Sign up

FORESTCHECK: terrestrial vertebrate associations with fox control and silviculture in jarrah (Eucalyptus marginata) forest.
Abstract:
Terrestrial vertebrate associations with silviculture and other factors were investigated as part of the FORESTCHECK monitoring program in the jarrah (Eucalytpus marginata) forests of south-west Western Australia. A total of 48 integrated monitoring grids form the basis of this study--sampled over five years (2001-2006), across five ecosystem-defined regions (one sampled per year), each with replicates of two silvicultural treatments (shelterwood, gap release) and external reference forest (uncut forest or structurally mature forest that had not been harvested for timber for over 40 y). Terrestrial vertebrates were surveyed in spring and autumn using pitfall traps and wire cages. Forty-one terrestrial vertebrate taxa (8 frogs, 22 reptiles, 11 mammals) comprising 1165 captures were recorded.

Fox (Vulpes vulpes) control had the strongest effect on terrestrial vertebrates, with baited areas supporting significantly more individuals (three-fold increase) than unbaited areas. The mammals Trichosurus vulpecula, Bettongia penicillata, Cercartetus concinnus and Dasyurus geoffroii, and the skink Tiliqua rugosa were particularly more abundant in fox-baited forest.

Several terrestrial vertebrate community attributes (species accumulations by grids and number of individuals, dominance--diversity plots, overall community structure and overall abundance) differed little among the three treatments (i.e. two silvicultural, plus external reference forest). However, external reference grids had significantly lower species richness than the shelterwood grids and a significantly different community structure. These differences resulted from a greater prevalence within shelterwood of some species such as the reptiles Egernia napoleonis, Menetia greyii, Ctenotus labillardieri and Ramphotyphlops australis. Forests that had never been harvested, a subset (8/15 grids) of the external reference treatment, had the lowest overall abundance, due largely to a confounding with fox control. The level of replication enabled differences between treatments of greater than 23% in species richness, and 37% in overall abundance, to be detected as statistically significant.

Significant ecosystem/year differences were found. Differences in community structure between ecosystems/years approximated the geographic/bioclimatic relationships between the grids, with the distinction between southern jarrah communities (Jarrah South/2001-02 and Jarrah Blackwood Plateau/2005-06) and the northern communities being particularly apparent. Time since last fire, live tree basal area, and the proportion of basal area removed by harvesting and silvicultural treatment were not correlated with vertebrate species richness, abundance or community structure. In comparison to the effect of fox control and regional/temporal variation, silvicultural treatment and the intensity of timber harvesting had minor impacts.

Suggestions for the improvement of this and similar studies are discussed, with a particular focus on reducing residual variance and increasing sample size.

Keywords: harvesting; silviculture; forest management; monitoring; pest control; fauna; mammals; foxes; reptiles; frogs; jarrah; Eucalyptus marginata; Western Australia

Article Type:
Report
Subject:
Sustainable forestry (Research)
Forest management (Research)
Vertebrates (Environmental aspects)
Arboriculture (Research)
Eucalyptus (Environmental aspects)
Authors:
Wayne, Adrian F.
Liddelow, Graeme L.
Williams, Matthew R.
Pub Date:
12/01/2011
Publication:
Name: Australian Forestry Publisher: Institute of Foresters of Australia Audience: Academic Format: Magazine/Journal Subject: Forest products industry Copyright: COPYRIGHT 2011 Institute of Foresters of Australia ISSN: 0004-9158
Issue:
Date: Dec, 2011 Source Volume: 74 Source Issue: 4
Topic:
Event Code: 310 Science & research
Geographic:
Geographic Scope: Australia Geographic Code: 8AUST Australia
Accession Number:
276438085
Full Text:
Introduction

Although forests cover only 5% of Australia's landmass, they support almost half of the terrestrial vertebrates, including 77%, 53% and 33% of all mammal, frog and reptile species, respectively (Lamb and Smyth 2003). Furthermore, 75% of these are endemic forest species (Lamb and Smyth 2003).

Timber harvesting is a significant form of disturbance in those Australian forests where it is practiced and its effects on fauna remains contentious (e.g. Recher 1985; Lindenmayer and Franklin 2002; Calver and Wardell-Johnson 2004). Mammal responses and associations to timber harvesting have received considerable attention both in Western Australia (WA) and elsewhere in Australia and include studies of small terrestrial mammals (e.g. Barry 1984; Kavanagh and Webb 1998; Lunney et al. 2009), bats (e.g. Lloyd et al. 2006; Adams et al. 2009) and particularly arboreal mammals (e.g. Kavanagh and Bamkin 1995; Laurance and Laurance 1996; Lindenmayer et al. 1999; Lindenmayer 2000; Wayne et al. 2006). Reviewers of this information include Lindenmayer (1994), Gibbons and Lindenmayer (2002; hollow-users), Lindenmayer and Franklin (2002), Lindenmayer and Hobbs (2004; plantation forests) and Lindenmayer et al. (2008; salvage logging).

Vertebrates that use hollows are considered to be especially sensitive to timber harvesting in eucalypt forests (e.g. Christensen 1997; Abbott and Whitford 2002; Gibbons and Lindenmayer 2002; Wardell-Johnson et al. 2004; Kavanagh and Stanton 2005). About 75% of mammals in the forests of south-west WA are considered to use hollows to some extent (Christensen 1997) and are similarly recognised as an important habitat element for many other taxa. As a consequence, hollows and the fauna that use them have been the primary focus of studies on the effects of silviculture on fauna in south-west WA. These include investigations of the distribution, abundance and characteristics of hollow logs (Faunt 1992; Williams and Faunt 1997) and tree hollows (Whitford and Williams 2000, 2002; Whitford 2001, 2002), and their predicted availability to hollow users (Abbott and Whitford 2002). Arboreal and semi-arboreal mammals, including the ngwayir, Pseudocheirus occidentalis (Inions et al. 1989; Jones et al. 1994a,b; Jones and Hillcox 1995; Wayne et al. 2000, 2005a, 2006; Jones 2004), koomal, Trichosurus vulpecula hypoleucus (Inions et al. 1989; Jones and Hillcox 1995; Morris et al. 2001; Wayne et al. 2005a, Wayne unpublished data) and wambenger, Phascogale sp. WAM M434, formerly P. tapoatafa (Rhind 1996a,b, 1998, 2002, 2004; Rhind and Bradley 2002), have also been the subject of detailed investigation.

In southern NSW, many small terrestrial mammals and reptiles (including arboreal, terrestrial and fossorial species) were not affected by logging or recovered quickly (Webb 1995; Goldingay et al. 1996; Kavanagh and Webb 1998). However, some reptiles were adversely affected at various periods after logging in coastal forests (Lunney et al. 1991; Webb 1995), but others in montane forests had significantly higher abundances at logged sites (Goldingay et al. 1996). In East Gippsland, no reptile species appeared to be disadvantaged by timber harvesting and one species may have benefited from forest regrowth, but reptile species diversity was adversely affected by harvesting in the lowland forests and favoured in the wet forests (Alexander et al. 2002).

Responses of frogs to disturbance are poorly known in Australia compared with elsewhere (Gillespie 2002; Hazell 2003). The few studies undertaken indicate a pattern of response broadly consistent with other taxa; some species are disadvantaged and some are favoured. In NSW, frog species richness and abundance were positively associated with logging disturbance (Lemckert 1999), but other species were considered to be disadvantaged (Lemckert 1999; Gillespie 2002; Kavanagh and Stanton 2005; Penman et al. 2005, 2008). Responses within the same taxon may also vary. For example, Crinea signifera is a well-studied frog that appears relatively tolerant to disturbance associated with silviculture (Baker and Lauck 2006; Lauck 2006), but its responses to logging varied (Kavanagh and Webb 1998; Lemckert 1999; Kavanagh and Stanton 2005; Lauck 2006).

Published empirical studies on terrestrial vertebrates in relation to silviculture in south-west WA are limited but do include reports on several mammals: mardo, Antechinus flavipes (Wardell-Johnson 1986), numbat, Myrmecobius fasciatus (Friend 1994), chuditch, Dasyurus geoffroii (Serena et al. 1991; Morris et al. 2003), woylie, Bettongia penicillata and quenda, Isoodon obesulus fusciventor (Morris et al. 2003). Some bats have also been recently studied (Webala et al. 2010, 2011). Reviews of the fauna responses to silviculture in south-west WA include those by Nichols and Muir (1989), Wardell-Johnson and Nichols (1991), Christensen (1997) and Calver and Dell (1998a,b). There are currently no peer-reviewed publications that quantitatively relate effects of silviculture on reptiles or amphibians in Western Australian forests, and up until recently were severely lacking for elsewhere in Australia (e.g. Gillespie 2002; Hazell 2003).

Overall, responses of terrestrial vertebrates to the disturbance associated with timber harvesting vary. Some species are favoured, others disadvantaged, and in many cases there is no apparent effect or insufficient data to reliably assess responses (e.g. Woinarski and Fisher 1995; Goldingay et al. 1996; Kavanagh and Webb 1998; Lindenmayer and Franklin 2002; Kavanagh and Stanton 2005).

The jarrah (Eucalyptus marginata) forest of south-west WA supports a particularly high level of biodiversity (Wardell-Johnson and Horwitz 1996), and as such represents one of the most important Australian bioregions for the conservation of numerous species that have disappeared from much of their former ranges (National Land and Water Resources Audit 2002). Although considerable research into disturbance ecology has been carried out in the region (Wardell-Johnson and Nichols 1991), the lack of appropriate studies of the causes of declining populations of native species seriously hampers effective conservation of local biodiversity (Christensen and Abbott 1989; Nichols and Muir 1989; Christensen 1997; Calver and Dell 1998a,b; Friend and Wayne 2003; Wardell-Johnson et al. 2004).

