Lizard responses to forest fire and timber harvesting: Complementary insights from species and community approaches

Forest Ecology and Management 379 (2016) 206–215

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Lizard responses to forest fire and timber harvesting: Complementary insights from species and community approaches Yang Hu a,⇑, Luke T. Kelly a, Graeme R. Gillespie a,c, Tim S. Jessop a,b a

School of BioSciences, University of Melbourne, Victoria 3010, Australia Centre for Integrative Ecology, School of Life and Environmental Sciences, Deakin University, Victoria 3216, Australia c Department of Land Resource Management, Palmerston, Northern Territory 0831, Australia b

a r t i c l e

i n f o

Article history: Received 7 January 2016 Received in revised form 24 July 2016 Accepted 27 July 2016

Keywords: Reptile assemblage Disturbance Burning Logging Functional diversity

a b s t r a c t Understanding the relationship between community composition and ecosystem function is essential for managing forests with complex disturbance regimes. Studies of animal responses to fire and timber harvesting in forest ecosystems typically focus on a single level of community diversity. Measures of species abundance and diversity at the community level, along with measures of functional diversity that incorporate information on species traits, provide opportunities for complementary insights into biodiversity responses to disturbances. We quantified community and functional responses of a temperate forest lizard community to fire and rotational logging using metrics including species-specific abundance, community abundance, species richness and evenness, as well as trait-based measures of functional diversity. We used non-linear regression models to examine the relationships between reptile data and time since fire and timber harvesting, using sites arrayed along a 30-years post-disturbance chronosequence. We modelled responses separately in two major vegetation types: coastal Banksia woodland and lowland eucalypt forests. Species and community measures offered different insights into the role of fire and logging. Species responses to disturbance differed between disturbance type and vegetation type. Four species exhibited significant population responses to either fire or timber harvesting, while the rest were unaffected by either disturbance. At the community level, species richness and community abundance increased significantly with time since fire in woodland vegetation. In forest vegetation, community abundance decreased with time since fire. Surprisingly, community evenness and functional diversity did not show marked responses to fire or timber harvesting. This is likely a result of trait homogeneity and the asynchrony in species responses to disturbance. We advocate using multiple measures of community composition - incorporating species-specific information, community metrics and functional traits - to ensure a more holistic understanding of disturbance ecology in forest landscapes. Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction Disturbance influences community dynamics by changing resource availability, patterns of species assemblage and the functional relationship between species and resources (Hughes et al., 2007; Turner, 2010). Studies of animal responses to disturbance largely focus on a single aspect of biodiversity such as species richness or community abundance. Although such studies have yielded a wealth of information, focusing on just one biological level can mask important community changes and functional differences (Cadotte et al., 2011; Gerisch et al., 2012). An approach that combines measures of individual species, community assemblage and functional diversity is likely to provide complementary insights ⇑ Corresponding author. E-mail address: [email protected] (Y. Hu). http://dx.doi.org/10.1016/j.foreco.2016.07.040 0378-1127/Ó 2016 Elsevier B.V. All rights reserved.

into the underlying mechanisms regulating community organization in relation to disturbances (McGill et al., 2007; Driscoll et al., 2010; Cadotte et al., 2011). Disturbances such as fire and timber harvesting often occur in the same landscapes and can act synergistically (Reich et al., 2001; Gardner et al., 2007). Both disturbances can cause largescale changes to habitat structure and community assemblages. Some of these changes, such as the removal of vegetation and habitat elements such as coarse-woody debris, can be similar under both types of disturbances (Keller et al., 2004; Bassett et al., 2015). Other ecological changes are markedly different between disturbance types. For example, low intensity fires often affected the understorey and shrub layer more so than the overstorey (Hart and Chen, 2008), while timber harvesting has more severe impacts across all vegetation strata (e.g., Alexander et al., 2002; Asner et al., 2005). Structural variations in natural landscapes are

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

intricately linked to disturbance regimes, including the severity, frequency and time since disturbance (Turner, 2010; Bassett et al., 2015; Chia et al., 2016). These habitat differences drive different species and community responses in various ways, such as by influencing resource availability and successional dynamics (Jellinek et al., 2004; Tews et al., 2004; Aponte et al., 2014). Understanding how fire and timber harvesting influence the numerical and functional diversity of animal communities continues to be important for forest ecology and management. Here we explore the responses of a lizard community to disturbances in temperate forests of south-eastern Australia that are frequently perturbed by fire and timber harvesting. Lizards are common elements of temperate forests, and are ectothermic and sedentary, which makes them a suitable group for developing a better understanding of animal responses to disturbance (Gardner et al., 2007). Previous works show that reptile community metrics, such as richness and abundance, exhibit different patterns in response to fire and timber harvesting (e.g., Kilpatrick et al., 2010; Santos and Cheylan, 2013; Sutton et al., 2014; Azor et al., 2015). In a sand-pine scrub community of North America, reptile community composition showed rapid changes in responses to prescribed burning (Steen et al., 2013). Increases in both richness and abundance were observed in a reptile assemblage in southeastern Australia in response to fire, but not to such a degree for timber harvesting (Hu et al., 2013a,b). Fine-scale factors such as site-specific habitat structure, climatic conditions and the local reptile assemblage are also important influences on lizard communities (Nimmo et al., 2013; Steen et al., 2013; Sutton et al., 2014). How might individual reptile species respond to fire and timber harvesting in temperate forests? The habitat accommodation model suggests a predictable sequence of animal succession as the structural elements of habitat recover to suit each of the component species respectively (Fox, 1982). However, responses of individual reptile species to fire have been shown to be variable, with some correlation with the intrinsic attributes of species such as ecological tolerance and habitat preferences (e.g., Driscoll and Henderson, 2008; Nimmo et al., 2012; Smith et al., 2013). A large scale study in North American temperate forests reported species-specific associations to both disturbance history and environmental variables (Sutton et al., 2013). Post-disturbance changes in species richness or evenness, or the abundance of individual species are expected to potentially influence the ecological function of communities (e.g., Chapin et al., 2000). This is because disturbance effects on communities, via changes to species diversity, composition and abundance, affect the range, distribution and abundance of biological and ecological traits of individual species (e.g., life history traits such as body size, diet and habitat use) (Petchey and Gaston, 2002, 2006; Moretti et al., 2006). Consequently, measures of functional diversity (FD) have been suggested as a complementary approach to further evaluate how disturbances influence communities (Mason et al., 2005). In this approach, species are treated as a distribution of quantifiable trait values that link communities to their potential ecological function (Petchey and Gaston, 2006). When used in conjunction with species diversity metrics, this combined approach potentially enhances our understanding of community organization in the face of disturbance events (Chapin et al., 2000; McGill et al., 2007; Cadotte et al., 2011). We model the effects of fire and timber harvesting on individual lizard species abundances, community assemblage and functional diversity. First, we examined species-specific patterns in response to time since fire and timber harvesting across two broad vegetation types. Second, we explored patterns in how species richness, total abundance of lizards (hereafter ‘community abundance’) and community evenness responded to disturbance history. Third,

