility of an opportunistic breeder. We present evidence of an extension of the 1991-92 breeding season to at least 10 months, from mid-October or earlier, to late July. In May 62.5% of females were pregnant or lactating whereas 25-44% were pregnant or lactating at different sites in July. Our data from sand-mined dunes extend from Bridge Hill Ridge at the northern end of the Myall Lakes National Park to Tomago near the mouth of the Hunter River. We collate data from studies over the last 20 years, documenting extended breeding in 8 different years, 6 from one region, and at least 4 from another. Twenty years of rainfall data from nearby weather stations in the two regions show extended breeding to be a response to climate, when rainfall between September and March exceeds a threshold value of 750-800mm (approximately 120% of the long-term average), with at least 420-480mm falling during the first three months of the year (approximately 130% of the long-term average). While rainfall appears to be the ultimate factor, we support Kemper''s hypothesis that the proximal factor is most likely to be the abundance or quality of food produced by above-average late summer rainfall.

The diet of the Hastings River mouse (Pseudomys oralis) has been assessed from 90 faecal
samples collected from 42 animals in nine different locations in New South Wales. Repeated
sampling of individuals over periods of 4-6 days gave consistent results confirming the reliability
of the dietary estimation techniques used in this study. No differences were found between
males and females when tested at two different sites. Eight dietary items were identified with
the mean percentage occurrences for summer and winter, respectively, being: leaf, 45, 77; seed,
44, 12; insect, 7, 2; stem, <l , 7; flower, 2, <l ; pollen, 1, <1; fern sporangia, <1, 1; and fungi,
<l , 1. A two-way ANOVA of geographical locality and season (summer v. winter) showed no
locality effect, but a strongly significant season effect for leaf (P <0.0001), seed (P <0.0001)
and insect (P < 0.007). A similar multivariate two-way ANOVA for overall diet also showed
a significant season effect, but no locality effect. The summer co-dominance of leaf and seed
shifted to almost complete dominance of the winter diet by leaf material. This differs from the
trends in dietary composition of all other species of Pseudomys studied, which broaden their diet
in winter to incorporate many different types of food. It is not clear whether this reflects an active
choice by P. oralis to select for increased amounts of leaf, or a default selection caused by the
lack of alternative dietary items, such as fungi, which are utilised in winter by the other species.