FORESTCHECK is an integrated monitoring project designed to inform forest managers about changes and trends in key elements of forest biodiversity associated with management activities in Western Australia (Abbott and Burrows 2004; McCaw et al. 2011). The initial focus has been on monitoring the effects of silvicultural treatment in jarrah forest at replicated grids and at a regional level in relation to vascular plants, cryptograms, epigeous macrofungi, invertebrates and vertebrates. While FORESTCHECK is designed as a long-term monitoring program, it also provides an opportunity to investigate fauna responses to silviculture. Therefore the main aim of this study is to determine whether terrestrial vertebrate associations in jarrah forest subjected to silvicultural treatments differ among treatments and relative to 'external reference' areas, these being areas of comparable mature forest that has either never been harvested or not harvested for more than 40 y. The response measures investigated included abundance of some individual species, overall species richness and abundance, and community structure. As such this is the first published account of an empirical study of the associations of frogs, reptiles and some mammal species with silviculture in a Western Australian forest.

Methods

Monitoring grids

Details of methods common to this study and other biota groups monitored in FORESTCHECK are provided by McCaw et al. (2011). Briefly, 48 monitoring grids were established in four jarrah forest ecosystems: Jarrah South (JS), Jarrah North West (JNW), Jarrah North East (JNE) and Jarrah Blackwood Plateau (JB) (McCaw et al. 2011). Grids were established in two locations in the JNW ecosystem--the central (JNW-C) and northern (JNW-N) subregions to cover its extensive north--south range. All grids from one ecosystem (or location) were monitored in any one year, with different ecosystems surveyed in successive years (2001-2006) (Table 1). Treatments were shelterwood (12 grids), selective cut (3 grids), gap release (14 grids), coupe buffer (4 grids) and external reference forest (15 grids). Coupe buffers were uncut temporary reserves within areas of harvested forest. External reference grids were established in mature forest in conservation reserves, national park and state forest, and included forest that had never been harvested (8 grids) or had not been harvested for at least 40 y (7 grids). All silvicultural treatments (shelterwood, selective cut and gap release) had been undertaken since 1988, and in each ecosystem were stratified across time since harvest (McCaw et al. 2011). Descriptive statistics for coupe buffer and selective cut treatments are presented but not included in comparative analyses due to the low representation of these treatments. Through timber harvesting, the basal area of shelterwood grids was reduced by an average of 45% and of gap release grids by 70%, and the average size of gaps was 9 ha. At the time of monitoring the average basal area of grids in the reference, shelterwood and gap release treatments was 41, 22 and 16 [m.sup.2] [ha.sup.-1] respectively (McCaw 2011). The attributes of each monitoring grid, including history and time since last treatment (silviculture and/or prescribed fire), and detailed descriptions of the forest ecosystems, silvicultural treatments and other site attributes are included in McCaw et al. (2011).

Terrestrial vertebrate surveys

At each of the 48 monitoring grids 15 pitfall traps were arranged in three transects of five pitfalls spaced 20 m apart along transects and 25 m between transects (see Fig. 3 in McCaw et al. 2011). Each pitfall trap comprised a 20 L plastic bucket (25 cm diameter x 40 cm deep) with a 5-m-long fly-wire 'drift' fence located centrally across the bucket. The drift fence was kept vertical with stakes to a height of at least 20 cm high above ground level and was buried to a depth of at least 5 cm where possible. A polystyrene meat tray or egg carton and a handful of sand and leaf litter were provided as shelter in the bottom of each bucket.

Fifteen wire cage traps (20 cm x 20 cm x 45 cm) were arranged in three transects of five traps each spaced 50 m apart both within and between transects (see Fig. 3 in McCaw et al. 2011). Traps were located in sheltered positions with thick hessian bags overlaid to provide protection to captive fauna and baited with a variant of 'universal' bait (peanut butter, rolled oats and emu oil). Pitfall and cage traps were run simultaneously on all grids in each ecosystem for four consecutive nights in both spring (November--December) and autumn (March-May). All captive animals were identified and processed at the point of capture, and released immediately afterward. Mammals were fitted with individual identity ear tags and their sex, weight and breeding status recorded.

Dunnarts (Sminthopsis spp.) were aggregated at the genus level given the difficulties of unequivocally distinguishing between S. gilbertii and S. griseoventer by field identification of live specimens. Similarly, 'Lerista spp.' skinks reported here is an aggregation of multiple possible species within the L. distinguenda group (including L. distinguenda, L. elegans and L. microtis), and 'Morethia spp.' skinks is an aggregation of M. obscura and possibly some M. linocellata specimens. This is a conservative taxonomic approach, given that of the vouchered specimens, all Lerista were confirmed by the WA Museum as L. distinguenda and all Morethia were confirmed as M. obscura, but some unvouchered field identifications recorded L. elegans, L. microtis and M. linocellata.

Data analysis

Terrestrial vertebrate responses were analysed according to silvicultural treatment: (1) gap release, (2) shelterwood and (3) external reference forest. Preliminary analyses also indicated significant differences between shelterwood and selective cut grids, and between coupe buffer and external reference grids. These findings were consistent with an earlier study in the JS ecosystem (Burrows et al. 1994) which indicated that vertebrate fauna respond differently within the coupe buffers to either silviculturally treated or contemporarily unharvested forest (Morris et al. 2001; Wayne unpublished data). While it may be appropriate to include coupe buffer grids as part of the reference treatment for sedentary or non-volant taxa (McCaw et al. 2011 and individual FORESTCHECK papers within this issue) this is not appropriate for terrestrial vertebrates. Within the remaining 41 FORESTCHECK monitoring grids upon which this analysis was based, there was some bias in the replication across ecosystems/regions in the shelterwood treatment (Table 1) that has been considered when analysing the data and interpreting the results.

Whether the grids were located in forest areas subject to routine aerial fox control as part of the Western Shield program (Armstrong 2004; Orell 2004) was not considered at the time of site selection but is indicated in Table 1. Shelterwood, gap release and external reference treatments are not confounded with fox control, with about half the grids in each being subject to regular fox baiting. However, within each of the external reference subsets there is confounding: six of the eight never-harvested grids and two of the seven previously harvested reference grids were not subject to fox control. There was also confounding within some ecosystems with respect to fox control (JNW-C, JNW-N and JB) and between ecosystems over time (i.e. one ecosystem sampled per year) (Table 1). The analyses and interpretation of the results from this study has, therefore, given due regard to these factors.

Analysis followed the standard protocol used for all biota groups examined in FORESTCHECK (see 'Data analysis', McCaw et al. 2011) and is summarised here. Species accumulation curves were determined for each treatment and for the total sampling effort over the five sample years using Estimates software (Colwell 2005). Estimates of total species richness were calculated using statistical extrapolation and the first-order jackknife estimator (JACK1) (Heltsche and Forrester 1983).

Differences in terrestrial vertebrate communities between treatments were examined using four complementary approaches. Dominance--diversity curves were constructed to give a visual representation of population abundance structure in each treatment. Two multivariate procedures, non-metric multidimensional scaling (nMDS) (Kruskal 1964) and canonical analysis of principal coordinates (CAP) (Anderson and Robinson 2003), were used to examine differences in species assemblages between treatments and forest ecosystems, and overall differences in species assemblages between treatments and ecosystems were tested using permutation multivariate analysis of variance (PerMANOVA) (Anderson 2001). All nMDS, CAP and PerMANOVA analyses used PRIMER software (Clarke and Gorley 2006).

Multivariate linear modelling was used to assess the effects of silvicultural treatment, ecosystem/year, live standing tree basal area, years since harvested, years since last burnt, and fox control on vertebrate abundance and species richness, using the GLM procedure in the SAS software (SAS Institute Inc. 2004). Live standing basal area and years since last burnt were continuous covariates, whereas all other effects were categorical; years since harvested was coded into four categories (0-4 y, 5-9 y, 10 y or more, or never harvested); and fox control as binary yes/no. Initially, a model including all main effects was fitted using the model-building strategy of Hosmer and Lemeshow (2000). Because the sample size was relatively small, any clearly non-significant effects (P > 0.15) other than treatment (the main variable of interest) were sequentially excluded from the main-effects model and the remaining effects re-evaluated. Once final models containing only the significant main effects and treatment were obtained, single degree-of-freedom contrasts were used to test specific a priori hypotheses of interest and the means and standard errors for each effect calculated, after adjusting for other effects. Examination of residuals was used to assess the assumptions of normality and constant variance using standard methods (Zuur et al. 2010), and abundance data were log-transformed to meet these assumptions. Sample size was insufficient to permit examination of interactions between effects.

Detectable effect sizes for species richness and abundance were determined using standard methods, based on the results of analysis of variance (ANOVA) (Osenberg et al. 1994).

Results

Species composition

Forty-one terrestrial vertebrate taxa (8 frogs, 22 reptiles, 11 mammals) were recorded from pitfall and wire cage traps (Table 2). Four taxa constituted 55% of all capture records (n = 1165): the mammal T. vulpecula (22%), Lerista spp. (14%) and Morethia spp. (11%) skinks, and the mammal B. penicillata (8%). Eight species were recorded only once (i.e. singletons); seven from the coupe buffer and external reference grids (1, 3 and 3 species respectively at a coupe buffer, an external reference grid harvested more than 40 y ago and a never-harvested external reference grid) and one, the snake Notechis scutatus, from a shelterwood grid. Except for the eight singletons and the frog Pseudophryne guentheri (two records), no species was detected exclusively within one treatment (Table 2). Eight of the 10 species recorded from only one ecosystem/year were from either JS (2001-02) or JB (2005-06). Another 10 species were recorded from two ecosystems/years, four species from three, 10 species from four and only seven species were recorded in all five ecosystems/ years.