207

we used information on species life history traits (including body size, diet and habitat niche) to calculate three functional diversity metrics and examined if they varied according to disturbance type and time since disturbance. 2. Materials and methods 2.1. Study area The study was conducted in Cape Conran Coastal Park (37°490 S, 148°440 E), Murrungowar State Forest (37°370 S, 148°440 E), Bemm State Forest (37°410 S, 148°520 E) and Club Terrace State Forest (37°640 S, 148°490 E), in East Gippsland, Victoria, Australia. Annual rainfall for the region averaged approximately 846 mm, and mean maximum and minimum temperatures ranged from 27.0 °C (January) to 4.7 °C (July). Climatic conditions are relatively uniform across the study area. Elevation ranged from sea level to 350 m. Our study focused on two widespread vegetation types: (1) lowland forest dominated by Eucalyptus sieberi and Eucalyptus globoidea, and (2) coastal woodland dominated by Banksia species, most notably Banksia serrata and Banksia integrifolia. Study sites comprise a mosaic of vegetation which had been disturbed at different times by fire and timber harvesting. The fires range from intense wildfires to relatively mild prescribed burns conducted by government agencies, both of which are relatively frequent in the recent decades. We used fire history mapping from 1930 to the present (Department of Environment, Land and Water, Orbost, unpublished data). Since the early 1900’s, several selective and nominally reduced-impact timber harvesting methods have been employed in the study area, such as single tree selections, seed tree retention and thinning from below (DSE, 2007, 2009). Clear-fell logging has not been undertaken at our study sites. Although logging techniques differ in their intensities and area of impact, a previous study showed that moderate differences between timber harvesting techniques in the study area did not differentially influence reptile community structure (Hu et al., 2013a). In the present study we group logged areas as a single disturbance type. 2.2. Study design We selected 97 sites for study. Sixty-three sites had been disturbed by fire in the last 30 years and thirty-four sites had been disturbed by timber harvesting in the last 30 years (Appendix 3). We selected sites that had been disturbed by only a single disturbance type within the last 30 years to avoid potential confounding interactions between fire and timber harvesting. The 63 sites disturbed by fires were located in both of the main vegetation types: lowland eucalypt forests (28 sites) and coastal Banksia woodland (35 sites). All timber harvesting sites were located in eucalypt forests, and no harvesting had been conducted in Banksia vegetation. Sites were spaced at least 250 m apart to ensure independence of sampling relative to the movement of small lizards. 2.3. Sampling protocol We used pitfall trapping to sample small lizards. A pitfall line was constructed at each site, consisting of five 20-L buckets spaced 5 m apart, and buried with the rim flush with the ground. A 30 cm plastic drift fence passed over the top of the buckets. Traps were kept at least 20 m from tracks to minimise the influence of edge effects. Trapping was conducted from late 2008 to early 2011, between the warmer months of November and February, for a total of 17,531 trap nights (each pitfall bucket is considered a single trap). Traps were checked daily and closed when not in use.

208

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

Animals were identified to species level following descriptions in Wilson and Swan (2008). Each animal capture was treated as a new individual. Recapture rates of reptiles in eucalypt and Banksia vegetation of Australia are usually very low (e.g., 10% in Bragg et al., 2005; 2% in Nicholson et al., 2006; 1% in Legge et al., 2008; <5% in Carthew et al., 2009; <20% in Smith et al., 2013). Moreover, each lizard captured was released at least 20 m away from the pitfall line, exceeding the short-term movements of most small skinks, thus reducing the likelihood of recapture. Therefore, we expect our analysis to be robust to the small number of recaptures potentially present in our sample. Other reptiles present on this landscape (mostly snakes) were not used for analyses as pitfall trapping is not the most effective sampling method for those species. 2.4. Species traits For each species, we collated data on biological traits representing differences in functional ecological output. Trait information was collated from the literature (Wilson and Swan, 2008; Cogger, 2014) and an online database (museumvictoria.com.au; Table 1): (1) (2) (3) (4) (5) (6)

Snout-to-vent length (SVL; mm) Primary habitat strata (terrestrial, fossorial or arboreal) Main shelter use (leaf litter, coarse woody debris or crevices) Reproductive mode (oviparous or viviparous) Mean clutch size Activity pattern (diurnal, nocturnal or crepuscular)

