Assessing the ecological health of rivers when they are dry

Ecological Indicators 85 (2018) 537–547

Contents lists available at ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Research paper

Assessing the ecological health of rivers when they are dry Alisha L. Steward

a,b,⁎

a

, Peter Negus , Jonathan C. Marshall

a,b

MARK a

c

, Sara E. Clifford , Catherine Dent

a

Water Planning Ecology, Queensland Department of Science, Information Technology and Innovation, Ecosciences Precinct, 41 Boggo Road, Dutton Park, Queensland 4102, Australia Australian Rivers Institute, Griffith University, 170 Kessels Road, Nathan, Queensland 4111, Australia c Remote Sensing Centre, Queensland Department of Science, Information Technology and Innovation, Ecosciences Precinct, 41 Boggo Road, Dutton Park, Queensland 4102, Australia b

A R T I C L E I N F O

A B S T R A C T

Keywords: Ecological assessment Biological monitoring Temporary river Murray-Darling Basin Feral pigs

Rivers and streams that dry up are found on every continent, and can form a large proportion of river networks. When rivers are dry, traditional indicators of river health – such as aquatic macroinvertebrates, fish or water quality – cannot be measured. Aquatic health indicators are widely used to assess wetted habitats, but currently no universally applicable indicators have been developed or applied to assess dry riverbed health. Dry riverbeds are often the ‘typical’ state of many intermittent rivers and streams; however, the ecological health of these habitats is rarely, if ever, assessed in monitoring programs. Resource managers have called for indicators of intermittent river health during the dry phase. The use of terrestrial invertebrate biota (e.g. ants, beetles, and spiders) as indicators in this study provides a novel solution to assessing rivers when they are dry. We developed a conceptual model of human-induced stressors (i.e. disturbance by livestock and feral mammals) on dry riverbed biota, which guided the selection of potential health indicators. Livestock and feral mammals are one of the most significant stressors on riverine ecosystems in Queensland, and impact riverbeds by altering the substrate through compaction, rooting and pugging. We trialled the use of metrics of terrestrial invertebrate assemblages as indicators of dry riverbed health in four Australian dryland catchments: Bulloo, Paroo, Warrego and Nebine. We used quantile regression and found that terrestrial invertebrate communities responded negatively (and significantly, p < 0.05) to a gradient of disturbance, defined by on-the-ground field measurements of livestock and feral mammal impacts. This response to stressors was predicted by the initial conceptual model. We conclude that terrestrial invertebrates in this study are suitable indicators of dry riverbed health, as they are impacted by disturbance from livestock and feral mammals. They can be used in the same way that indicators, such as aquatic macroinvertebrates, are traditionally used to assess river health. We also successfully combined indicators of wet and dry habitats to provide a holistic assessment of the health of intermittent river ecosystems incorporating all sections of the river network. We suggest that this approach should be adopted by other river health monitoring programs in rivers around the world.

1. Introduction 1.1. Intermittent rivers are widespread, and there will be more of them in the future Rivers that temporarily cease to flow and dry up are a global phenomenon, being found on every continent and nearly every watershed (Datry et al., 2014). Intermittent rivers have been described as being more representative of the world’s river systems than those with perennial flows (Williams, 1988). Their spatial extent is likely to further increase as a result of the combined effects of altered land-use patterns,

climate change, and increased water extraction for human uses (Meehl et al., 2007; Palmer et al., 2008; Larned et al., 2010). These effects can increase the duration of dry spells in intermittent rivers, and can potentially convert perennial rivers to intermittent ones. 1.2. Dry riverbeds are not included in traditional river monitoring programs Environmental monitoring and assessment of aquatic ecosystems is undertaken to inform management: either by identifying reductions in river health in response to anthropogenic stressors, or sometimes to demonstrate the effectiveness of restoration actions. As such, it is a

⁎ Corresponding author at: Water Planning Ecology, Queensland Department of Science, Information Technology and Innovation, Ecosciences Precinct, 41 Boggo Road, Dutton Park, Queensland 4102, Australia. E-mail address: alisha.steward@griffithuni.edu.au (A.L. Steward).

https://doi.org/10.1016/j.ecolind.2017.10.053 Received 8 December 2016; Received in revised form 2 September 2017; Accepted 24 October 2017 1470-160X/ Crown Copyright © 2017 Published by Elsevier Ltd. All rights reserved.

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

However, few indicators used for river health monitoring are truly specific; for example, aquatic macroinvertebrates, which are widely used as indicators, respond to many varied stressors. The widespread occurrence of intermittent rivers and streams discussed above, together with the emerging realisation that terrestrial invertebrates of dry riverbeds are ubiquitous and contain elements of specialist fauna for this habitat (Wishart, 2000; Steward et al., 2011; Corti et al., 2013; Steward et al., 2017), suggest a broad relevance of these biota as indicators of dry river health. While terrestrial invertebrates have not previously been investigated as river health indicators, they have been successfully utilised for biomonitoring in other habitats. For example, ants (Formicidae) are routinely used as biological indicators of rehabilitated mine sites throughout Australia, including forests, semi-arid heathlands, subtropical shrublands and tropical savannah woodlands (Andersen et al., 2004; Andersen and Majer, 2004). Similarly, ground beetles (Carabidae) have been employed as biological indicators of exposed riverine sediments and riparian zones in Europe (Boscaini et al., 2000; Eyre and Luff, 2002; Kleinwächter and Rickfelder, 2007). However, the sensitivity of terrestrial invertebrates to stressors in dry riverbeds has not yet been investigated.