Several species were detected in all but one treatment (Table 2); the frog Heleioporus eyrei was not detected in an external reference grid (n = 15), the snake Ramphotyphlops australis was not detected in a coupe buffer grid (n = 4) and the frogs Crinia georgiana and Limnodynastes dorsalis, the reptiles Acritoscincus trilineatum, Menetia greyii and Pogona minor, and the mammals B. penicillata and D. geoffroii were not detected in a selective cut grid (n = 3).

The mean abundance of the reptile Egernia napoleonis was greater on the shelterwood grids relative to external reference and gap release treatments (P < 0.01 and P < 0.02, respectively). Similarly, R. australis was more abundant in the shelterwood treatment relative to external reference treatment (P = 0.01). There were no other significant pairwise differences in the mean abundance of terrestrial vertebrate species between the two main silvicultural treatments (shelterwood, gap release) and external reference forest.

Species accumulation curves

Rates of species accumulation were similar for gap release, shelterwood and external reference grids (Fig. 1), but there was a trend for the harvested treatments (shelterwood and gap release) to initially increase more rapidly but have slightly fewer species at 300 individuals compared to the external reference treatment. Thirty-four species were detected at the 15 external reference grids, 32 species at the 14 gap release grids and 29 species at the 12 shelterwood grids. At n = 12 grids, the modelled species richness was 31 for both external reference and gap release treatments.

The number of species increased relative to the number of captured individuals marginally more rapidly in the external reference treatment compared with those subjected to harvesting, particularly the shelterwood treatment (Fig. 2). Species accumulation rates for each treatment did not approach asymptotes within the range of individuals sampled (286-375 individuals each treatment), although when all grids were combined (i.e. increased sample size) the species accumulation curve approached an approximate asymptote when more than 750 individuals were sampled.

Dominance--diversity plots

Dominance--diversity curves were similar for gap release, shelterwood and external reference treatments (Fig. 3). The same three species were the most common in all treatments but their rankings varied. Within the external reference treatment, the mammal T. vulpecula and the skinks Lerista spp. and Morethia spp. were the three most common species respectively. Within the shelterwood treatment the descending ranking was Lerista spp., T. vulpecula and Morethia spp. Within the gap release treatment, Lerista spp. and Morethia spp. were equally the most abundant, followed by T. vulpecula. The relative dominance of the most abundant species was greatest in the external reference treatment (T. vulpecula was roughly twice as abundant as the next most common species) and least in the gap release (the abundances of the three most common species were roughly equal) (Fig. 3). The proportion of species recorded only once or twice was 34% for both shelterwood and gap release treatments and 47% in the external reference treatment.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Species richness and abundance

Multivariate linear modelling was used to assess the effects of silvicultural treatment, ecosystem/year, live tree basal area, years since harvested, years since last burnt, and fox control on vertebrate abundance and species richness. Fox control (P < 0.01), time since harvest (P < 0.01) and ecosystem/year (P = 0.02) significantly affected abundance but silvicultural treatment did not (P = 0.26), whereas only silvicultural treatment affected species richness (P = 0.03). Live basal area and years since last burnt did not affect abundance (P = 0.19 and P = 0.46, respectively) or richness (P = 0.27 and P = 0.56, respectively).

The means for each of these effects were compared using single-df contrasts. Fox-baited sites had significantly higher abundance (33 vs 10 individuals, P < 0.01) and slightly higher richness (9 vs 7 species, P = 0.17). The sites that had never been harvested had the lowest abundance (11), less than the 18 for 1-4 y, and significantly less than the 23 and 26 for 5-9 or 10+ years since harvest, respectively. Although abundance varied significantly between ecosystem/years (range 11-35), there was no difference between silvicultural treatments (Tables 3, 4). A combined comparison of harvested treatments with external reference forest also showed no significant difference (P = 0.78). Richness, however, was significantly lower in external reference grids (by 30%; Table 4).

[FIGURE 3 OMITTED]

Compared with areas baited for foxes (n = 21 grids), unbaited grids (n = 20) had, on average, substantially fewer koomals (T vulpecula) (mean 8.57 vs 0.10, respectively; P < 0.01), Tiliqua rugosa skinks (mean 0.52 vs 0.05, respectively; P = 0.01), and woylies (B. penicillata) (mean 3.90 vs 0.15, respectively; P = 0.03), and marginally fewer pygmy possums (Cercartetus concinnus) (mean 0.66 vs 0.05, respectively; P = 0.05), and chuditches (D. geoffroii) (mean 0.62 vs 0.15, respectively; P = 0.06).

Never-harvested grids were confounded with fox baiting history (6/8 grids were not fox-baited) and predominantly located in the JNW-C and JNE ecosystems. The most common taxa recorded were either not detected at the never-harvested grids (T. vulpecula and B. penicillata) or much less frequently so than at other grids (Lerista spp. and H. peroni skinks).

Community structure

Modelling (PerMANOVA) of community structure involving silvicultural treatments, the five ecosystems/years and the interaction term between these two factors revealed no significant relationship for the interactive term (P = 0.58), a significant difference in community structure according to ecosystem/ year (P < 0.01), but no overall significant difference between silvicultural treatments (P = 0.08). Nonetheless, a priori pair-wise comparisons of community structure between treatments revealed a significant difference between external reference forest and shelterwood (P = 0.01) (Table 4). The two pair-wise comparisons with gap release were not significant (P = 0.30 and P = 0.49 for shelterwood and external reference treatments, respectively).

The nMDS ordination of community structure did not discriminate silvicultural treatments (two-dimension stress = 0.22) (Fig. 4). The CAP ordination also showed no clear separation between the silvicultural treatments (P = 0.22) (Fig. 5), but, consistent with the perMANOVA, shelterwood and external reference treatments did tend to differentiate to some extent along axis one. To determine which species were associated with silvicultural treatment, correlations between abundances and the axis scores were calculated. The reptiles E. napoleonis, M. greyii, Ctenotus labillardieri and to a lesser extent R. australis were more prevalent or abundant on harvested grids (i.e. their abundances were significantly and negatively correlated with axis 1). No taxa were positively associated with external reference forest.

Consistent with the PerMANOVA analysis, the CAP ordination of community structure revealed highly significant differences between ecosystem/year of sampling (P = 0.0001) (Fig. 6). The ordination clearly distinguished and grouped grids within the same ecosystem/year and generally reflected the geographic/ bioclimatic relationship between grids (axis one and two approximate the north--south and west--east geographic spread of grids, respectively). Communities at the two locations in the JNW ecosystem were not readily distinguished from each other. The prevalence and abundance of the reptile Hemiergis peroni and koomals (T. vulpecula) were most strongly associated with grids towards the left-hand side of axis one (i.e. southern grids) and Lerista spp.and M. greyi skinks were more strongly associated with grids toward the right-hand side of axis one (i.e. had correlations greater than 0.58 or less than -0.58). Key taxa discriminating grids on axis two were the woylie (B. penicillata) (more prevalent at grids toward the bottom of the graph) and the skink C. labillardieri (more prevalent at grids toward the top of the graph) (i.e. had correlations greater than 0.40 or less than -0.40).

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Terrestrial vertebrates in relation to standing tree basal area

Plots revealed no clear relationship between live tree basal area and either species richness or total terrestrial vertebrate abundance, consistent with the multivariate modelling described above. There was also no significant relationship between community structure and live standing tree basal area (PerMANOVA, P = 0.09, n = 41 grids) after accounting for the significant ecosystem/year factor (P < 0.01) and its interaction with the basal area covariate (P = 0.40). Similarly there was no significant relationship between community structure and the proportion of basal area removed during the silvicultural treatment (PerMANOVA, P = 0.37, n = 41 grids) after accounting for the significant ecosystem/year factor (P < 0.01) and their interaction (P = 0.26).

Inferential power

The detectable effect size (with 95% confidence) (Osenberg et al. 1994) for the current Forestcheck terrestrial vertebrate monitoring design and data (41 grids, 3 treatments, 12-15 grids per treatment) is 23% for species richness and for 37% for abundance. Table 5 provides estimates of the number of monitoring grids required to achieve various detectable effect sizes.

Discussion

Silvicultural treatment differences

Several terrestrial vertebrate community attributes were similar in silvicultural treatments and reference forests, including species accumulations by grids and number of individuals, dominance--diversity plots, overall community structure and overall abundance. External reference grids, however, had significantly lower species richness than shelterwood grids and a significantly different community structure. These differences resulted from a greater prevalence within shelterwood of some species such as the reptiles Egernia napoleonis, Morethia greyii, Ctenotus labillardieri and Ramphotyphlops australis. These likely benefit from the larger amounts of coarse woody debris in harvested grids (McCaw 2011) and the intermediate state of forest structure and diversity available in shelterwoods (e.g. moderately open forest structure; increased ectothermic opportunities such as basking sites and the nature, diversity and distribution of micro habitats). Egernia napoleonis shelters in and forages from the crevices of larger logs (Bush et al. 1995; pers. obs.), the abundance of which is likely to be a key factor in determining skink densities (Craig et al. 2011). Although C. labillardieri is predominantly a rock-dwelling species it also shelters under logs and bark (Cogger 1994). A subterranean specialist, the distribution and abundance of R. australis is associated with soil and invertebrate prey, particularly ants and termites (Ehmann and Bamford 1993; Cogger 1994) that may be associated with coarse woody debris. Soil disturbance used to promote jarrah seedling establishment in shelterwoods may also benefit R. australis. Morethia greyi is a widespread habitat generalist found in a range of environments from dry sclerophyll forest to arid scrub (Cogger 1994) and can be found in disturbed sites, including suburban backyards (Bush et al. 1995).

