Grey-headed Flying-fox Management Strategy for the Lower Hunter Grey-headed Flying-fox

14.To not delete these three lines (this is hidden text)

15.

The primary aim of this study is to provide a strategy for conserving the important roles foraging habitats in the Lower Hunter region currently play in the ecology and biology of GHFFs. This region contains a number of species in the blossom diet of GHFFs that produce abundant nectar relatively frequently, and that these plant species play a key role in supporting the seasonal pattern of occupation of the camps in the region, including important periods in the reproductive cycle. We also recognise that the region contains extensive tracts of Spotted Gum (Corymbia maculata), a species which flowers infrequently, but provides an unusually rich feeding resource supporting large numbers of animals for extended periods of time (Eby 1991, Pook et al.1997). The significance of this phenomenon to the species warrants targeted conservation efforts.

• 
  contain high densities of highly productive food plants;
  
• 
  are highly productive during key periods in the reproductive cycle of GHFFs (spring to autumn); and / or
  
• 
  contain high densities of Spotted Gum.
  

Flying-foxes are nocturnal, cryptic foragers. Recent satellite telemetry studies have confirmed that a large proportion of individuals feed in extensive tracts of native vegetation in relatively remote areas, where they are unlikely to be observed (Roberts 2012b, J. Martin Royal Botanic Gardens Sydney unpublished data). There are insufficient field observations of GHFFs feeding in the Lower Hunter region on which to construct an assessment of the significance of vegetation communities for the species (OEH Wildlife Atlas 2012; note Table 10.6 contains GHFF records, however few are noted as foraging observations). The evaluation presented in this report is therefore based on methods developed by Eby & Law (2008) in which the significance of vegetation communities is predicted on the basis of the flowering and fruiting characteristics of the diet plants they contain.

Native vegetation within the study area was ranked according to the quality of foraging habitat it provides for GHFFs. The methods follow Eby and Law (2008) and are described briefly here. A more detailed description can be found in their report. Eby and Law (2008) developed an index of habitat quality that is primarily a function of the flowering and fruiting characteristics of diet plants, and their patterns of distribution.

1. 
   compile a comprehensive list of plant species in the diet of GHFFs;
   
2. 
   assess and score the flowering or fruiting characteristics of diet plants, including seasonal patterns of phenology;
   
3. 
   score habitat quality on the basis of the presence and relative densities of scored dietary species as provided in vegetation classifications;
   
4. 
   incorporate key biological and ecological considerations for GHFFs into habitat scores; and
   
5. 
   classify habitat scores into ranks. Fruit and blossom diets are assessed independently with the results integrated in the final ranking.
   

15.2.1Flower Scores

Various characteristics of nectar production are significant to the assessment. High-quality dietary species are those that:

1. 
   provide relatively large volumes of food (Productivity score),
   
2. 
   are annually reliable in their productivity (Reliability score), and
   
3. 
   are productive for lengthy periods (Duration score).
   

Five attributes were used to describe these three broad characteristics and data on these attributes were acquired from quantitative field data and expert opinion (see Eby & Law 2008). Data were grouped into three or four scores per attribute, scaled from 0 (score assigned to non-diet plants) to 1 (score assigned to optimum condition).

Productivity is a function of the maximum abundance of resource available to GHFFs from an individual tree, and the spatial synchrony of flowering of the tree species in a local area, defined by a nominal commuting distance between day roost and feeding area. Abundance is considered the most significant variable in the assessment of flowering characteristics and is weighted accordingly. The other variables serve to moderate this productive potential.

Australian trees vary substantially in the frequency with which they flower from year to year and the reliability of a species moderates its productivity through time (over many years). Reliability is a measure of the frequency of substantial flowering events. It is a function of annual frequency of flowering and the proportion of those events that produce significant resources for GHFFs. Dietary species that flower reliably are likely to be of particular importance at times when many other species fail to flower for environmental reasons. Very sparse flowering (flowers present in <10% of canopy area) is unlikely to attract migratory GHFFs (Eby 1991), and is not considered.

Productivity and reliability describe different features of flowering, each of which is important to this assessment. The two scores were combined to create a single value which could be used to describe the overall characteristics of individual species within vegetation types. Productivity was weighted more highly than reliability in the calculation.

Duration is the length, in months, of a single flowering event. This variable is assessed excluding months of very sparse flowering (<10% of foliage). Species that are productive for > 3 months are assigned the highest score of 1.

The majority of species in the diet of GHFFs have clear seasonal phenologies. The annual flowering schedules of dietary plants were summarised as presence / absence of data at bi-monthly intervals. Months when flowering is sparse (<10% canopy cover) or infrequent (<20% of years) were not included.

Feeding habitats of GHFFs were defined by the vegetation communities, or Map Units (MUs), described in the Greater Hunter Native Vegetation Mapping Project version 4 (GHMv4; Sommerville 2009; Sivertsen et al. 2011). The full map for the project, which encompasses the Greater Hunter region, was clipped to the boundary of the Lower Hunter study area. One hundred and five (105) MUs occur in the study area. Their significance as feeding habitat for GHFFs was described from the species richness, relative density and flower scores of the dietary species they contain.