valuable tool for directing and supporting natural resource management (Apitz et al., 2006; Field et al., 2007; Norris et al., 2007). The unpredictability of flow, and subsequently of surface water presence and distribution along intermittent rivers have been recognised as challenges for environmental monitoring (Sheldon, 2005). Despite their prevalence, intermittent rivers, and in particular dry riverbeds, have often been neglected and frequently ignored in river management, policy, and monitoring programs throughout the world (Steward et al., 2012; Acuña et al., 2014; Mazor et al., 2014). Hence, gaps often exist in monitoring data sets used to assess river health when sites are dry and consequently not sampled during particular occasions, seasons or years. This problem is particularly likely to occur in semi-arid and arid regions, areas with Mediterranean climates, during the dry season in monsoonal ‘wet-dry’ tropics, and in other regions during drought conditions. Under these circumstances, monitoring and assessment of intermittent river ecosystems typically seeks out and considers only the wet parts of the system (e.g. by using aquatic macroinvertebrates as biological indicators: Chessman, 1995; Reynoldson et al., 1995), and does not represent the entire river network. This is an important deficiency because wet parts of an intermittent river network may be unrepresentative of the ecological health of the system in totum. Furthermore, even in the absence of surface water, dry riverbed habitats can be ‘healthy’ and can have ecological values that may otherwise be overlooked, such as unique biodiversity, their use as dispersal corridors for terrestrial biota, and sites for the storage and processing of organic matter and nutrients (McClain et al., 2003; Steward et al., 2011, 2012; Acuña et al., 2014; Sánchez-Montoya et al., 2016). Conversely, dry riverbed habitats may be degraded by various stressors and therefore ‘unhealthy’ (Chiu et al., 2017; Steward et al., 2017). There is thus a recognised need to develop indicators of intermittent river health during the dry phase (Acuña et al., 2014). Several potential solutions have been proposed, but they rely on targeting specific habitats that may not be present in all intermittent river systems, or may be unrepresentative of the overall health of the system being assessed. For example, Robson et al. (2011) suggested sampling remnant pool ‘drought refuges’ for aquatic invertebrates during the dry phase. These pools can have sparse, variable and patchy distributions at the catchment scale, so by their very nature are unrepresentative of the overall intermittent river network where they occur. Some reaches may not contain any surface water to sample. Variable taxonomic composition in refuge pools results from stochastic founder effects and strong biotic interactions (Sheldon et al., 2010). Furthermore, sampling itself may threaten refuge function and therefore system resilience by depleting the supply of future colonists utilising refuges, so their use as monitoring habitats may be undesirable. Leigh et al. (2013) partially overcame these issues with their suggestion to adopt hyporheic invertebrates as indicators of dry river health. However, not every dry riverbed has a functioning hyporheic zone due to a lack of hyporheic capacity (e.g. in bedrock or clay substrates), or to temporal absence due to drying subsurface water during extended periods without flow (Boulton and Stanley, 1995). To better overcome these issues we demonstrate the use of terrestrial riverbed invertebrates (sensu Steward et al., 2011) as sensitive dry river health indicators, followed by an approach to integrate these with more traditional aquatic indicators to assess the entire river network – representing both wet and dry reaches of an intermittent river system.

1.4. A conceptual model of livestock and feral mammal impacts on intermittent rivers 1.4.1. Natural intermittent rivers To understand how stressors from livestock and feral mammals impact intermittent rivers and streams during the dry phase, we developed a conceptual understanding of their ecological structure and function (Fig. 1). Natural dry riverbeds can contain interstitial spaces which are inhabited by terrestrial invertebrates, both when the bed substrate is coarse (e.g. cobble, pebble) or fine (e.g. sand, silt/clay) (Steward et al., 2011). A diversity of substrates and other habitat types supports a comparatively diverse invertebrate fauna. Coarse substrates, such as cobbles and pebbles, can provide structural complexity and interstitial spaces for terrestrial invertebrates (Paetzold et al., 2008). Sandy substrates are used by invertebrates that dig (e.g. Mecynotarsus spp. (Coleoptera: Anthicidae), Hashimoto and Hayashi, 2012; Steward, 2014). Silt/clay substrates can crack once dry, and produce long furrows in which invertebrates can reside, acting as cool microhabitats. Catchment vegetation cover minimises runoff, and therefore erosion, suggesting that more substrate types are available in areas rarely smothered by sediment. Functioning riparian zones provide buffers, which minimise nutrient and sediment loads entering waterways, and inputs of leaf litter and woody debris, which provide habitat and potential food resources for terrestrial invertebrates. 1.4.2. Impacted intermittent rivers Out of all human-mediated disturbances, land use change has one of the largest impacts on species richness (Murphy and Romanuk, 2014). The use of land for livestock grazing (particularly cattle) impacts upon both aquatic and terrestrial ecosystems (Fleischner, 1994; Agouridis et al., 2005), with 80% of rivers and riparian zones damaged by grazing in the United States (Belsky et al., 1999). Cattle grazing for meat production is the dominant land use in arid and semi-arid parts of Australia, where most of the river network is dry most of the time. More than 83% of the total area of the Australian state of Queensland, representing almost 145 million hectares, is managed for beef cattle and sheep grazing (Barson, 2013). Dry riverbeds can be subjected to numerous stressors, such as livestock trampling, overgrazing, weed infestation, gravel and sand extraction, wastewater discharge, inundation by dams and weirs, cropping, and their use as roads (Steward et al., 2012). However, stressors caused by livestock and feral mammals appear to be both widespread and ecologically important in our study area of Queensland, and are the focus of this study. Feral mammals are common inhabitants of Queensland, and include pigs (Sus scrofa), goats (Capra aegagrus hircus),