The most abundant terrestrial mammals recorded in this study were Trichosurus vulpecula and Bettongia penicillata. Neither species displayed a strong response to silvicultural treatment. In comparable areas of jarrah forest, B. penicillata abundance was similar in harvested forest and external reference grids but T. vulpecula decreased significantly immediately (less than 2 y) after shelterwood treatment, in gap releases where retained habitat trees were experimentally removed and in adjacent unharvested buffers (Morris et al. 2001). Because of the replicated BACI (before--after, control--impact) design of the Morris et al. (2001) study, the results are likely to be sensitive and reflect the true short-term responses of these mammals (e.g. Underwood et al. 1996; Kavanagh and Stanton 2005). Since this study found no detectable effects on T. vulpecula at sites silviculturally treated more than 2 y previously, it appears that any effects may be transitory.

For frogs, both generally and with this study, difficulties with sampling and highly variable measures of abundance over space (e.g. Goldingay et al. 1996) and time (e.g. diurnal to annual variation due to weather, season, etc.; deMaynadier and Hunter 1995) have resulted in there being insufficient information to understand their responses to silvicultural treatment (e.g. Kavanagh and Webb 1998).

In this study, lower vertebrate abundances on never-harvested grids largely reflected the absence of otherwise common mammals T. vulpecula and B. penicillata, and lower abundances of otherwise common reptiles. These differences are best explained by the confounding of the never-harvested external reference treatment with fox control (6/8 grids not fox-baited). Also, most of the never-harvested grids were located in the JNW-C and JNE ecosystems, which accounts for the lower abundance of common reptiles such as the skinks, Hemiergisperoni (more prevalent in JS ecosystem) and Lerista spp. (more prevalent in JNW ecosystem). While there is confounding within the external reference treatment subsets, differences between gap release and shelterwood silvicultural treatments and the external reference treatment have been analysed with fox control as a covariate, enabling treatment effects to be examined independently.

Kavanagh and Stanton (2005) reported the response of 227 vertebrate species (including bats and birds) to selective timber harvesting at 487 sites across north-eastern New South Wales (NSW). In frequency of occurrence, 18% were significantly disadvantaged, 18% were favoured and 65% were apparently unaffected, but for many there were insufficient data to assess their response (Kavanagh and Stanton 2005). Also, more species were less common in selectively harvested forest compared with both unharvested and intensively harvested forests, and more mammal species were disadvantaged than favoured by harvesting (11 versus 4 species, respectively). Reptiles comprised the largest proportion (39%) of species sensitive to harvesting, with many species responding differently according to the intensity of harvesting. No species displayed consistently strong preferences for selectively harvested forest. Their findings contrast with our study, which found higher mean species richness and overall abundances in shelterwood (a treatment of moderate intensity), and two reptiles (E. napoleonis and R. australis) were more abundant in shelterwood compared with the external reference treatment: the only significant pair-wise differences.

Fox control

In this study, fox control had the strongest effect on terrestrial vertebrate abundance. Areas regularly aerially baited (Armstrong 2004; Wyre 2004) supported significantly more individuals (three-fold increase) and around 20% more species than unbaited areas, although this latter effect is not statistically significant. Several conservation-listed medium-sized mammals (T. vulpecula, B. penicillata and Dasyuris geoffroii), the western pygmy possum (Cercartetus concinnus) and at least one large lizard (Tiliqua rugosa) were particularly more abundant in fox-baited forest. This finding is consistent with other studies that indicate foxes and fox control influence the conservation status, distribution and abundance of many species, including T. vulpecula, B. penicillata and D. geoffroii (e.g. How and Hillcox 2000; Burrows and Christensen 2002; Morris et al. 2003; Orell 2004; Wheeler and Priddel 2009). At less than 20 g adult weight, the increased abundance of C. concinnus is particularly interesting given that it is below the so called 'critical weight range' (35-5500 g) (Burbidge and McKenzie 1989; Kinnear et al. 2002, 2010; Johnson and Isaac 2009). The current distribution and abundance of some species is also likely to be an historical artefact of past and present declines and the managed recoveries of these species. For example, the current distribution of B. penicillata is a consequence of range contraction within the jarrah forest to all but one area east of Manjimup by the 1970s (Start et al. 1998) and subsequent successful translocations or reintroductions (Mawson 2004).

Ecosystem/year differences

There were considerable differences in terrestrial vertebrate species composition, abundance and community structure between the ecosystems/years. Half of the species (20/41) were recorded from only one or two ecosystems/years, and only 17% were common to all five ecosystems/years. Notably, JNW-C (2002-03) and JS (2001-02) had significantly higher abundances and, while there was no overall significant difference in species richness, JNW-C (2002-03) had significantly more species than JNE (2004-05). Community structure distinctions between ecosystem/years, particularly southern jarrah (JS/2001-02 and JB/2005-06) from the northern communities, appear to be strongly influenced by geographically related factors (e.g. climate, soils and nutrients; see McCaw et al. 2011). However, temporal influences also need to be considered. For instance, the abundance of B. penicillata has declined in all large populations in jarrah forest by more than 90%, including within the JS (2001-02) and JNE (2004-05) ecosystems, within the 5-y time frame of this study. The species was recently listed as Critically Endangered (Groom 2010).

Similar overall mean species richness, composition, abundance and community structure of terrestrial vertebrates in JNW-C (2002-03) and JNW-N (2003-04) support the classification of these two areas as one ecosystem, despite its comparatively large north--south extent. This was also the case for the understorey vascular flora communities of JNW-C, JNW-N and JNE (Ward et al. 2011) but were markedly distinct for other taxa including cryptograms (Cranfield et al. 2011), epigeous macrofungi (Robinson and Williams 2011), invertebrates (Farr et al. 2011) and birds (Abbott et al. 2011).

Design and data considerations

Detection error, which leads to biased estimates of abundance, is considered one of the two main sources of variation in monitoring data (Yoccoz et al. 2001). The assumption that probability of detection is constant over space and time is common to many field studies that use indices, such as raw count data, as measures of abundance. Although this assumption is often invalid (Pollock et al. 2002), there is some evidence that it may be valid for at least some taxa in this study. For example, Craig et al. (2009) demonstrated that count data from pitfall traps accurately reflected total reptile density as well as that of some individual species, including H. initialis and Lerista spp. (which were also common in the FORESTCHECK study), and provided relative measures of abundances that were not influenced by the vegetation structure of either restored mine pits or unmined jarrah forest. In the southern jarrah forest, capture probabilities for the mammals T. vulpecula and B. penicillata, based on mark-recapture modelling from a much larger dataset, were generally similar or constant over space and time across different silvicultural treatments (Wayne unpublished data). Nonetheless, future work associated with FORESTCHECK should test the veracity of this assumption.

Strengths of the FORESTCHECK study include standard sampling efforts, continuity of field expertise, and stratified and replicated trapping sites. All help to reduce detection error and spatial variation, the second major source of variation in monitoring data (Yoccoz et al. 2001), and increase confidence in subsequent inferences from analysis. Nonetheless, few treatment or other effects were detected in our study (e.g. time since last fire, live tree basal area or the proportion of basal area removed relative to pre-silvicultural treatment levels). This may reflect real lack of treatment effects on terrestrial vertebrates, but may also be due to limitations of our data, including insufficient records for many species, no records for others known to occur in jarrah forest, and the limited inferential power available with the existing data.

Within the treatments, 34-17% of all species were recorded only once or twice. The reason that eight of the nine species recorded in a single treatment were from reference forest is likely to be a function of chance and insufficient sampling rather than them actually being absent from harvested forest. In a complementary study in the southern jarrah forest (Wayne et al. 2001; A. Wayne unpublished data), six of these same eight species were recorded in harvested forest; but that study was outside the geographic range of the reptiles H. quadrilleata and L. burtonis. The sampling effort of Wayne et al. (2001) was 3.5 times more than in this study.

Low abundance data for many individual species also made it difficult to investigate species-level responses to silvicultural treatments and forest ecosystems. For example, only three species had more than 100 records across all grids (i.e. an average of more than two individuals per grid). At least 3 0 native species known to occur in the areas sampled were not detected in this study, although some of these (e.g. large macropods) are unlikely to be captured in wire cages or pitfalls. Evidence for incomplete species detection also includes the lack of an asymptote in species accumulation curves within treatments, and the four FORESTCHECK grids located within the Wayne et al. (2001) study area recorded less than half of the same species (using identical trapping methods).

As a result, statistical power was only sufficient to detect differences in species richness and abundance between treatments that were greater than 23% and 37%, respectively. Relative to other taxa the effort required to generate these data was large. However, FORESTCHECK is a long-term project and ongoing repeated sampling of grids will mean that data quality will improve and statistical power will increase over time.

Learning opportunities for this and other studies of fauna responses

Ecological monitoring programs for conservation or management are often considered inadequate (e.g. Yoccoz et al. 2001; Bennett and Adams 2004; Legg and Nagy 2006; Field et al. 2007). Common weaknesses include poorly defined objectives, insufficient hypothesis formulation, survey design, data quality and statistical power. Statistical power was a key limitation of terrestrial vertebrate data in this study, but it can be improved by changes that reduce residual variance and increase sample size (Legg and Nagy 2006).

Increasing the number of replicates within a well-balanced stratified design, with well-defined selection criteria to eliminate extraneous sources of variation, will increase statistical power and thus enable the use of simple and more powerful statistical models and parametric tests (e.g. Legg and Nagy 2006). The importance of design considerations including all major effects, and not just those of principal interest, is highlighted by the confounding of fox control within external reference subsets and between ecosystem/ years in this study. The consideration of JNW-C and JNW-N grids as replicates within the one JNW ecosystem category would also be appropriate to improve efficiency, given the similarities of these locations. Confounding between ecosystem and year could be overcome by surveying all or a subset of the treatments (e.g. external reference) within the same year. However, subsequent rounds of monitoring will also aid in clarifying ecosystem differences and improve statistical power, through aggregation of data from multiple years.