These traits were selected because they represent important biological and ecological characteristics for lizards. Previous work showed that body size is related to reptile dispersal ability and thermal tolerance; reproductive mode and clutch size are indicators of reproductive potential; habitat and shelter use and activity pattern represent species’ habitat preferences and tolerance to microclimate (Howard et al., 2003; Moretti et al., 2006; OlallaTárraga et al., 2006). 2.5. Data analysis We examined the relationships between disturbance history and three major components of reptile diversity, including (1) individual species responses, (2) total species richness, evenness and community abundance, and (3) functional diversity. We used generalized additive models (GAM) with a Gaussian distribution to model the responses of the various metrics above in response to fire and timber harvesting. GAM is a non-parametric approach that does not require adherence to assumptions of linearity. Several studies show that GAMs are a particularly useful method for quantifying relationships between animal abundance and disturbance history, which are often complex and non-linear (Kelly et al., 2011; Nimmo et al., 2012; Hu et al., 2013a). All analyses were

conducted separately for each vegetation type (lowland forest vs. coastal Banksia woodland). First, we modelled how the relative abundance of individual species (per 100 trap nights) responded to time since fire and timber harvesting. This was conducted for seven species each with >50 captures. Other species were captured in insufficient numbers (<15 captures) for analysis (Appendix 1; see Appendix 2 for the list of all response variables tested in relation to disturbance history and habitat type). Second, we quantified three measures of reptile community assemblage: species richness, total relative abundance (per 100 trap nights), and species evenness, and correlated their responses to time since fire and time since timber harvesting. Third, to investigate functional diversity changes postdisturbance, we calculated three functional diversity metrics: (1) functional richness, (2) functional evenness, and (3) functional divergence. These metrics are the three primary components of functional diversity (Mason et al., 2005). Functional richness indicates the degree to which the resources available to a community are used. Functional evenness is viewed as the distribution of organisms in niche space, with lower functional evenness indicating an under-utilisation of resources (Mason et al., 2005). Functional divergence reflects niche differentiation, and thus the degree of resource competition, with high functional divergence indicating greater ecological functionality, as a result of more efficient resource use (Mason et al., 2005). Functional evenness and divergence were calculated from the distribution of species across all sites, in multivariate functional trait space, and were weighted by the relative abundance of each species (Villeger et al., 2008; Bishop, 2012; Gerisch et al., 2012). Functional richness was not weighted by species abundances (Mason et al., 2005; Villeger et al., 2008; Gerisch et al., 2012). Functional evenness and divergence values were bounded between 0 and 1, with lower values representing lower evenness and divergence. To explore the possibility of dominant species masking functional responses we also recalculated all functional diversity metrics after excluding the single most common species from the dataset and ran additional regression analyses on these data. GAMs were used to determine the relationship between functional indices, disturbance type and time since disturbance. We also tested for correlations between these three functional metrics by regressing them against each other using generalized linear models (GLM), both before and after removing the dominant species. Inferring patterns in ecological research is often confounded by the effect of spatial autocorrelation, in which site level attributes may vary according to geographic proximity (Dormann et al., 2007). Generalized additive mixed models (GAMM) can incorporate different correlation structures as random effects to improve model fit. We used the geographical coordinates of each site (longitude and latitude) to capture the spatial relationships among sites, for all regression models, following Zuur et al. (2009). For each GAM, we compared a more complex set of GAMM models that

Table 1 Study species and their ecological traits.

Eulamprus heatwolei Lampropholis guichenoti Lampropholis delicata Saproscincus mustelinus Egernia saxatilis Anepischtos maccoyi Pseudemoia spenceri Liopholis whitii Pseudemoia rawlinsoni Acritoscincus duperreyi

SVL (mm)

Habitat

Shelter use

Reproductive mode

Clutch size

Activity pattern

100 48 51 55 135 50 65 113 62 80

Terrestrial Terrestrial Terrestrial Terrestrial Terrestrial Fossorial Arboreal Terrestrial Terrestrial Terrestrial

Debris Leaf litter Leaf litter Leaf litter Crevices Debris Crevices Crevices Leaf litter Leaf litter

Viviparous Oviparous Oviparous Oviparous Viviparous Oviparous Viviparous Viviparous Viviparous Oviparous

3 6 4 4 2.5 3 2 4 6 6

Diurnal Diurnal Diurnal Crepuscular Diurnal Nocturnal Diurnal Diurnal Diurnal Diurnal

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

incorporated spatial correlation structure (either as Gaussian, spherical, exponential, linear or ratio distributions). Spatial correlation was considered significant when the best ranked model incorporating spatial correlation was within 2 AIC units of the nonspatially considered model. Where spatial autocorrelations were detected, we re-ran the analyses using GAMMs to incorporate the spatial elements (Zuur et al., 2009). Due to statistical constraints, percentage deviance explained for GAMM models were taken from the equivalent GAM models instead. All analyses were conducted in the statistical software R (R Development Core Team, 2015). Functional indices were calculated in the package FD (Laliberte and Shipley, 2011).

209

3.3. Community responses to disturbance Species richness significantly increased with time since fire in coastal Banksia woodland, as did community abundance (Fig. 2). Species evenness showed no response (P > 0.05; Table 2). In lowland forests, species richness did not correlate with time since fire, whereas community abundance decreased significantly with time since fire. Species evenness showed no response (Table 2). Neither community abundance, nor species richness, nor species evenness responded to time since timber harvesting (all P > 0.05; Fig. 2; Table 2). 3.4. Functional diversity responses to disturbance