1.3. Terrestrial invertebrates as indicators for river health monitoring For an indicator to be effective for biological monitoring it needs to be relevant and sensitive (Andersen, 1999; Dobbie et al., 2013). To be relevant for river health assessment it must be applicable to rivers and streams within the region being assessed, and to be sensitive it needs to change in a measureable way along gradients of a stressor. An ideal indicator is also specific, so that it is responsive to gradients of a single stressor, allowing for direct diagnosis of changes in river health. 538

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Fig. 1. Conceptual model illustrating cause-effect relationships of livestock and feral mammals on terrestrial invertebrates in intermittent river ecosystems: upstream of the fence (background) represents an ungrazed and thus undisturbed dry riverbed, while downstream of the fence (foreground) is impacted by stressors resulting from livestock and feral mammals.

disturbance of banks and riverbeds by feral pigs that use their snout and feet to turn over the substrate in search of food (Fig. 2). Both pugging and rooting behaviours alter the bed habitat that invertebrates use, destroying the interstitial spaces that large-bodied invertebrates of dry riverbeds require. The rooting impact of pigs is somewhat similar to pugging; however, rooting behaviour can mix and turn over sediments while pugging compacts sediments. Livestock and feral mammals are attracted to riparian zones, riverbanks and riverbeds for shade, cooler temperatures, forage, and as corridors for movement and to access water (Kauffman and Krueger, 1984; Fleischner, 1994; Belsky et al., 1999; Steward et al., 2011). Livestock can form extensive tracks of bare ground in and around intermittent rivers as they tend to repeatedly use the same path (Trimble and Mendel, 1995; Fig. 2). Grazing can impact the function of riparian

sheep (Ovis aries) and cattle (Bos taurus) (Choquenot et al., 1996; Robertson and Rowling, 2000; Mitchell et al., 2007a, 2007b; Gentle and Pople, 2013), and generate similar stressors on river ecosystems as livestock grazing (Figs. 1 and 2, ). There are no native ungulates in Australia. The direct impacts of livestock and feral mammals on dry river ecosystems include a reduction in riparian vegetation cover; the compaction, pugging and rooting of soil; and eutrophication from their urination and defecation (Kauffman et al., 1983a; Belsky et al., 1999; Figs. 1 and 2). ‘Pugging’ is a term used to describe the physical disturbance of banks and riverbeds by the hoofs of animals, typically when the substrate is moist. This impact results in characteristic hoof indentations or ‘pug’ marks left in the substrate that can persist long after it has dried (Fig. 2). ‘Rooting’ is a term used to describe the physical 539

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Fig. 2. The effects of livestock and feral pigs on riverbeds and banks: a) cattle pugging across the width of a dry riverbed in the Cooper Creek catchment, Australia; b) riverbed disturbance from pig rooting near a waterhole in the Diamantina River catchment, Australia; and c) the difference between grazed (left) and ungrazed (right) land separated by a fence in the centre of the image, located in the Warrego River catchment, Australia. Note the exposed tree roots and bare ground on the left, in comparison to the increased ground vegetation and leaf litter, and natural benches up the bank on the right.

persistent waterholes in intermittent rivers, so are therefore highly subjected to trampling, pugging and rooting (Steward et al., 2011). Livestock and feral mammal urine and manure deposited directly into rivers, or washed in from adjacent areas, can affect water quality by increasing nitrate and phosphate levels (Schepers and Francis, 1982; Muenz et al., 2006) and lowering the dissolved oxygen content, causing fish mortality (Taylor et al., 1989). Feasibly, these added nutrients also have an impact on terrestrial invertebrate communities once riverbeds are dry (Fig. 1), potentially disrupting dry riverbed foodwebs. For example, added nutrients could foster the growth of cyanobacteria, which is relatively unpalatable to consumers once stranded as riverbeds dry up, as opposed to more palatable types of algae which represent higher quality food (Guo et al., 2017). Livestock and feral mammals contribute to an increase in bare ground through the consumption and trampling of vegetation (Kauffman et al., 1983a; Popolizio et al., 1994; Trimble and Mendel, 1995; Agouridis et al., 2005), and can hinder vegetation succession (Kauffman et al., 1983b). Barrios-Garcia et al. (2014) found that feeding and rooting by feral pigs in Hawaii reduced plant biomass by 60% relative to areas where pigs were excluded. In a study in Colorado, Schulz and Leininger (1990) found that grazed areas had four times more bare ground than areas that had been ungrazed for 30 years. The extent to which cattle can generate bare ground is further evident in a study by Mills et al. (1989) in south-west Queensland, where more than half of the properties examined had 60% of their land grazed bare, representing almost two million hectares. Pettit (2002) found that intensive cattle grazing along a river waterhole in the Cooper Creek catchment in western Queensland led to a significant reduction in vegetation cover and an increase in the percentage of bare ground. Increases in bare ground can threaten river ecosystems by accelerating surface water run-off and soil erosion during rainfall events and subsequently increasing fine sediment supply to waterways (Ludwig and