Increased sample effort at the grid level is also important. As well as improving abundance data, more frequent sampling within a given year will increase the likelihood of detecting more species. This is especially so for herpetofauna, that are active for relatively limited periods and or have peak activity periods that vary between species according to seasons (e.g. Main 1965) and particular meteorological factors (e.g. Lemckert 2001; Penman et al. 2006). The timing of surveys in relation to seasonal differences in abundance may also be important for some species, such as Phascogale sp. with an annual male die-off that can result in at least a 50% variation in population abundance within an annual cycle (Rhind 2002). The use of additional complementary survey methods would enhance the dataset, and records from spotlighting, sand pads, and sighting surveys from vehicles conducted as part of Forestcheck could also be incorporated into future analyses. Species-specific monitoring and research may also be necessary and particularly valuable for priority taxa that are more readily detected by more specific methods such as nest boxes for Phascogale sp. and scat surveys for the possum Pseudocheirus occidentalis (Wayne et al. 2005b).

Other analytical approaches may also help (e.g. improved control of covariates and Bayesian and information theoretic approaches). More reliable measures of abundance may be possible using mark-recapture methods (e.g. Pollock et al. 1990) or by developing better estimates of detection probability by other means, such as using the double sampling approach (Pollock et al. 2002).

The completion of the second cycle of monitoring will provide an increased dataset to investigate vertebrate associations and enable preliminary estimations of how repeated sampling over time is likely to improve statistical power (Field et al. 2007). The integrated nature of Forestcheck also allows more detailed investigations of relationships across major taxonomic groups and other forest attributes (such as soils, nutrients and coarse woody debris) that should provide a better understanding of the ecology of jarrah forest.

It is important that monitoring programs are associated with experimental work (Legg and Nagy 2006; Nichols and Williams 2006 and references within). Experimental research is especially valuable in verifying the causes and identifying the reasons for observed changes or differences (i.e. the processes behind responses). Forestcheck is directly complemented by experimental work investigating the effects of silviculture within the southern jarrah forest (Burrows et al. 1994, 2001; Morris et al. 2001; Abbott et al. 2003a,b; Craig and Roberts 2005). Future investigations should include a more direct integration and synthesis between these programs as well as developing similar research studies in other locations and or regarding other major factors in common with Forestcheck. However, tension remains between the value of making any of these improvements and the feasibility of being able to do so with available resources.

It is important to also emphasise that only the most common and widespread species furnish sufficient data for rigorous statistical analysis (Kavanagh and Stanton 2005). Rare, patchy and potentially sensitive or specialist species may be too difficult to directly measure their responses to disturbance. Alternative approaches are therefore necessary, such as the investigation of community-level responses (as in this study), the aggregation of taxa according to ecological guilds or by some other biologically meaningful criterion, the possible use of surrogates and indicator species, use of ecological first principles and other less direct or lateral means, such as comparative life history studies (Lauck 2005). Nonetheless, management necessarily continues within the context of incomplete and imperfect information. Therefore, while specific information may be necessary, complete knowledge is not essential to ensure sustainable forest management. Continued application of the precautionary principle (e.g. Deville and Harding 1997) and active adaptive management (Walters and Holling 1990; Lee 1999) can progressively improve knowledge and management.

Summary of responses

The greatest sources of variation in terrestrial vertebrate richness, abundance and community structure in this study were the beneficial effect of fox control on those species subject to fox predation, and differences attributable to ecosystem/year effects. In comparison, silvicultural treatment and the intensity of timber harvesting had minor effects. Some species were not recorded from particular treatments but this may simply reflect non-detection rather than true absence. Similarly, there was no relationship between live tree basal area after timber harvesting and either species richness or total abundance. Significant associations between terrestrial vertebrates and silvicultural treatment were nonetheless evident. The significantly different community structure between shelterwood grids and external reference forest was largely reflected in the former having higher mean species richness, two reptiles being more abundant, and greater overall vertebrate abundance. Within the limits of the data, no species were negatively associated with silvicultural treatment compared with external reference forest. Indeed, the fraction of species apparently unaffected by harvesting was much higher than reported by comparable studies in eastern Australia. Therefore it is likely that the current silvicultural treatments in the jarrah forest have been within the tolerance thresholds to disturbance for many of the terrestrial vertebrates as suggested for other taxa in these forests (e.g. Abbott et al. 2011).

Acknowledgements

We thank the animal handlers who collected data in the field (Bruce Ward, Colin Ward, Chris Vellios, Marika Maxwell and Christina Gilbert); Nature Conservation Program Leaders and employees at Donnelly, Wellington, Perth Hills and Blackwood Districts of the Department of Environment and Conservation for essential assistance and support of field activities (including Ian Wilson, Jennifer Langton, Drew Griffiths, Steve Gunn, Steve Thomas, Frank Collier and Kelly Bennett); Verna Tunsell for database management; Amanda Mellican for preliminary analyses of some of the data; and Richard Robinson, Lachie McCaw and Ian Abbott for comments on drafts of this manuscript. Thank you also to the referees who helped to improve this paper.

References

Abbott, I. and Burrows, N. (2004) Monitoring biodiversity in jarrah forest in south-west Western Australia: the FORESTCHECK initiative. In: Lunney, D. (ed.) Conservation of Australia's Forest Fauna. 2nd edn. Royal Zoological Society of New South Wales, Mosman, NSW, Australia, pp. 947-958.

Abbott, I. and Whitford, K. (2002) Conservation of vertebrate fauna using hollows in forests of south-west Western Australia: strategic risk assessment in relation to ecology, policy, planning, and operations management. Pacific Conservation Biology 7, 240-255.

Abbott, I., Burbidge, T., Strehlow, K., Mellican, A. and Wills, A. (2003a) Logging and burning impacts on cockroaches, crickets and grasshoppers, and spiders in jarrah forest, Western Australia. Forest Ecology and Management 174, 383-399.

Abbott, I., Mellican, A., Craig, M.D., Williams, M., Liddelow, G. and Wheeler, I. (2003b) Short-term logging and burning impacts on species richness, abundance and community structure of birds in open eucalypt forest in Western Australia. Wildlife Research 30, 321-329.

Abbott, I., Liddelow, G. L.,Vellios, C.V., Mellican, A.E. and Williams, M.R. (2011) FORESTCHECK : the response of birds to silviculture in jarrah (Eucalyptus marginata) forest. Australian Forestry 74, 328-335.

Adams, M.D., Law, B.S. and French, K.O. (2009) Vegetation structure influences the vertical stratification of open- and edge-space aerial-foraging bats in harvested forests. Forest Ecology and Management 258, 2090-2100.

Alexander, J.S.A., Scotts, D.J. and Loyn, R.H. (2002) Impacts of timber harvesting on mammals, reptiles and nocturnal birds in native hardwood forests of East Gippsland, Victoria: a retrospective approach. Australian Forestry 65, 182-210.

Anderson, M.J. (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecology 26, 32-46.

Anderson, M.J. and Robinson, J. (2003) Generalized discriminant analysis based on distances. Australian and New Zealand Journal of Statistics 45, 301-318.

Armstrong, R. (2004) Baiting operations: Western Shield review--February 2003. Conservation Science WA 5, 31-50.

Baker, S. and Lauck, B. (2006) Association of common brown froglets, Crinia signifera, with clearcut forest edges in Tasmania, Australia. Wildlife Research 33, 29-34.

Barry, S.J. (1984) Small mammals in a south-east Queensland rainforest: the effects of soil fertility and past logging disturbance. Australian Wildlife Research 11, 31-39.

Bennett, L.T. and Adams, M.A. (2004) Assessment of ecological effects due to forest harvesting: approaches and statistical issues. Journal of Applied Ecology 41, 585-598.

Burbidge, A.A. and McKenzie, N.L. (1989) Patterns in the modern decline of Western Australia's vertebrate fauna: causes and conservation implications. Biological Conservation 50, 143-198.

Burrows, N.D. and Christensen, P.E.S. (2002) Long-term trends in native mammal capture rates in a jarrah forest in south-western Australia. Australian Forestry 65, 211-219.

Burrows, N., Friend, G., Morris, K., Stoneman, G., Craig, M. and Williams, M. (1994) An Integrated Study of the Effects of Timber Harvesting on the Jarrah Forest Ecosystem. Department of Conservation and Land Management, Perth, Western Australia.

Burrows, N.D., Ward, B. and Cranfield, R. (2001) Short-term impacts of logging on understory vegetation in a jarrah forest. Australian Forestry 65, 47-58.

Bush, B., Maryan, B., Browne-Cooper, R. and Robinson, D. (1995) A Guide to the Reptiles and Frogs of the Perth Region. University of Western Australia Press.

Calver, M.C. and Dell, J. (1998a) Conservation status of mammals and birds in southwestern Australian forests. I. Is there evidence of direct links between forestry practices and species decline and extinction? Pacific Conservation Biology 4, 296-314.

Calver, M.C. and Dell, J. (1998b) Conservation status of mammals and birds in southwestern Australian forests. II. Are there unstudied, indirect or long-term links between forestry practices and species decline and extinction? Pacific Conservation Biology 4, 315-325.

Calver, M.C. and Wardell-Johnson, G. (2004) Sustained unsustainability? An evaluation of evidence for a history of overcutting in the jarrah forests of Western Australia and its consequences for fauna conservation. In: Lunney, D. (ed.) Conservation of Australia's Forest Fauna. Royal Zoological Society of New South Wales, Sydney, pp. 94-114.

Christensen, P.E.S. (1997) A Review of the Knowledge of the Effects of Key Disturbances on Fauna in the South-West Forest Region. Western Australian Government Department of Conservation of Land Management, Perth.