3. Results 3.1. Species composition and summary statistics We captured 3677 lizards representing 10 skink species. Four species dominated the assemblage, comprising 94.5% of all captures, and were each present at >80% of all sites, including yellow bellied water skink (Eulamprus heatwolei; 41% of captures), garden skink (Lampropholis guichenoti; 32.8% of captures), weasel skink (Saproscincus mustelinus; 11.1% of captures), and delicate skink (Lampropholis delicata; 8.1% of captures). Three other species were also captured in sufficient numbers for analysis, including the black rock skink (Egernia saxatilis), Spencer’s skink (Pseudemoia spenceri), and McCoy’s skink (Anepischtos maccoyi; Appendix 1). Eulamprus heatwolei abundance (per 100 trap nights) averaged 6.3 ± 5.1 (SD) per site in Banksia woodland, 4.3 ± 3.9 in eucalypt forest sites investigated for the effects of fire, and 6.8 ± 5.6 in eucalypt forest sites investigated for timber harvesting. Lampropholis guichenoti abundance per site averaged 4.8 ± 3.5 in Banksia woodland, 3.6 ± 4.0 in eucalypt forest sites for the study of fire, and 5.1 ± 4.1 in eucalypt forest sites for the study of timber harvesting. Lampropholis delicata abundance per site averaged 1.0 ± 1.1 in Banksia woodland, 1.0 ± 1.4 in eucalypt forest sites for study of fire, and 1.4 ± 1.1 in eucalypt forest sites for study of timber harvesting. Saproscincus mustelinus abundance per site averaged 1.3 ± 1.3 in Banksia woodland, 1.2 ± 1.0 in eucalypt forests sites for study of fire, and 2.1 ± 1.5 in eucalypt forest sites for study of timber harvesting. On average, the capture rate of lizards (per 100 trap nights) was 14.18 ± 5.58 in Banksia woodland, 10.83 ± 5.89 in eucalypt forest sites for study of fire and 16.63 ± 7.24 in eucalypt forest sites for study of timber harvesting.

No significant relationships were found between functional diversity and fire or timber harvesting (Table 2; Fig. 3). Apart from functional divergence, all relationships between functional metrics and disturbance history were unchanged after removing the dominant species (Eulamprus heatwolei; Table 2). We examined correlations among functional indices. Functional evenness was marginally correlated with functional divergence in sites subject to timber harvesting in lowland eucalypt forest (t = 2.086, P = 0.046). No other significant correlations were identified among functional indices. This was true even after removing and checking the influence of the most common species (Eulamprus heatwolei). We also examined correlations between functional indices and species richness and evenness. Species richness was significantly correlated with functional richness in Banksia woodland (estimate = 11.08, t = 3.269, P = 0.003). Species evenness was significantly correlated with functional evenness in coastal Banksia woodland (estimate = 0.43, t = 4.135, P < 0.001) and in eucalypt vegetation subject to timber harvesting (estimate = 0.45, t = 2.963, P = 0.006). 4. Discussion Conservation of biodiversity in both natural and intensively managed forests depends on understanding the mechanisms underlying community responses to disturbance. Here we demonstrate that using measures that focus on different aspects of biodiversity responses to disturbance informs our understanding of the relationships between fire, timber harvesting and the assemblage of animal communities. 4.1. Species responses

3.2. Individual species responses to disturbance In coastal Banksia woodland, three of the seven species of skinks analysed responded to time since fire (Table 2). The relative abundances of Eulamprus heatwolei and Egernia saxatilis increased with time since fire in woodland vegetation. Saproscincus mustelinus populations showed a curvilinear trend, peaking approximately 15 years post-fire. The remaining four species did not respond to time since fire in woodland vegetation (Fig. 1a). In eucalypt forest, Eulamprus heatwolei abundance increased with time since fire, while that of Lampropholis guichenoti was negative. No post-fire responses were detected for other species in eucalypt vegetation (Table 2; Fig. 1b). Both Eulamprus heatwolei and Saproscincus mustelinus displayed significant responses to timber harvesting (logging was only conducted in eucalypt forests). The population trend of Eulamprus heatwolei was curvilinear, peaking at approximately 15 years post-harvesting (Fig. 1c), while that of Saproscincus mustelinus increased with age since timber harvesting (Fig. 1c). No responses were detected for other species (Table 2; Fig. 1c).

Responses to fire and timber harvesting by individual species of lizards were variable. For example, Eulamprus heatwolei was more abundant in later successional stages post-fire, but not in logged areas. This is likely to be due to its microhabitat preference for coarse woody debris (Heatwole and Taylor, 1987), which usually occur at higher cover in older post-fire vegetation (Bassett et al., 2015). Saproscincus mustelinus abundance was influenced by both time since fire and time since timber harvesting, but not in both habitat types. This species may be dependent on ground layer vegetation and favours shaded habitats, which are removed by both disturbances and recover slowly (Urlus, 2009). Although data on the fine-scale requirements of Saproscincus mustelinus is scarce, the ecology of its congeners suggests that it occurs at higher abundance in long undisturbed areas because of its limited tolerance to changes in thermal and hydric conditions (Moussalli et al., 2005). Lampropholis guichenoti favoured recently disturbed sites. This is consistent with the species’ known preference for open habitats (Cogger, 2014) and previous studies which showed that it inhabits recently burnt vegetation with low ground cover (Penn et al.,

210

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

Table 2 Modelled relationships between reptiles and disturbance history. Where model type was represented by GAMM, this indicated that spatial autocorrelation had been detected. Level Community

Response variable Species richness

Community abundance

Species evenness

Functional

Functional richness

Functional evenness

Functional divergence

Functional richness (minus Eulamprus heatwolei)

Functional evenness (minus Eulamprus heatwolei)

Functional divergence (minus Eulamprus heatwolei) Species

Eulamprus heatwolei abundance

Lampropholis guichenoti abundance

Lampropholis delicata abundance

Saproscincus mustelinus abundance

Egernia saxatilis abundance

Pseudemoia spenceri abundance

Anepischtos maccoyi abundance

* ** ***

Analysis

edf

F

P-value ***

Model type

R2 (adj)