vegetation buffers and thus increase nutrient and sediment loads to adjacent streams (Lowrance et al., 1984; Naiman and Décamps, 1997; Muenz et al., 2006). Additionally, riparian vegetation has often been cleared by land managers to promote pasture growth for livestock, and this has been occurring over the last 200 or more years since grazing commenced in Australia (NLWA, 2002; Martin and McIntyre, 2007). A reduction in shade due to a loss of vegetation impacts terrestrial invertebrates, as dry riverbeds can reach high temperatures that exceed the 60 °C thermal tolerance of most eukaryotic organisms (Tansey and Brock, 1972; Steward et al., 2017) and can have large diurnal fluctuations, making them stressful environments. Trampling and compaction by livestock and feral mammals can compress sediments (Trimble and Mendel, 1995) and decrease soil porosity (Orr, 1960). This limits the availability of interstitial spaces and furrows for terrestrial invertebrates to inhabit, and water to infiltrate (Kauffman and Krueger, 1984), and in turn, decreases invertebrate abundance and biomass (Schon et al., 2010). Livestock trampling also disturbs the invertebrates living near the soil surface, reducing their abundance and species richness (Cluzeau et al., 1992). Trampling and feeding reduce vegetation cover, and as a result this reduces the amount and diversity of litter and soil organic matter content (Naeth et al., 1991; Green and Kauffman, 1995), which are likely to be important for the provision of habitat and food resources to terrestrial invertebrate communities. Bromham et al. (1999) found a more diverse invertebrate fauna in ungrazed woodlands compared to grazed woodlands, most likely due to a higher diversity of food and habitat in the areas of lower disturbance. Similarly, Lindsay and Lindsay and Cunningham, (2009) found that excluding livestock benefitted components of the invertebrate community, vegetation condition and the process of litter decomposition. Dry riverbeds represent important movement corridors used by native wildlife (SánchezMontoya et al., 2016), livestock and feral animals to access sparse,

540

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

the ground surface, and feral pig, sheep and goat impacts were assessed by the density of their tracks and rooting damage. Measurements of the stressor intensity were categorised as: none (1), light (2), moderate (3), or severe (4), and were averaged across all 80 quadrats to produce a score for each site. Subjective but consistent judgements by experienced field staff were used to describe the categories: light – visible presence of livestock or feral mammals, such as limited rooted areas or sparse hoof prints; moderate – obvious damage from pugging and rooting; and severe – obvious and extensive with pugging and rooting over 50% of the area covered by the transect (e.g. Fig. 2a, b).

Tongway, 2002; NLWA, 2002). The increased accumulation of deposited sediment in river channels is well documented to impact aquatic biota (Wood and Armitage, 1997) via changes to particle size distribution (Schälchli, 1992). We expect that deposited fine sediments such as silt and clay, which blanket the channel substrate and fill interstitial spaces (Richards and Bacon, 1994; Wood and Armitage, 1997; Bartley and Rutherfurd, 2005), will similarly impact dry riverbed terrestrial invertebrates by changing their habitat and thus their assemblage composition (Fig. 1). Degradation of dry riverbeds by livestock and feral mammals has not been studied previously, but given these high exposure rates and the analogous impacts known from other habitats, we expect that trampling, compaction, pugging and rooting will impact dry riverbed invertebrates directly by squashing (and therefore killing them), and indirectly by decreasing the availability of food and habitat (Fig. 1).

2.1.2. Terrestrial invertebrate sampling Terrestrial invertebrates were sampled in dry riverbeds using pitfall traps after Steward et al. (2011). Six replicate pitfall traps were positioned in the dry channel at each site and set for approximately 24 h. Traps were positioned in diverse micro-habitats within the dry channel, such as leaf litter, bare ground and woody debris, and across different substrate types. The pitfall traps consisted of 250 mL plastic jars, 77 mm high and 67 mm in diameter, filled to 100 mL with 70% Ethanol, 3% Glycerol and 27% water as per Steward et al. (2011) and Wishart (2000). Ethanol was used as a preservative and a killing agent. Glycerol was used to keep the invertebrate bodies supple for handling and identification. A drop of detergent was added to break the surface tension, which prevented captured invertebrates from escaping. A plastic cover was positioned approximately 100 mm over each trap to prevent rain, leaf litter and other debris from blocking the trap and reducing its efficiency (Williams, 1959). Specimens were identified in the laboratory with a stereo-microscope to order and then family level where possible, using taxonomic keys (CSIRO Division of Entomology, 1991). Ant subfamilies and genera, and beetle families were verified by a museum entomologist (Chris Burwell, Queensland Museum, pers. comm.).

1.5. Research questions We predicted that terrestrial invertebrate communities of dry riverbeds would respond to stressors related to the impacts of livestock and feral mammals. In this paper we ask: Do the terrestrial invertebrate communities of dry riverbeds show an ecological response to measures of this stressor, in terms of reduced taxon richness or abundance? If there is a response, can these metrics be used for river health assessment? 2. Methods This study tested the sensitivity of terrestrial invertebrate richness and abundance to stressors generated by livestock and feral mammals in four catchments in southern Queensland, Australia, and then applied these findings to the monitoring and assessment of river condition.