Christensen, P.E.S. and Abbott, I. (1989) Impact of fire in the eucalypt forest ecosystem of southern Western Australia: a critical review. Australian Forestry 52, 103-121.

Clarke, K.R. and Gorley, R.N. (2006) PRIMER v6: User Manual/Tutorial. PRIMER-E, Plymouth.

Cogger, H.G. (1994) Reptiles and Amphibians of Australia. Reed Books, Chatswood NSW.

Colwell, R.K. (2005) EstimateS: Statistical estimation of species richness and shared species from samples. Version 7.5. User's Guide. http:// purl.oclc.org/estimates.

Craig, M.D. and Roberts, J.D. (2005) The short-term impacts of logging on the jarrah forest avifauna in south-west Western Australia: implications for the design and analysis of logging experiments. Biological Conservation 124, 177-188.

Craig, M.D., Grigg, A.H., Garkaklis, M.J., Hobbs, R.J., Grant, C.D., Flemming, P.A. and Hardy, G.E.S. (2009) Does habitat structure influence capture probabilities? A study of reptiles in a eucalypt forest. Wildlife Research 36, 509-515.

Craig, M.D., Benkovic, A.M., Grigg, A.H., Hardy, G.E.S., Fleming, P.A. and Hobbs, R.J. (2011) How many mature microhabitats does a slow-recolonising reptile require? Implications for restoration of bauxite minesites in south-western Australia. Australian Journal of Zoology 59, 9-17.

Cranfield, R.J., Robinson, R.M., Williams, M.R. and Tunsell, V.L. (2011) Forestcheck : the response of lichens and bryophytes to silviculture in jarrah (Eucalyptus marginata) forest. Australian Forestry 74, 303-314.

deMaynadier, P.G. and Hunter Jr., M.L. (1995) The relationship between forest management and amphibian ecology: a review of the North American literature. Environmental Reviews 3, 230-261.

Deville, A. and Harding, R. (1997) Applying the Precautionary Principle. Federation Press, Sydney.

Ehmann, H. and Bamford, M.J. (1993) Family Typhlopidae. In: Glasby C.J., Ross, G.J.B. and Beesley, P.L. (eds) Fauna of Australia. Vol 2A. Amphibia and Reptilia. Australian Government Publishing Service, Canberra, pp. 280-283.

Farr, J.D., Wills, A.J., Van Heurck, P.F., Mellican, A.E. and Williams, M.R. (2011) Forestcheck : the response of macro-invertebrates to silviculture in jarrah (Eucalyptus marginata) forest. Australian Forestry 74, 315-327.

Faunt, K. (1992) Formation, Frequency and Longevity of Hollows in Jarrah. Western Australian Government Department of Conservation and Land Management, Perth.

Field, S.A., O'Connor, P.J., Tyre, A.J. and Possingham, H.P. (2007) Making monitoring meaningful. Austral Ecology 32, 485-491.

Friend, J.A. (1994) Recovery Plan for the Numbat (Myrmecobius fasciatus) 1995-2004. Western Australian Wildlife Management Program No. 18. Department of Conservation and Land Management, Western Australia.

Friend, G.R. and Wayne, A.F. (2003) Relationships between mammals and fire in south-west Western Australian ecosystems: what we know and what we need to know. In: Abbott, I. and Burrows, N.D. (eds) Fire in Ecosystems of South-West Western Australia: Impacts and Management. Backhuys Publishers, Leiden, The Netherlands, pp. 363-380.

Gibbons, P. and Lindenmayer, D.B. (2002) Tree Hollows and Wildlife Conservation in Australia. CSIRO Publishing, Melbourne.

Gillespie, G.R. (2002) Impacts of sediment loads, tadpole density, and food type on the growth and development of tadpoles of the spotted tree frog Litoria spenceri: an in-stream experiment. Biological Conservation 106, 141-150.

Goldingay, R.G., Daly, G. and Lemckert, F. (1996) Assessing the impacts of logging on reptiles and frogs in the montane forests of southern New South Wales. Wildlife Research 23, 495-510.

Groom, C. (2010) Justification for continued conservation efforts following the delisting of a threatened species: a case study of the woylie, Bettongia penicillata ogilbyi (Marsupialia: Potoroidae). Wildlife Research 37, 183-193.

Hazell, D. (2003) Frog ecology in modified Australian landscapes: a review. Wildlife Research 30, 193-205.

Heltsche, J.F. and Forrester, N.E. (1983) Estimating species richness using the jackknife procedure. Biometrics 39, 1-12.

Hosmer, D.W. and Lemeshow, S. (2000) Applied Logistic Regression. 2nd edn. Wiley, New York.

How, R.A. and Hillcox, S.J. (2000) Brushtail possum, Trichosurus vulpecula, populations in south-western Australia: demography, diet and conservation status. Wildlife Research 27, 81-89.

Inions, G.B., Tanton, M.T. and Davey, S.M. (1989) Effect of fire on the availability of hollows in trees used by the common brushtail possum, Trichosurus vulpecula Kerr, 1792, and the ringtail possum, Pseudocheirus peregrinus Boddaerts, 1785. Australian Wildlife Research 16, 449-458.

Johnson, C.N. and Isaac, J.L. (2009) Body mass and extinction risk in Australian marsupials: the 'critical weight range' revisited. Austral Ecology 34, 35-40.

Jones, B. (2004) The possum fauna of Western Australia: decline, persistence and status. In: Goldingay, R.L. and Jackson, S.M. (eds) The Biology of Australian Possums and Gliders. Surrey Beatty and Sons, Sydney, pp. 149-160.

Jones, B. and Hillcox, S. (1995) A survey of the possums Trichosurus vulpecula and Pseudocheirus occidentalis and their habitats in forest at Ludlow, Western Australia. Western Australian Naturalist 20, 139-150.

Jones, B.A., How, R.A. and Kitchener, D.J. (1994a) A field study of Pseudocheirus occidentalis (Marsupialia: Petauridae). I. Distribution and habitat. Wildlife Research 21, 175-187.

Jones, B.A., How, R.A. and Kitchener, D.J. (1994b) A field study of Pseudocheirus occidentalis (Marsupialia: Petauridae). II. Population studies. Wildlife Research 21, 189-201.

Kavanagh, R.P. and Bamkin, K.L. (1995) Distribution of nocturnal forest birds and mammals in relation to the logging mosaic in south-eastern New South Wales, Australia. Biological Conservation 71, 41-53.

Kavanagh, R.P. and Stanton, M.A. (2005) Vertebrate species assemblages and species sensitivity to logging in the forests of north-eastern New South Wales. Forest Ecology and Management 209, 309-341.

Kavanagh, R.P. and Webb, G.A. (1998) Effects of variable-intensity logging on mammals, reptiles and amphibians at Waratah Creek, southeastern New South Wales. Pacific Conservation Biology 4, 326-347.

Kinnear, J.E., Sumner, N.R. and Onus, M.L. (2002) The red fox in Australia--an exotic predator turned biocontrol agent. Biological Conservation 108, 335-359.

Kinnear, J.E., Krebs, C.J., Pentland, C., Orell, P., Holme, C. and Karvinen, R. (2010) Predator-baiting experiments for the conservation of rock-wallabies in Western Australia: a 25-year review with recent advances. Wildlife Research 37, 57-67.

Kruskal, J.B. (1964) Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika 29, 1-27.

Lamb, D. and Smyth, A. (2003) Forests. In: Attiwill, P. and Wilson, B. (eds) Ecology: An Australian Perspective. Oxford University Press, Melbourne, pp. 331-351.

Lauck, B. (2005). Impact of forestry practices on amphibian life histories. PhD thesis, University of Tasmania.

Lauck, B. (2006) Fluctuating asymmetry of the frog Crinia signifera in response to logging. Wildlife Research 33, 313-320.

Laurance, W.F. and Laurance, S.G.W. (1996) Responses of five arboreal marsupials to recent selective logging in tropical Australia. Biotropica 28, 310-322.

Lee, K.N. (1999) Appraising adaptive management. Conservation Ecology 3, 3.

Legg, C.J. and Nagy, L. (2006) Why most conservation monitoring is, but need not be, a waste of time. Journal of Environmental Management 78, 194-199.

Lemckert, F. (1999) Impacts of selective logging on frogs in a forested area of northern New South Wales. Biological Conservation 89, 321-328.

Lemckert, F.L. (2001) The influence of micrometeorological factors on the calling activity of the frog Crinia signifera (Anura: Myobatrachidae). Australian Zoologist 31, 625-631.

Lindenmayer, D.B. (1994) Wildlife corridors and the mitigation of logging impacts on fauna in wood-production forests in southeastern Australia--a review. Wildlife Research 21, 323-340.

Lindenmayer, D.B. (2000) Factors at multiple scales affecting distribution patterns and their implications for animal conservation--Leadbeater's possum as a case study. Biodiversity and Conservation 9, 15-35.

Lindenmayer, D.B. and Franklin, J.F. (2002) Conserving Forest Biodiversity: A Comprehensive Multiscaled Approach. Island Press, Washington.

Lindenmayer, D.B. and Hobbs, R.J. (2004) Fauna conservation in Australian plantation forests--a review. Biological Conservation 119, 151-168.

Lindenmayer, D.B., Cunningham, R.B. and McCarthy, M.A. (1999) The conservation of arboreal marsupials in the montane ash forests of the central highlands of Victoria, south-eastern Australia. VIII. Landscape analysis of the occurrence of arboreal marsupials. Biological Conservation 89, 83-92.

Lindenmayer, D.B., Burton, P.J. and Franklin, J.F. (2008) Salvage Logging and its Ecological Consequences. Island Press, Washington DC.

Lloyd, A., Law, B. and Goldingay, R. (2006) Bat activity on riparian zones and upper slopes in Australian timber production forests and the effectiveness of riparian buffers. Biological Conservation 129, 207-220.