The frequency–cover abundance method was used to estimate the relative densities of dietary species (Eby and Law 2008). The occurrence of species in MUs is summarised numerically in tables of standard data collected in field samples (Sommerville 2009). Frequency values (f) of canopy species are the proportion of field samples of the MU in which the species was recorded. Cover abundance scores (C/A) are medians from the field samples, scored on a 6 class Braun-Blanquet scale (Poore 1955). Tree species with f <0.3 or C/A scores <2 were excluded from calculations of habitat scores due to their infrequent or sparse occurrence in the MU.

1. the density estimate of each canopy species in a vegetation type was calculated as:

d = f * C/A;

2. the relative density of each dietary species was calculated as the density estimate of that species divided by the sum of densities of all canopy species in the MU:

Rd = di/Σ(d_1-k);

3. the density-weighted nectar score for each dietary species was calculated as:

NS = Rd * species nectar score; and

4. finally, the total nectar score for the habitat is the sum of density-weighted nectar scores:

Ts = Σ(NS_1-k).
Nine MUs present in the GHVM are not described in the profiles provided by Sommerville (2009). Information on species present in the canopy of these MUs was acquired from a spreadsheet of data provided by OEH (GreaterHunter_Draft_100812.xls; OEH 2012). The density estimates of dietary species were produced by averaging plants listed as characteristic of the upper stratum (averaging method of Eby & Law 2008).
Scores of productivity, reliability, and wt p*r were calculated separately for each MU. Bi-monthly habitat scores were generated by including in calculations only those species that are productive in each bi-month. Dietary species that were not productive in a given bi-month were assigned scores of zero. For each MU, bi-monthly habitat scores were produced for productivity, reliability and wt p*r.
15.2.3Fruit Habitat Score

Fruit-bearing vegetation types (usually rainforest) were scored on the basis of the species richness of dietary plants. Types that contained >10 species were assigned the highest score, habitats with 5-9 species were assigned an intermediate score, habitats with <5 species were assigned a low score.

A primary aim of this project was to identify habitat necessary to secure into the future forage for GHFFs during key periods in the reproductive cycle (spring to autumn). This consideration introduces a temporal element to the ranking process which we accommodated by conducting separate assessments of habitat quality in each bi-monthly period. Final habitat ranks were assigned to MUs using the highest rank achieved in any bi-monthly interval. This procedure ensured that the critical, short-term role of highly seasonal habitats was captured in the ranking process.

The weighted productivity * reliability scores of habitats were used to assign ranks to MUs in each bi-monthly period. Habitat scores were classified into four ranks of equal land area, with a value of 1 being assigned to the highest ranking habitat.

1. sort wt p*r habitat scores in descending order;

2. calculate the total area of habitat that has the potential to be productive in the bi-month being considered (the total area of each MU was derived from the digital map layer);

3. using the total productive area as the base, calculate the boundary of quartiles of equal area; and

4. allocate MUs to ranks in descending order of wt p*r scores until bounds for each category as defined by the equal area value were reached.
Using this method, the most productive 25% of vegetation productive in each bi-monthly interval is identified as having the highest rank for habitat conservation. The final nectar rank of a vegetation type was taken as the highest bi-monthly rank assigned to it. This ensured that the maximum value of a vegetation type was considered in assessments of conservation significance.
15.2.4.2Bimonthly Habitat Ranks (Fruit)

The reliable nature of fruiting phenologies in dietary plants was considered of particular benefit to GHFFs, providing relatively predictable feeding habitat. A rank of two was subjectively assigned to rainforest habitats containing >5 diet plants, a rank of three was assigned to habitats with <5 diet plants. Some MUs contained rainforest with wet sclerophyll emergents. Both nectar and fruit scores were generated for these communities and the highest rank was taken as the final rank for the MUs.

The method described above addresses the first two aims of this assessment. It sets high conservation ranks for vegetation that contains high densities of productive food plants, and it takes into consideration variations in productivity that occur through time, targeting vegetation that provides foraging opportunities for GHFFs during key periods in their reproductive cycle (spring to autumn). Additional consideration was given to the third aim, to conserve vegetation that contains high densities of Spotted Gum. Nine MUs were identified as containing Spotted Gum at sufficient frequency and C/A scores to include them in habitat assessments. The relative density of Spotted Gum in each of these MUs was calculated and the rank assigned to MUs containing the species was noted. Where there was a clear discrepancy between the density of the species and the rank assigned to the MU, consideration was given to elevating the conservation rank that was assigned.

15.3.1Diet Plants

Twenty-seven (27) species of plants in the nectar diet of GHFF occur in MUs in the Lower Hunter study area (refer to Table 15.4). The list contains 27 species in the Myrtaceae: two Angophora, three Corymbia, 19 Eucalyptus and one each of Melaleuca and Syncarpia. There is also one species of Banksia.

The 27 dietary plants in the region vary widely in their productivity scores (range 0.35 - 1.0, median = 0.7) and reliability scores (range = 0.15 – 1.0, median = 0.6). Weighted productivity x reliability scores range from 0.35 (A. costata) to 1.0 (E. robusta). The median value is 0.56. Twelve species (44%) score in the upper quartile of all species in the nectar diet of the animals (wt p*r ≥0.65, Eby & Law 2008). These species are considered as highly productive food plants for the purpose of this assessment.

The ranges of 38 species of rainforest trees and lianas in the fruit diet of GHFFs fall within the Lower Hunter region (refer to Table 15.6). The regional list comprises members of 27 families and 31 genera. Four genera are represented by more than one species. The most species rich genus is Ficus (6 spp.).