2.1.3. Data analyses Terrestrial invertebrate taxon richness and abundance were calculated as the average of the six replicates per site and then log10-transformed for regression analysis. We used quantile regression to test whether livestock and feral mammal intensity was a limiting factor on the invertebrate community. Quantile regression is a statistical technique that can be used to estimate the response effects of limiting factors (Cade and Noon, 2003; Brooks and Haeusler, 2016). It can determine whether there are relationships between a response and limiting factors that are evident across the distribution of the data and is especially useful where the data have heterogeneous variation (potentially due to influences of several limiting factors) (Cade and Noon, 2003). Quantile regression determines a linear regression function between a limiting factor and the proportion of the measured response data points (equivalent to the quantile) that fall below the regression line (Brooks and Haeusler, 2016). For our application, a quantile of 0.8 as the maximum limit with the total site number of 31 was selected based on the formula (maximum quantile < 1–(5/number of sites)) for determining quantile extremes (Rogers, 1992). We calculated the 0.8 linear quantile regression relationships between the stressor (mean livestock and feral mammal intensity score at each site) and the responses (mean log10 taxon richness and mean log10 abundance) using the rq function within the quantreg package (Koenker, 2016) using R software (R Core Team, 2013).

2.1. Testing the sensitivity of invertebrate richness and abundance to stressor intensity Testing of the sensitivity of the richness and abundance of dry riverbed terrestrial invertebrates as indicators of ecosystem response to disturbance from livestock and feral mammals was conducted in the Bulloo, Paroo, Warrego and Nebine river catchments (Fig. 3). All four have semi-arid climates, and contain rivers and streams with intermittent flow. Kennard et al. (2010) classified flow in the region as highly to extremely intermittent. The Bulloo catchment is an isolated drainage that flows into several terminal lakes. The Paroo, Warrego and Nebine catchments are part of the northern most drainages of Australia’s largest river system – the Murray-Darling Basin. Most of the area is used as grazing land for cattle and sheep, with feral pigs and goats recognised as important pest species (Negus et al., 2013). Sample sites were selected to be statistically representative of the region using a Generalised Random Tessellated Stratification (GRTS) design adapted for use on stream networks (Dobbie and Stevens, 2006; Dobbie and Negus, 2013). GRTS was used to ensure spatial balance and efficiency in site selection while accounting for operational difficulties that can occur during field sampling of specific sites (Dobbie et al., 2013). Thirty-one sites were sampled: six in the Bulloo catchment, eight in the Paroo, eleven in the Warrego and six in the Nebine catchment (Fig. 3, Supplementary Table 1). 2.1.1. Field-based estimates of stressor intensity Direct measures of stressor intensity were made at each site during September–November (austral spring) in 2012. Four 200-m transects were run parallel to the river channel on the bank at each site, with two transects along each bank where possible. Measurements of riverbed impact were made in contiguous 10 m x 10 m quadrats along each transect (20 quadrats per transect). Livestock impacts in each quadrat were assessed based on the density of pug markings and tracks along

2.2. Application of dry riverbed invertebrates as indicators for government reporting of multi-stressor river health assessments We undertook a multi-stressor assessment as part of the Queensland Government’s ongoing river health assessment program (Fig. 4). Stressors (i.e. grazing intensity) were used to measure the threat (livestock and feral mammals), and the biological response (e.g. terrestrial invertebrate richness) was used to measure dry riverbed condition. In 541

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Fig. 3. Map of sites sampled in the Bulloo, Paroo, Warrego and Nebine catchments (closed circles). Bold lines indicate catchment boundaries. The position of the catchments in the Australian state of Queensland is provided (inset).

this study, we utilised dry riverbed terrestrial invertebrates as one indicator to represent the ecological impacts of livestock and feral mammals. We used the dry riverbed terrestrial invertebrate richness and abundance metrics derived from the Bulloo, Paroo, Warrego and Nebine catchment samples, as described above. Additional indicators sampled at the same time and within the same reach as terrestrial invertebrates were used to represent the impacts of other stressors, and these were brought together into an overall assessment (Negus et al., 2015; Fig. 4). Our multi-stressor assessment identified priority threats to the health of each river system. Indicators of river health were chosen based on the threats which were evaluated to pose the greatest risk to these ecosystems, and were aggregated to present an ‘all of catchment’ assessment. Conceptual models (e.g. Fig. 1) were used to link cause to effect, allowing for a risk-based diagnosis of the most likely causes of any degradation in condition. Dry riverbed terrestrial invertebrate metrics were used to define river health based on the stressor of introduced riparian fauna, quantified using the livestock and feral mammal damage metrics described above. This was then combined with assessments of health based on other important stressors in the catchments (introduced aquatic fauna and deposited sediment in permanent waterholes) to give an overall picture of river health, incorporating both wet and dry parts of the rivers. For government reporting (Fig. 4), assessments used a reference condition approach (Hawkins et al., 2010; Dobbie et al., 2013), with expected values derived from sites formally evaluated to be minimally disturbed according to set criteria (DSITIA (Department of Science, Information Technology, Innovation and the Arts) (2013)). Site metrics were range standardised (Eq. (1)) to theoretically vary between zero (worst possible health) and one (health equivalent to reference) and aggregated to provide an assessment of catchment health.