Lunney, D., Eby, P. and O'Connell, M. (1991) Effects of logging, fire and drought on three species of lizards in Mumbulla State Forest on the south coast of New South Wales. Australian Journal of Ecology 16, 33-46.

Lunney, D., Matthews, A., Eby, P. and Penn, A. (2009) The long-term effects of logging for woodchips on small mammal populations. Wildlife Research 36, 691-701.

Main, A.R. (1965) Frogs of Southern Western Australia. Western Australian Naturalist Club, Perth.

Mawson, P.R. (2004) Translocations and fauna reconstruction sites: Western Shield review--February 2003. Conservation Science WA 5, 108-121.

McCaw, W.L. (2011) Characteristics of jarrah (Eucalyptus marginata) forest at Forestcheck monitoring sites in south-west Western Australia: stand structure, litter, woody debris, soil and foliar nutrients. Australian Forestry 74, 254-265.

McCaw, W.L., Robinson, R.M. and Williams, M.R. (2011) Integrated biodiversity monitoring for the jarrah (Eucalyptus marginata) forest in south-west Western Australia: the Forestcheck project. Australian Forestry 74, 240-253.

Morris, K., Johnson, B., Rooney, J. and Ward, C. (2001) The short-term impacts of timber harvesting and associated activities on the abundance of medium-sized mammals in the jarrah forest of Western Australia. In: Craig J.L., Mitchell, N. and Saunders, D.A. (eds) Nature Conservation 5: Conservation in the Production Environments: Managing the Matrix. Surrey Beatty and Sons, Sydney, pp. 60-70.

Morris, K., Johnson, B., Orell, P., Gaikhorst, G., Wayne, A. and Moro, D. (2003) Recovery of the threatened chuditch (Dasyurus geoffroii): a case study. In: Archer M., Jones, M.E. and Dickman, C.R. (eds) Predators with Pouches: The Biology of Carnivorous Marsupials. CSIRO Publishing, Melbourne, pp. 435-451.

National Land and Water Resources Audit (2002) Australian Terrestrial Biodiversity Assessment 2002. National Land and Water Resources Audit, Canberra.

Nichols, J.D. and Williams, B.K. (2006) Monitoring for conservation. Trends in Ecology and Evolution 21, 668-673.

Nichols, O.G. and Muir, B. (1989) Vertebrates of the jarrah forest. In: Dell B., Havel, J.J. and Malajczuk, N. (eds) The Jarrah Forest: A Complex Mediterranean System. Kluwer Academic Publishers, Dordrecht, pp. 133-153.

Orell, P. (2004) Fauna monitoring and staff training: Western Shield review--February 2003. Conservation Science Western Australia 5, 51-95.

Osenberg, C.W., Schmitt, R.J., Holbrook, S.J., Abu-Saba, K.E. and Flegal, A.R. (1994) Detection of environmental impacts: natural variability, effect size, and power analysis. Ecological Applications 4, 16-30.

Penman, T.D., Mahony, M.J. and Lemckert, F.L. (2005) Soil disturbance in integrated logging operations and the potential impacts on a fossorial Australian frog. AppliedHerpetology 2, 415-424.

Penman, T.D., Lemckert, F.L. and Mahony, M.J. (2006) Meteorological effects on the activity of the giant burrowing frog (Heleioporus australiacus) in south-eastern Australia. Wildlife Research 33, 35-40.

Penman, T.D., Lemckert, F.L. and Mahony, M.J. (2008) Applied conservation management of a threatened forest dependent frog, Heleioporus australiacus. Endangered Species Research 5, 45-53.

Pollock, K.H., Nichols, J.D., Brownie, C. and Hines, J.E. (1990) Statistical inference for capture-recapture experiments. Wildlife Monographs 107, 1-97.

Pollock, K.H., Nichols, J.D., Simons, T.R., Farnsworth, G.L., Bailey, L.L. and Sauer, J.R. (2002) Large scale wildlife monitoring studies: statistical methods for design and analysis. Environmentrics 13, 105-119.

Recher, H.F. (1985) A diminishing resource: mature forest and its role in forest management. In: Kikkawa, J. (ed.) Wildlife Management in the Forests and Forest-Controlled Lands in the Tropics and Southern Hemisphere. IUFRO and University of Queensland, Brisbane, pp. 28-33.

Rhind, S.G. (1996a) Habitat tree requirements and the effects of removal during logging on the marsupial brush-tailed phascogale (Phascogale tapoatafa tapoatafa) in Western Australia. Western Australian Naturalist 21, 1-22.

Rhind, S.G. (1996b) Addendum--Errata: habitat tree requirements and the effects of removal during logging on the marsupial brush-tailed phascogale (Phascogale tapoatafa tapoatafa) in Western Australia. Western Australian Naturalist 21, 1.

Rhind, S.G. (1998) Ecology of the brush-tailed phascogale in jarrah forest of south-western Australia. Thesis, Murdoch University, Perth, Western Australia.

Rhind, S.G. (2002) Reproductive demographics among brush-tailed phascogales (Phascogale tapoatafa) in south-western Australia. Wildlife Research 29, 247-257.

Rhind, S.G. (2004) Direct impacts of timber harvesting and forest management of the brush-tailed phascogale Phascogale tapotafa and other arboreal marsupials in a jarrah forest of Western Australia. In: Lunney, D. (ed.) Conservation of Australia's Forest Fauna. Royal Zoological Society of New South Wales, Sydney, pp. 639-655.

Rhind, S.G. and Bradley, J.S. (2002) The effect of drought on body size, growth and abundance of wild brush-tailed phascogales (Phascogale tapoatafa) in south-western Australia. Wildlife Research 29, 235-245.

Robinson, R.M. and Williams, M.R. (2011) Forestcheck : the response of epigeous macrofungi to silviculture in jarrah (Eucalyptus marginata) forest. Australian Forestry 74, 288-302.

SAS Institute Inc. (2004) SAS/STAT 9.1 User's guide, Volumes 1-7. SAS Publishing, Cary, NC.

Serena, M., Soderquist, T.R. and Morris, K. (1991) The chuditch: Dasyurus geoffroii. Department of Conservation and Land Management, Western Australia.

Start, T., Burbidge, A. and Armstrong, D. (1998) A review of the conservation status of the woylie, Bettongia penicillata ogilbyi (Marsupialia: Potoroidae) using IUCN criteria. CALMScience 2, 277-289.

Underwood, A.J., Russell, J.S. and Craig, W.O. (1996) On beyond BACI: sampling designs that might reliably detect environmental disturbances. In: Schmitt, R.J. and Osenberg, C.W. (eds) Detecting Ecological Impacts. Academic Press, San Diego, pp. 151-175.

Walters, C.J. and Holling, C.S. (1990) Large-scale management experiments and learning by doing. Ecology 71, 2060-2068.

Ward, B., Robinson, R.M., Cranfield, R.J. and Williams, M.R. (2011) Forestcheck : the response of vascular flora to silviculture in jarrah (Eucalyptus marginata) forest. Australian Forestry 74, 276-287.

Wardell-Johnson, G. (1986) Use of nest boxes by mardos, Antechinus flavipes leucogaster, in regenerating karri forest in south western Australia. Australian Wildlife Research 13, 407-417.

Wardell-Johnson, G. and Horwitz, P. (1996) Conserving biodiversity and the recognition of heterogeneity in ancient landscapes: a case study from south-western Australia. Forest Ecology and Management 85, 219-238.

Wardell-Johnson, G. and Nichols, O. (1991) Forest wildlife and habitat management in southwestern Australia: knowledge, research and direction. In: Lunney, D. (ed.) Conservation of Australia's Forest Fauna. Royal Zoological Society of New South Wales, Sydney, pp. 161-192.

Wardell-Johnson, G., Calver, M.C., Saunders, D.A., Conroy, S. and Jones, B. (2004) Why integration of demographic and site-based studies of disturbance is essential for the conservation of jarrah forest fauna. In: Lunney, D. (ed.) Conservation of Australia's Forest Fauna. Royal Zoological Society of New South Wales, Sydney, pp. 394-417.

Wayne, A.F., Ward, C., Rooney, J. and Wheeler, I. (2000) The immediate impacts of timber harvesting and associated activities on the ngwayir (Pseudocheirus occidentalis) in the jarrah forest of Kingston State Forest Block. Unpublished progress report. Western Australian Government Department of Conservation and Land Management, Manjimup.

Wayne, A.F., Wheeler, I.B., Ward, C.G., Rooney, J.F. and Mellican, A. (2001) The impacts of timber harvesting and associated activities on the small terrestrial vertebrates of the jarrah forest: Kingston Project progress report. Western Australian Government Department of Conservation and Land Management, Manjimup.

Wayne, A.F., Cowling, A., Rooney, J.F., Ward, C.G., Wheeler, I.B., Lindenmayer, D.B. and Donnelly, C.F. (2005a) Factors affecting the detection of possums by spotlighting in Western Australia. Wildlife Research 32, 689-700.

Wayne, A.F., Cowling, A., Ward, C.G., Rooney, J.F., Vellios, C.V., Lindenmayer, D.B. and Donnelly, C.F. (2005b). A comparison of survey methods for arboreal possums in jarrah forest, Western Australia. Wildlife Research 32, 701-714.

Wayne, A.F., Cowling, A., Lindenmayer, D.B., Ward, C.G., Vellios, C.V., Donnelly, C.F. and Calver, M.C. (2006) The abundance of a threatened arboreal marsupial in relation to anthropogenic disturbances at local and landscape scales in mediterranean-type forests in south-western Australia. Biological Conservation 127, 463-476.