% deviance explained

Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest

1 1 1 1 1 2.02 1 1 1

15.86 0.26 0.08 21.38 10.97 3.15 3.38 1.98 3.43

<0.001 0.613 0.779 <0.001*** 0.003** 0.051 0.075 0.172 0.074

GAMM GAM GAM GAMM GAMM GAMM GAMM GAMM GAMM

0.25 0.03 0.03 0.14 0.28 0.17 0.06 0.08 0.06

27.2 1.0 0.3 16.8 31.0 21.8 8.5 13.8 8.8

Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest

1 1 1.37 1 1 1 1 1 1 1 1 1.30 1 1 1 1 1 1

0.03 2.4 1.89 1.97 1.02 0.97 1.07 0.35 0.1 0.33 3.77 2.27 0.28 0.33 1.12 0.04 0.06 4.74

0.864 0.133 0.167 0.17 0.323 0.334 0.31 0.557 0.751 0.568 0.065 0.196 0.596 0.57 0.3 0.852 0.807 0.038*

GAMM GAMM GAM GAMM GAM GAMM GAM GAMM GAM GAMM GAMM GAM GAM GAM GAMM GAMM GAM GAM

0.03 0.05 0.05 0.03 0.00 0.00 0.00 0.02 0.03 0.01 0.06 0.06 0.01 0.03 0.01 0.01 0.04 0.12

0.1 8.2 9.4 0.9 3.8 2.4 3.1 1.8 0.4 3.1 9.9 10.8 0.4 1.5 4.4 0.0 2.4 14.9

Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest Fire Banksia woodland Fire lowland forest Logging lowland forest

1 1 2.49 1 1 1 1 1 1.58 2.25 1 1 1 1 1 1 1 1 1 1.6 1

28.8 7.52 6.05 2.51 6.6 0.01 0.01 0.05 0.67 7.12 0.1 8.27 7.99 0.87 1.33 1.24 0.07 0.15 0.27 0.81 1.24

<0.001*** 0.011* 0.004** 0.122 0.016* 0.922 0.931 0.818 0.47 0.004** 0.751 0.007** 0.008** 0.359 0.257 0.273 0.793 0.7 0.605 0.418 0.274

GAMM GAM GAMM GAM GAMM GAMM GAMM GAMM GAMM GAM GAMM GAMM GAMM GAMM GAMM GAM GAMM GAM GAMM GAMM GAMM

0.2 0.19 0.3 0.04 0.19 0.03 0.03 0.05 0.08 0.31 0.05 0.17 0.14 0.00 0.04 0.00 0.05 0.03 0.03 0.10 0.01

23.7 22.4 35.9 7.1 22.3 0.0 38.7 4.8 12.6 35.1 2.1 19.2 16.8 3.3 6.9 3.6 0.8 0.5 0.0 16.5 4.0

P < 0.05. P < 0.01. P < 0.001.

2003). Lampropholis guichenoti has previously been found to be more abundant in sites with simpler ground layer, with little leaf litter (Anderson and Burgin, 2002). 4.2. Community responses to disturbance The relationships between disturbance and lizard community richness, evenness and abundance were context-dependent. For example, in both eucalypt and woodland vegetation, fires had substantial effects on richness and abundance that lasted for decades. This is evidence that relatively old forests (undisturbed for over 30 years) will benefit populations of some lizard species. Metaanalysis shows that species richness increases with time since disturbance (Svensson et al., 2012), as long undisturbed habitats potentially contain vegetation of greater complexity, and thus greater niche space for different species (Hu et al., 2013a). Higher richness is often used as an indicator of habitat quality (Elmqvist

et al., 2003). Nevertheless, recently disturbed areas provide habitat for several species in eucalypt forests and Banksia woodland. This includes Lampropholis guichenoti that is attracted by open vegetation in recently disturbed areas, increasing species richness in these areas. Thus, species richness did not necessarily increase with time since disturbance because lizard species inhabited a range of successional states. This is consistent with studies of reptiles in other ecosystems (e.g., Pike et al., 2011; Jofre et al., 2016). Fire and timber harvesting influenced the reptile community in different ways, with different abundance trends observed in each disturbance type. Whereas clear changes were observed in the reptile community post-fire, timber harvesting had no significant effect on reptile community responses. This could be due to the intrinsic differences among the disturbance types. Previous studies in the study region showed that fires modified understorey vegetation more so than selective logging of the overstorey, and differed in their effects on the mid-storey (Hu et al., 2013a,b). In turn, these

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

211

Fig. 1. Responses in relative abundances of the seven most common lizard species to age post-disturbance in context of fire in coastal Banksia woodland, fire in lowland forest, and timber harvesting in lowland forest. Solid lines represent predictions of GAMs. Dotted lines represent ±2 SE. P-values and their respective r2 have been placed on graphs which showed statistically significant correlations (P < 0.05).

212

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

Fig. 2. Responses of lizard species richness, community abundances (captures per 100 trap nights) and evenness to age post-disturbance for analyses of: (column a) fire in coastal Banksia woodland, (column b) fire in lowland forest, and (column c) timber harvesting in lowland forest. Solid lines represent model predictions based on generalized additive models. Dotted lines represent ±2 SE. P-values and their respective r2 have been placed on graphs which showed statistically significant correlations (P < 0.05).

differences will result in different patterns of reptilian secondary succession. Broad vegetation types (Banksia woodland vs. eucalypt forest) had an important effect on reptile community response to fire. The influence of vegetation type on reptile communities is consistent with other studies from southern Australia (e.g., Mac Nally and Brown, 2001). The differences in reptile abundance and postdisturbance responses between coastal Banksia woodland and lowland forests could be due to microhabitat differences at the ground level such as litter and coarse woody debris that provide shelter, food and basking opportunities (Facelli and Pickett, 1991). Indeed, on average, our study sites in eucalypt forests contained thicker leaf litter layer and higher quantities of woody

debris than those in Banksia woodland (Hu et al., unpublished data). Habitat availability is a major driving force of reptile responses to disturbances. Our earlier work in the same study region shows that the additive effect of canopy cover and shrub layer complexity is important in explaining the overall diversity and abundance of this reptile assemblage, and variations in responses to fire and timber harvesting are likely related to those different rates of vegetation recovery following disturbance (Hu et al., 2013a,b). In addition, as discovered in the Mediterranean, the heliothermic requirements of reptile species can explain their lower abundance in long unburnt forests, which often had denser canopy, allowing less opportunity for basking (Santos and Cheylan, 2013).