S tan dardised score = 1 −

X − reference WCV − reference

(1)

Where: X = site value, Reference = reference value, WCV = value under realistic worst case scenario Reference values were zero for the stressor metric (the total livestock and feral mammal scores from six transects per site), as our scoring where there are no impacts of livestock or feral mammals would be zero – an “historical” reference condition using the definitions of Stoddard et al. (2006). The 20th percentile of mean log10 taxon richness (5.5) and abundance (15.17) from reference sites was used as the reference values for terrestrial invertebrates. A percentile value is used rather than the full range to account for anomalous observations influencing assessments (Dobbie et al., 2013). Worst case values were six for the stressor metric based on the value that would be obtained if all quadrats assessed were recorded as having severe stressor intensity, and zero for the mean sample taxon richness and abundance based on the hypothetical potential of no terrestrial invertebrates being recorded at a site. 3. Results 3.1. Stressor intensity Stressors attributed to livestock and feral mammals along river channels were assessed in all four catchments, with some sites appearing to be in natural condition – having no impacts recorded – while other sites had severe degradation measured in one or more of the assessed quadrats. Quadrats with severe degradation were recorded at sites in all four catchments, although it was not widespread within the sites. The average livestock and feral mammal impact score used to 542

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Fig. 4. Summary table from a Queensland Government technical report (Negus et al., 2015) on the health of river ecosystems based on the assessment of threats from introduced aquatic fauna, introduced riparian fauna (i.e. livestock and feral mammals), and deposited sediment. The summary table shows the results of each indicator sampled in each catchment, to demonstrate how terrestrial invertebrates were incorporated into the assessment of river health. The ‘current threat’ for introduced riparian fauna was a measure of the intensity of damage from livestock and feral mammals.

represent sites ranged from 1 (no impact) to 2 (light impact) (Table 1). The Nebine catchment had the lowest impact score (1.08), followed by the Warrego (1.16), Bulloo (1.27), and Paroo (1.37).

a site in the Paroo catchment.

3.2. Terrestrial invertebrates

The 0.8 quantile regression models for both invertebrate richness and abundance were significant (i.e. p < 0.05) (Table 2, Fig. 5). This indicates that the response of dry riverbed invertebrates is limited by the intensity of livestock and feral mammal disturbance.

3.3. Sensitivity to the stressor

Ninety invertebrate taxa, represented by 12,979 individuals, were recorded from the sampling sites (Supplementary Table 2). The following taxonomic groups were represented: Acari, Amphipoda, Araneae, Blattodea, Chilipoda, Coleoptera, Dermaptera, Diplura, Diptera, Entomobryomorpha, Hemiptera, Hymenoptera, Isopoda, Lepidoptera, Mantodea, Neuroptera, Orthoptera, Poduromorpha, Pseudoscorpionida, Psocoptera, Symphypleona, Thysanoptera, and Trichoptera. Ants (Formicidae) and springtails (Collembola) were collected from every site, with Iridomyrmex ants and Entomobryidae springtails present at 29 of the 31 sites (Supplementary Table 2). Flies (Diptera) and beetles (Coleoptera) were collected from 30 sites each, of which phorid flies were collected from 28 sites and anthicid beetles collected from 24 sites. Other commonly-encountered taxa included Acari (27 sites), Apocrita (27 sites), Cecidomyiidae (25 sites), and Lycosidae (22 sites). Uncommon taxa only found at a single site included Aleyrodidae, Chilipoda, Chrysopidae, Gryllotalpidae, Hydrophilidae, Mantinae, Micropholcommatidae, Miridae, Miturgidae, Mycetophagidae, Nematoda, Oecobiidae, Podomyrma, Pseudomyrmecinae, Reduviidae, Spercheidae, Tipulidae, and Trichoptera. The mean taxon richness of terrestrial invertebrates per pitfall trap at a site varied between sites and catchments, ranging from 3.8 taxa at a site in the Bulloo, to 13.8 taxa at a site in the Nebine catchment (Table 1). Mean terrestrial invertebrate abundance per pitfall trap ranged from 5.5 individuals at a site in the Bulloo, to 385 individuals at

3.4. River health assessment The intensity of the stressor (livestock and feral mammals) was assessed as ‘moderate’ in the Warrego, Paroo and Bulloo catchments, and ‘slight’ in the Nebine (‘current threat’, Fig. 4). The condition of dry riverbeds was assessed as being ‘slightly disturbed’ in all four catchments, based on the terrestrial invertebrate results. These results contributed to the overall river assessment across multiple stressors indicating the relative risk to the health of rivers in these catchments from different threatening processes (for more information refer to Negus et al., 2015). 4. Discussion 4.1. Suitability of terrestrial invertebrates as indicators for river health monitoring Several authors have recently recognised the need for indicators of intermittent river health during the dry phase (Steward et al., 2012; Acuña et al., 2014; Datry et al., 2014; Steward et al., 2017). This study fills that need. The terrestrial invertebrates of dry riverbeds meet the criteria as a suitable indicator because: 543

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Table 1 Terrestrial invertebrate index results for sites in the Bulloo, Nebine, Paroo and Warrego River catchments in Queensland, Australia. Catchment

Site name

Site number

Mean livestock and feral mammal impact score

Mean log10 taxon richness

Mean log10 abundance

Bulloo Bulloo Bulloo Bulloo Bulloo Bulloo Nebine Nebine Nebine Nebine Nebine Nebine Paroo Paroo Paroo Paroo Paroo Paroo Paroo Paroo Warrego Warrego Warrego Warrego Warrego Warrego Warrego Warrego Warrego Warrego Warrego