Webala, P., Craig, M.D., Law, B.S., Wayne, A.F. and Bradley, S.J. (2010) Roost site selection by southern forest bat Vespadelus regulus and Gould's long-eared bat Nyctophilus gouldi in logged jarrah forests, south-western Australia Forest Ecology and Management 260, 1780-1790.

Webala, P., Craig, M.D., Law, B.S., Armstrong, K.N., Wayne, A.F. and Bradley, J.S. (2011). Bat habitat use in logged jarrah eucalypt forests of south-western Australia. Journal of Applied Ecology 48, 398-406.

Webb, G.A. (1995) Effects of logging on lizards in eucalypt forest at Eden, New South Wales. Australian Forestry 58, 155-159.

Wheeler, R. and Priddel, D. (2009) The impact of introduced predators on two threatened prey species: a case study from western New South Wales. Ecological Management and Restoration 10, S117-123.

Whitford, K.R. (2001) Dimensions of tree hollows used by fauna in the jarrah forest: improving the dimensional description of potentially usable hollows. CALMScience 3, 499-511.

Whitford, K.R. (2002) Hollows in jarrah (Eucalyptus marginata) and marri (Corymbia calophylla) trees: I. Hollow sizes, tree attributes and ages. Forest Ecology and Management 160, 201-214.

Whitford, K.R. and Williams, M.R. (2000) Survival of jarrah (Eucalyptus marginata Sm.) and marri (Corymbia calophylla Lindl.) habitat trees retained after logging. Forest Ecology and Management 5218, 1-17.

Whitford, K.R. and Williams, M.R. (2002) Hollows in jarrah (Eucalyptus marginata) and marri (Corymbia calophylla) trees--II. Selecting trees to retain for hollow dependent fauna. Forest Ecology and Management 160, 215-232.

Williams, M.R. and Faunt, K. (1997) Factors affecting the abundance of hollows in logs in jarrah forest of south-western Australia. Forest Ecology and Management 95, 153-160.

Woinarski, J.C.Z. and Fisher, A. (1995) Wildlife of lancewood (Acacia shirleyi) thickets and woodlands in Northern Australia. 1. Variation in vertebrate species composition across the environmental range occupied by lancewood vegetation in the Northern Territory. Wildlife Research 22, 379-411.

Wyre, G. (2004) Management of the Western Shield program: Western Shield review--February 2003. Conservation Science Western Australia 5, 20-30.

Yoccoz, N.G., Nichols, J.D. and Boulinier, T. (2001) Monitoring biological diversity in space and time. Trends in Ecology and Evolution 16, 446-453.

Zuur, A.F., Leno, E.N. and Elphick, C.S. (2010) A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution 1, 3-14.

Adrian F. Wayne (1,3), Graeme L. Liddelow (1) and Matthew R. Williams (2)

(1) Department of Environment and Conservation, Science Division, Locked Bag 2, Manjimup, WA 6258, Australia

(2) Department of Environment and Conservation, Science Division, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia

(3) Email: adrian.wayne@dec.wa.gov.au

Revised manuscript received 27 September 2011
Table 1. Summary of Forestcheck monitoring grids according to jarrah
forest ecosystem/year of sampling, silvicultural treatment, and fox
control (number of grids subject/not subject to routine fox/baiting
out of the total number of grids for that particular ecosystem/year
x silvicultural treatment)

                                   Year sampled and ecosystem

Treatment                        2001-02    2002-03    2003-04
                                    JS       JNW-C      JNW-N

Gap release                        1/2        0/3        2/0
Shelterwood                        1/0        0/3        3/0
Selective cut
Coupe buffer                       1/2
External reference--harvested
  > 40 y ago                       1/1                   2/0
External reference--never-
  harvested                        0/1        0/3        1/0
Total                              4/6        0/9        8/0

                                  Year sampled and
                                     ecosystem

Treatment                        2004-05    2005-06     Total
                                   JNE         JB

Gap release                        1/2        3/0        7/7
Shelterwood                        1/2        2/0        7/5
Selective cut                                 3/0        3/0
Coupe buffer                       1/0                   2/2
External reference--harvested
  > 40 y ago                       0/1        2/0        5/2
External reference--never-
  harvested                        0/2        1/0        2/6
Total                              3/7       11/0       26/22

Table 2. Count of all terrestrial vertebrate taxa recorded in all
treatments from 48 Forestcheck grids

                                    Gap                    Selective
                                  release    Shelterwood      cut
Species                           (n = 14)    (n = 12)      (n = 3)

Frogs
  Crinia georgiana                   7            2
  Crinia glauerti
  Crinia subinsignifera
  Heleioporus eyrei                 13            6            1
  Limnodynastes dorsalis             5            1
  Litoria adelaidensis               2
  Litoria moorei
  Pseudophryne guentheri
Reptiles
  Acritoscincus trilineatum          5            2
  Apraisia pulchella                 7            6            2
  Christinus marmoratus              6            1            1
  Ctenotus catenifer                 1            1
  Ctenotus labillardieri             6           21            2
  Diplodactylus polyophthalmus       1            2
  Egernia napoleonis                 9           22            9
  Glaphyromorphus gracilipes         1
  Hemiergis initialis                1            9
  Hemiergis peroni                  15            8            2
  Hemiergis quadrillineata
  Lerista spp.                      49           71            3
  Lialis burtonis
  Menetia greyii                    23           34
  Morethia spp.                     49           38            4
  Notechis scutatus                               1
  Parasuta gouldii                   1
  Pogona minor                       2            3
  Ramphotyphlops australis           7           10            4
  Rhinoplocephalus bicolor           2                         1
  Tiliqua rugosa                     2            6
  Varanus rosenbergi                 1            1
Mammals
  Antechinus flavipes                6            4
  Bettongia penicillata             32           31
  Cercartetus concinnus              7           04            7
  Dasyurus geoffroii                 6            8
  Isoodon obesulus
  Mus musculus                       3            1
  Rattus fuscipes
  Rattus rattus                      7           13
  Sminthopsis spp.                   7           12            1
  Tachyglossus aculeatus             2            1
  Trichosurus vulpecula             46           56           24

                                   Coupe    External
                                  buffer    reference
Species                           (n = 4)   (n = 15)

Frogs
  Crinia georgiana                   7          4
  Crinia glauerti                    1
  Crinia subinsignifera                         1
  Heleioporus eyrei                  3
  Limnodynastes dorsalis             3          3
  Litoria adelaidensis                          1
  Litoria moorei                                1
  Pseudophryne guentheri                        2
Reptiles
  Acritoscincus trilineatum          2          4
  Apraisia pulchella                           12
  Christinus marmoratus              1          3
  Ctenotus catenifer
  Ctenotus labillardieri                        9
  Diplodactylus polyophthalmus                  2
  Egernia napoleonis                 1          6
  Glaphyromorphus gracilipes                    1
  Hemiergis initialis                           6
  Hemiergis peroni                   8         17
  Hemiergis quadrillineata                      1
  Lerista spp.                       3         41
  Lialis burtonis                               1
  Menetia greyii                     2         17
  Morethia spp.                      1         35
  Notechis scutatus
  Parasuta gouldii                              1
  Pogona minor                       1          1
  Ramphotyphlops australis                      2
  Rhinoplocephalus bicolor
  Tiliqua rugosa                                4
  Varanus rosenbergi                            1
Mammals
  Antechinus flavipes                           5
  Bettongia penicillata              9         22
  Cercartetus concinnus              1          4
  Dasyurus geoffroii                 4          2
  Isoodon obesulus                              1
  Mus musculus                       4          1
  Rattus fuscipes                               1
  Rattus rattus
  Sminthopsis spp.                   3          6
  Tachyglossus aculeatus
  Trichosurus vulpecula             46         80

Table 3. Mean species richness and abundance ([+ or -] se)
of terrestrial
vertebrates for the five forest ecosystems/years of sampling
(n = 41). N = number of grids monitored; values in the same column
followed by the same letter are not significantly different
([alpha] = 0.05).

Ecosystem/
year           N  Richness                 Abundance

JS/2001-02     7  8.42 [+ or -] 0.87 (ab)  21.69 [+ or -] 3.83 (ab)
JNW-C/2002-03  9  9.60 [+ or -] 0.91 (a)   34.93 [+ or -] 6.48 (a)
JNW-N/2003-04  8  7.81 [+ or -] 1.16 (ab)  11.44 [+ or -] 2.70 (c)
JNE/2004-05    9  6.70 [+ or -] 0.82 (b)   19.99 [+ or -] 3.29 (bc)
JB/2005-06     8  7.93 [+ or -] 1.01 (ab)  12.25 [+ or -] 2.50 (c)

Table 4. Mean species richness and abundance ([+ or -] se) of
terrestrial vertebrates for the two main silvicultural treatments
and external reference forest (n = 41). N = number of grids
monitored; values in the same column followed by the same letter
are not significantly different ([alpha] = 0.05).

Treatment     N           Richness                  Abundance

External
  reference   15   6.34 [+ or -] 0.87 (a)    17.61 [+ or -] 3.10 (a)
Shelterwood   12   9.63 [+ or -] 0.73 (b)    21.64 [+ or -] 3.21 (a)
Gap release   14   8.31 [+ or -] 0.72 (ab)   16.36 [+ or -] 2.38 (a)

Table 5. Estimates of the number of study grids required to get
detectable effect sizes (Osenberg et al. 1994) ranging from 50% to
5% (i.e., an increase or decrease with respect to the external
reference treatment), with 95% confidence, based on the current
Forestcheck terrestrial vertebrate monitoring design and data
(n = 41 grids, 3 treatments).

Attribute                   Detectable effect size (%)

                      50     40     30     20     10     5

Species richness      12     15     24    051    200    0800
Abundance             21     33     69    168    750    3200
Gale Copyright:
Copyright 2011 Gale, Cengage Learning. All rights reserved.