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

213

Fig. 3. Responses of the three functional indices to age post-disturbance in context of (a) fire in coastal Banksia woodland, (b) fire in lowland forest, and (c) timber harvesting in lowland forest. Solid lines represent predictions of GAMs. Dotted lines represent ±2 SE.

4.3. Functional responses to disturbances None of the functional diversity metrics measured varied with time elapsed since disturbance in either vegetation type. The pattern remained similar even after excluding the most abundant species from the analyses (see e.g., Steen et al., 2010), suggesting that our results are relatively robust to the influence of common species. This finding contrasts with results from previous works in other vertebrate taxa, where significant relationships between functional evenness and landscape disturbances have been observed (e.g., urbanization; Devictor et al., 2008; FilippiCodaccioni et al., 2008; Gerisch et al., 2012). The observed lack of responses in most functional indices to disturbance could be due to the study community being relatively species-poor, with four generalist species dominating the assemblage. Even across the

ten species examined, the range and variance in trait values was relatively low, making it difficult to separate the species into functional groups. Determining fine-scale dietary preferences would enable the use of greater trait diversity and may increase resolution of functional responses in this community. 4.4. Synthesis There are several explanations for why various components of both community and functional diversity were largely unaffected by disturbance. First, many lizard species in the study region are generalist and opportunistic predators with broad ecological tolerances (Heatwole and Taylor, 1987), enabling them to inhabit a wide range of habitats, including disturbed areas (Howard et al., 2003). Second, high levels of habitat heterogeneity created by

214

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215

patchy burns and timber harvesting may allow quick recolonisation from adjacent patches, especially if those patches contain long undisturbed vegetation (Friend, 1993). Third, prescribed fires in the study region were relatively low in intensity, and may not have had as large an impact on vegetation structure as larger wildfires. It was beyond the scope of the present study to quantify differences between prescribed fires and wildfires and it remains a fruitful area for future study. Modelling fire severity, fine-scale habitat heterogeneity and microclimatic fluxes will likely yield additional insights into reptile distributions in eucalypt forests and woodlands. We detected few correlations between species richness and functional diversity measures under both disturbance regimes and in different vegetation types. The weak relationships between species richness and functional richness may result from species richness being relatively invariant among the disturbance types and ages. Gerisch et al. (2012) found the functional trait-based approach to be less effective in species-poor systems. Functional changes in the reptile community are expected to contribute to ecological processes and energy fluxes in forest and woodland landscapes. Reptiles constitute substantial proportions of ground level animal biomass in the study area and their functional influence may be largely through changes in total abundance with age since disturbance rather than changes in functional traits per se (Hu et al., 2013a,b). The abundance of reptiles will exert influence on rates of leaf herbivory, detritus turnover and seed dispersal (Heatwole and Taylor, 1987; Wilson and Swan, 2008). Overall, this study demonstrates that forest disturbances play an important role in driving dynamics in lizard communities in a temperate forest and woodland ecosystem. Indicators at different levels of biodiversity provide different, but incomplete, insights into community dynamics after disturbance. Thus, studies that incorporate multiple elements of biodiversity, including individual species responses, as well as community and functional responses, provide a more holistic understanding of the consequences of biodiversity changes in natural and managed landscapes.

Acknowledgments This project was funded and supported by Zoos Victoria, University of Melbourne and the Department of Sustainability and Environment, Victoria. We are grateful to Tim Lockwood for assistance in establishing the infrastructure and also with sampling, Peter Jenkins and Christopher Anderson for providing site-level information, and the volunteers who assisted with field work. Three reviewers made valuable comments that improved the manuscript.

Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foreco.2016.07. 040.

References Alexander, J.S.A., Scotts, D.J., Loyn, R.H., 2002. Impacts of timber harvesting on mammals, reptiles and nocturnal birds in native hard-wood forests of East Gippsland, Victoria: a retrospective approach. Aust. For. 65, 182–210. Anderson, L., Burgin, S., 2002. Influence of woodland remnant edges on small skinks (Richmond, New South Wales). Austral Ecol. 27, 630–637. Aponte, C., Tolhurst, K.G., Bennett, L.T., 2014. Repeated prescribed fires decrease stocks and change attributes of coarse woody debris in a temperate eucalypt forest. Ecol. Appl. 24, 976–989. Asner, G.P., Knapp, D.E., Broadbent, E.N., Oliveira, P.J.C., Keller, M., Silva, J.N., 2005. Selective logging in the Brazilian Amazon. Science 310, 480–482.