Blackwater Creek at Adavale Bulloo River at Norley Station Bulloo River at Pinketta Station Bulloo River at Quilpie Bulloo River at Thargomindah Fifteen Mile Creek at Hay Paddock Hut Waterhole Nebine Creek at Aqua Downs Nebine Creek at Murra Murra Nebine Creek at Roseleigh Crossing Wallum Creek at Bollon Reserve Wallum Creek at Homeboin Waterhole Wallum Creek at One Mile Waterhole Beechal Creek at Boolbury Waterhole Gumholes Creek at Bowra Moonjaree Creek at Moonjaree Paroo River at 14 Mile Waterhole Paroo River at Wimmera Waterhole Pingine Waterhole Rolwegan Creek at Unnamed Waterhole Yowah Creek at Thandy Waterhole North Ambathalla at Ambathalla Creek Cuttaburra Creek Tinnenburra Waterhole at Tinnenburra Langlo River at Rylestone Waterhole Ward River at Bayrick Fish Hole Ward River at Binnowee Warrego River at Augathella Warrego River at Charleville Warrego River at Cunnamulla Weir Warrego River at Merwah waterhole Warrego River at Wallen Warrego River at Wyandra

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

1.0031 1.4813 1.0594 1 1.0348 2.0281 1 1.1563 1.1375 1 1.1563 1.0063 1.0438 1.6781 1.5906 1.0206 1.9 1.1406 1 1.5813 1.2875 1.5375 1.2656 1.0031 1.1313 1.1063 1.0031 1 1.1281 1.0688 1.2188

9.2 6.5 3.8 6.3 8.2 6.3 13.8 7.5 5.5 13.7 12.0 11.2 7.7 5.3 9.8 3.8 6.5 6.8 8.3 4.7 8.2 10.5 3.8 6.8 4.5 11.3 9.5 12.2 7.8 5.7 6.3

330.2 108.5 5.5 11.5 29.8 34.7 91.7 114.5 90.3 42.2 47.3 10.5 32.7 15.2 21.2 31.5 7.3 101.3 5.8 24.8 35.5 385.0 120.7 92.2 30.0 55.3 250.0 26.3 35.8 15.2 17.5

Table 2 Quantile (0.8) regression analysis results for a) mean log10 taxon richness and b) mean log10 abundance. An asterisk indicates significant results (p < 0.05). Coefficients

Value

Standard Error

t value

Pr( > |t|)

a) Mean log10 taxon richness Intercept Livestock and feral mammal impact score

1.35 −0.27

0.17 0.13

7.84 −2.12

0.00 0.04*

b) Mean log10 abundance Intercept Livestock and feral mammal impact score

2.63 −0.71

0.31 0.24

8.59 −2.95

0.00 0.006*

However, we don’t currently know if it is other stressors or natural variation that are the unmeasured influences. Our results show that terrestrial invertebrates of dry riverbeds meet Andersen’s (1999) criteria for indicator selection. They are widely distributed, abundant and taxa-rich − terrestrial invertebrates were found at every site sampled in this study, and have been collected in high abundance and richness at these sites as well as at sites in many other catchments throughout Australia and other parts of the world (Wishart, 2000; Lalley et al., 2006; Larned et al., 2007; Steward et al., 2011; Corti et al., 2013; Corti and Datry, 2016). Terrestrial invertebrates are functionally important in ecosystems − ants in particular are recognised as the dominant terrestrial invertebrate group in the Australian environment (Andersen, 1997; Andersen et al., 2004). Particular taxonomic and functional groups of terrestrial invertebrates are known to be sensitive and responsive to environmental changes − for example, beetles from the family Carabidae (Boscaini et al., 2000; Eyre and Luff, 2002; Bates et al., 2007; Kleinwächter and Rickfelder, 2007) and functional groups of ants (King et al., 1998; Hoffmann and Andersen, 2003). Lastly, terrestrial invertebrates are easily sampled (traps are deployed one day and retrieved the next) using inexpensive field equipment (e.g. Steward et al., 2017), and are easy to sort and identify − the laboratory identification effort is comparable to that required for aquatic invertebrate samples. We did not address all possible stressors impacting the terrestrial invertebrates of dry riverbeds in Queensland in the conceptual model. Our model focused on habitat degradation caused by livestock and feral mammals. Of course there may be other relevant stressors that need to be included in future conceptual models. Additional stressors to riverine ecosystems in Queensland that may influence terrestrial invertebrates include:

1) They are relevant. Terrestrial invertebrates were found in all dry riverbeds sampled, they were abundant, and were common. They have also been recorded as abundant and common elsewhere in Australia and Europe (Steward et al., 2011, 2017). Terrestrial invertebrates are important components of intermittent river ecosystems (Steward et al., 2017). 2) They are sensitive. Variability in terrestrial invertebrate metrics was associated with a gradient representing intensity of livestock and feral mammal damage to dry riverbeds, thereby supporting our conceptual model. Terrestrial invertebrates have also responded to stressors in other habitats (Andersen et al., 2004; Andersen and Majer, 2004). 3) They are specific. We used quantile regression to account for natural variability in substrate composition, hydrology, etc. of the sampled sites. We included samples from different catchments to broaden the applicability of our results. The quantile regression showed that intensity of livestock and feral mammal damage limits invertebrate richness and abundance, but that other unmeasured influences exist. 544

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

Fig. 5. Quantile regression results for a) Mean log10 taxon richness and b) Mean log10 abundance in relation to grazing pressure. The 0.8 quantile regressions (solid lines) are both significant (p < 0.05).