Azor, J.S., Santos, X., Pleguezuelos, J.M., 2015. Conifer-plantation thinning restores reptile biodiversity in Mediterranean landscapes. Forest Ecol. Manage. 354, 185–189. Bassett, M., Chia, E.K., Leonard, S.W., Nimmo, D.G., Holland, G.J., Ritchie, E.G., Clarke, M.F., Bennett, A.F., 2015. The effects of topographic variation and the fire regime on coarse woody debris: insights from a large wildfire. Forest Ecol. Manage. 340, 126–134. Bishop, T.R., 2012. Functional Diversity and Community Assembly Patterns in Ant (Hymenoptera: Formicidae) Communities Across a Forest Disturbance Gradient in Sabah, Malaysia Doctoral Dissertation, Masters’ Thesis. Imperial College of London. Bragg, J.G., Taylor, J.E., Fox, B.J., 2005. Distributions of lizard species across edges delimiting open-forest and sand-mined areas. Austral Ecol. 30, 188–200. Cadotte, M.W., Carscadden, K., Mirotchnick, N., 2011. Beyond species: functional diversity and the maintenance of ecological processes. J. Appl. Ecol. 48, 1079– 1087. Carthew, S.M., Horner, B., Jones, K.M.W., 2009. Do utility corridors affect movements of small terrestrial fauna? Wildlife Res. 36, 488–495. Chapin III, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., 2000. Consequences of changing biodiversity. Nature 405, 234–242. Chia, E.K., Bassett, M., Leonard, S.W., Holland, G.J., Ritchie, E.G., Clarke, M.F., Bennett, A.F., 2016. Effects of the fire regime on mammal occurrence after wildfire: site effects vs landscape context in fire-prone forests. Forest Ecol. Manage. 363, 130–139. Cogger, H., 2014. Reptiles and Amphibians of Australia. CSIRO Publishing. Devictor, V., Julliard, R., Jiguet, F., 2008. Distribution of specialist and generalist species along spatial gradients of habitat disturbance and fragmentation. Oikos 117, 507–514. Dormann, C.F. et al., 2007. Methods to account for spatial autocorrelation in the analysis of species distributional data: a review. Ecography 30, 609–628. Driscoll, D.A., Henderson, M.K., 2008. How many common reptile species are fire specialists? A replicated natural experiment highlights the predictive weakness of a fire succession model. Biol. Conserv. 141, 460–471. Driscoll, D.A., Lindenmayer, D.B., Bennett, A.F., Bode, M., Bradstock, R.A., Cary, G.J., Clarke, M.F., Dexter, N., Fensham, R., Friend, G., Gill, M., 2010. Fire management for biodiversity conservation: key research questions and our capacity to answer them. Biol. Conserv. 143, 1928–1939. DSE, 2007. Code of Practice for Timber Production. Department of Sustainability and Environment, Victoria. DSE, 2009. Management Procedures for Timber Harvesting, Roading and Regeneration in Victoria’s State Forests. Department of Sustainability and Environment, Victoria. Elmqvist, T., Folke, C., Nyström, M., Peterson, G., Bengtsson, J., Walker, B., Norberg, J., 2003. Response diversity, ecosystem change, and resilience. Front. Ecol. Environ. 1, 488–494. Facelli, J.M., Pickett, S.T.A., 1991. Plant litter: its dynamics and effects on plant community structure. Bot. Rev. 57, 1–32. Filippi-Codaccioni, O., Devictor, V., Clobert, J., Julliard, R., 2008. Effects of age and intensity of urbanization on farmland bird communities. Biol. Conserv. 141, 2698–2707. Fox, B.J., 1982. Fire and mammalian secondary succession in an Australian coastal heath. Ecology, 1332–1341. Friend, G.R., 1993. Impact of fire on small vertebrates in mallee woodlands and heathlands of temperate Australia: a review. Biol. Conserv. 65, 99–114. Gardner, T.A., Barlow, J., Peres, C.A., 2007. Paradox, presumption and pitfalls in conservation biology: the importance of habitat change for amphibians and reptiles. Biol. Conserv. 138, 166–179. Gerisch, M., Agostinelli, V., Henle, K., Dziock, F., 2012. More species, but all do the same: contrasting effects of flood disturbance on ground beetle functional and species diversity. Oikos 121, 508–515. Hart, S.A., Chen, H.Y.H., 2008. Fire, logging, and overstorey affect understorey abundance, diversity, and composition in boreal forests. Ecol. Monogr. 78, 123– 140. Heatwole, H.F., Taylor, J., 1987. Ecology of Reptiles. Surrey Beatty and Sons Pty Limited. Howard, R., Williamson, I., Mather, P., 2003. Structural aspects of microhabitat selection by the skink Lampropholis delicata. J. Herpetol. 37, 613–617. Hu, Y., Magaton, S., Gillespie, G., Jessop, T.S., 2013a. Small reptile community responses to rotational logging. Biol. Conserv. 166, 76–83. Hu, Y., Urlus, J., Gillespie, G., Letnic, M., Jessop, T.S., 2013b. Evaluating the role of fire disturbance in structuring small reptile communities in temperate forests. Biodivers. Conserv. 22, 1949–1963. Hughes, A.R., Byrnes, J.E., Kimbro, D.L., Stachowicz, J.J., 2007. Reciprocal relationships and potential feedbacks between biodiversity and disturbance. Ecol. Lett. 10, 849–864. Jellinek, S., Driscoll, D.A., Kirkpatrick, J.B., 2004. Environmental and vegetation variables have a greater influence than habitat fragmentation in structuring lizard communities in remnant urban bushland. Austral Ecol. 29, 294–304. Jofre, G.M., Warn, M.R., Reading, C.J., 2016. The role of managed coniferous forest in the conservation of reptiles. Forest Ecol. Manage. 362, 69–78. Kelly, L.T., Nimmo, D.G., Spence-Bailey, L.M., Haslem, A., Watson, S.J., Clarke, M. F., Bennett, A.F., 2011. Influence of fire history on small mammal distributions: insights from a 100-year post-fire chronosequence. Divers. Distrib. 17, 462–473.

Y. Hu et al. / Forest Ecology and Management 379 (2016) 206–215 Keller, M., Palace, M., Asner, G.P., Pereira, R., Silva, J.N.M., 2004. Coarse woody debris in undisturbed and logged forests in the eastern Brazilian Amazon. Global Change Biol. 10, 784–795. Kilpatrick, E.A., Waldrop, T.S., Lanham, J.D., Greenberg, C.H., Contreras, T.H., 2010. Short-term effects of fuel reduction treatments on herpetofauna from the southeastern United States. For. Sci. 56, 122–130. Laliberte, E., Shipley, B., 2011. FD: Measuring functional diversity (FD) from multiple traits, and other tools for functional ecology, R Package, . Legge, S., Murphy, S., Heathcote, J., Flaxman, E., Augusteyn, J., Crossman, M., 2008. The short-term effects of an extensive and high-intensity fire on vertebrates in the tropical savannas of the central Kimberley, northern Australia. Wildlife Res. 35, 33–43. Mac Nally, R., Brown, G.W., 2001. Reptile and habitat fragmentation in the boxironbark forests of central Victoria, Australia: predictions, compositional change and faunal nestedness. Oecologia 128, 116–125. Mason, N.W.H. et al., 2005. Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos 111, 112– 118. McGill, B.J. et al., 2007. Species abundance distributions: moving beyond single prediction theories to integration within an ecological framework. Ecol. Lett. 10, 995–1015. Moretti, M., Duelli, P., Obrist, M.K., 2006. Biodiversity and resilience of arthropod communities after fire-disturbance in temperate forests. Oecologia 149, 312– 327. Moussalli, A., Hugall, A.F., Mortiz, C., 2005. A mitochondrial phylogeny of the rainforest skink genus Saproscincus, Wells and Wellington (1984). Mol. Phylogenet. Evol. 34, 190–202. Nicholson, E., Lill, A., Andersen, A., 2006. Do tropical savanna skink assemblages show a short-term response to low-intensity fire? Wildlife Res. 33, 331–338. Nimmo, D., Kelly, L.T., Spence-Bailey, L.M., Watson, S.J., Haslem, A., White, J.G., Clarke, M.F., Bennett, A.F., 2012. Predicting the century-long post-fire responses of reptiles. Global Ecol. Biogeogr. 21, 1062–1073. Nimmo, D.G., Kelly, L.T., Farnsworth, L.M., Watson, S.J., Bennett, A.F., 2013. Why do some species have geographically varying responses to fire history? Ecography 37, 805–813. Olalla-Tárraga, M.A., Rodriguez, M.A., Hawkins, B.A., 2006. Broad-scale patterns of body size in squamate reptiles of Europe and North America. J. Biogeogr. 33, 781–793. Penn, A.M., Sherwin, W.B., Lunney, D., Banks, P.B., 2003. The effects of a lowintensity fire on small mammals and lizards in a logged, burnt forest. Wildlife Res. 30, 477–486. Petchey, O.L., Gaston, K.J., 2002. Functional diversity (FD), species richness and community composition. Ecol. Lett. 5, 402–411.

215

Petchey, O.L., Gaston, K.J., 2006. Functional diversity: back to basics and looking forward. Ecol. Lett. 9, 741–758. Pike, D.A., Webb, J.K., Shine, R., 2011. Removing forest canopy cover restores a reptile assemblage. Ecol. Appl. 21, 274–280. R Development Core Team, 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Reich, P.B., Bakken, P., Carlson, D., Frelich, L.E., Friedman, S.K., Grigal, D.F., 2001. Influence of logging, fire, and forest type on biodiversity and productivity in southern boreal forests. Ecology 82, 2731–2748. Santos, X., Cheylan, M., 2013. Taxonomic and functional response of a Mediterranean reptile assemblage to a repeated fire regime. Biol. Conserv. 168, 90–98. Smith, A.L., Bull, C.M., Driscoll, D.A., 2013. Successional specialization in a reptile community cautions against widespread planned burning and complete fire suppression. J. Appl. Ecol. 50, 1178–1186. Steen, D.A., Rall-McGee, A.E., Hermann, S.M., Stiles, J.A., Stiles, S.H., Guyer, C., 2010. Effects of forest management on amphibians and reptiles: generalist species obscure trends among native forest associates. Open Environ. Sci. 4, 24–30. Steen, D.A., Smith, L.L., Conner, L.M., Litt, A.R., Provencher, L., Hiers, J.K., Pokswinski, S., Guyer, C., 2013. Reptile assemblage response to restoration of firesuppressed longleaf pine sandhills. Ecol. Appl. 23, 148–158. Sutton, W.B., Wang, Y., Schweitzer, C.J., 2013. Amphibian and reptile responses to thinning and prescribed burning in mixed pine-hardwood forests of northwestern Alabama, USA. Forest Ecol. Manage. 295, 213–227. Sutton, W.B., Wang, Y., Schweitzer, C.J., Steen, D.A., 2014. Lizard microhabitat and microclimate relationships in southeastern pine-hardwood forests managed with prescribed burning and thinning. For. Sci. 60, 180–190. Svensson, J.R., Lindegarth, M., Jonsson, P.R., Pavia, H., 2012. Disturbance-diversity models: what do they really predict and how are they tested? Proc. Roy. Soc. B 279, 2163–2170. Tews, J., Brose, U., Grimm, V., Tielborger, K., Wichmann, M.C., Schwager, M., Jeltsch, F., 2004. Animal species diversity driven by habitat heterogeneity/diversity: the importance of keystone structures. J. Biogeogr. 31, 79–92. Turner, M.G., 2010. Disturbance and landscape dynamics in a changing world. Ecology 91, 2833–2849. Urlus, J., 2009. The Influence of Vegetation Structure and Land Management on East Gippsland Reptile Communities Honours Thesis. Deakin University. Villeger, S., Mason, N.W., Mouillot, D., 2008. New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology 89, 2290–2301. Wilson, S., Swan, G., 2008. A Complete Guide to Reptiles of Australia. New Holland Publishers, Chatswood, NSW. Zuur, A., Ieno, E.N., Walker, N., Savaliev, A.A., Smith, G.M., 2009. Mixed Effects Models and Extensions in Ecology with R. Springer.