• Acid soil run-off – affecting the pH of riverbed sediments; • Deposited sediment (resulting from land clearing) – altering dry riverbed habitats through in-filling of interstitial spaces; • Flow management – the frequency, timing, and duration of wet/dry • • • •

Nebine catchments where dry riverbeds were subjected to higher grazing impacts, as determined through the on-the-ground field measurements of livestock and feral mammal disturbance, were limited both in terms of terrestrial invertebrate taxon richness and the abundance of individuals. Our findings of terrestrial invertebrate responses to livestock and feral mammal impacts are well supported in the literature for other habitat types. For example, Schon et al. (2010) found high soil porosity was clearly associated with high abundance and biomass of soil invertebrates, and that these indices were lower in areas grazed by dairy cows as a result of soil compaction. Similarly, Cluzeau et al. (1992) found that cattle trampling compacted the soil and affected its structure, which greatly reduced earthworm density, biomass, and species richness. Leach et al. (2013) found more species of ants and beetles in forest than in nearby areas that had been cleared for grazing. Lindsay and Cunningham (2009) found that woodlands without recent grazing had a higher abundance of beetles. Bromham et al. (1999) found a more diverse ground invertebrate fauna in ungrazed than grazed woodlands, and this was attributed to differences in vegetation and ground litter. Feral pigs impact terrestrial invertebrates in a similar way: Vtorov (1993) found a higher abundance and richness of soil invertebrates, particularly springtails, from rainforests in Hawaii where feral pigs had been removed. Further research into the responses of terrestrial invertebrates to stressors could involve the collection of dry riverbed habitat data, such as substrate composition, from natural and degraded sites. This could determine whether some substrate types are more prone to degradation than others.

phases affecting the temporal and spatial availability of dry riverbed habitats; Habitat removal or disturbance (e.g. gravel extraction) – causing physical damage to riverbed habitats; Riparian habitat fragmentation (e.g. riparian tree clearing) – changing allochthonous food supplies; Salinity – potential toxic effects from salt in riverbed sediments; and Toxicants (e.g. pesticides and heavy metals) – potential toxic effects from contaminants in riverbed sediments.

To further explore the use of terrestrial invertebrates as biological indicators, conceptual cause/effect responses should be investigated with laboratory and/or field experiments (Dobbie et al., 2013). Other features of terrestrial invertebrate assemblages, such as traits (e.g. functional feeding groups), could also be examined. Future studies could use controlled experimental manipulation; for example, sites could be fenced from livestock and sampled as ground cover improves. Many studies have advocated the use of invertebrates as biological indicators in various terrestrial environments, whether it be entire assemblages (Obrist and Duelli, 2010) or specific taxonomic groups (ants: Majer, 1983; Peck et al., 1998; Andersen et al., 2002; Andersen et al., 2004; Andersen and Majer, 2004; Rainio and Niemelä, 2003; Michaels, 2007; Marc et al., 1999; Greenslade, 2007). It has been suggested that groups such as ants and beetles are likely to respond to habitat changes in different ways, and that using both ants and beetles together could strengthen assessments (Leach et al., 2013). We now recommend that terrestrial invertebrates be adopted as health indicators of dry riverbeds in Queensland and suggest they be tested in other catchments, and in other parts of the world, with both similar and different stressors to those assessed in this study.

4.3. Inclusion of dry riverbeds in river monitoring programs We successfully applied the results of our study to a government river health assessment in four catchments (Negus et al., 2015). We combined the terrestrial invertebrate results with other indicators for reporting purposes, illustrating how wet and dry river health indicators can be integrated into one assessment (Fig. 4). As far as we are aware, this is the first time that river monitoring programs have combined wet and dry indicators to represent the health of intermittent river channels. In the government report we also combined the assessments across multiple stressors: in summary, the threat intensity was quantified at reporting scales for each stressor, as was the resulting ecosystem condition based on the ecological response metrics and their deviation

4.2. Response of terrestrial invertebrates to livestock and feral mammal disturbance This study found a strong association between the response of terrestrial invertebrates that inhabit dry riverbeds and the impacts caused by livestock and feral mammals. Sites in the Bulloo, Paroo, Warrego and 545

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

from reference expectations. By considering these separately, it is possible to diagnose the causes of degradation and to prioritise management actions to conserve or restore ecosystems. The use of terrestrial invertebrates from dry riverbeds as indicators solves previous dilemmas over how to assess river health in the absence of surface water. We have confirmed that the impacts of stressors on the dry riverbeds of intermittent rivers can be quantified. Terrestrial invertebrates have now been adopted as indicators for government river health assessments in other catchments in Queensland.

http://www.agriculture.gov.au/SiteCollectionDocuments/natural-resources/soils/ trends-factsheets/qld-practices-grazing.pdf. Bartley, R., Rutherfurd, I., 2005. Re-evaluation of the wave model as a tool for quantifying the geomorphic recovery potential of streams disturbed by sediment slugs. Geomorphology 64, 221–242. Bates, A.J., Sadler, J.P., Fowles, A.P., 2007. Livestock trampling reduces the conservation value of beetle communities on high quality exposed riverine sediments. Biodiv. Conserv. 16, 1491–1509. Belsky, A.J., Matzke, A., Uselman, S., 1999. Survey of livestock influences on stream and riparian ecosystems in the western United States. J. Soil Water Conserv. 54, 419–431. Boscaini, A., Franceschini, A., Maiolini, B., 2000. River ecotones: carabid beetles as a tool for quality assessment. Hydrobiologia 422, 173–181. Boulton, A.J., Stanley, E.H., 1995. Hyporheic processes during flooding and drying in a Sonoran Desert stream: II Faunal dynamics. Arch. fuer Hydrobiol. 134, 27–52. Bromham, L., Cardillo, M., Bennett, A.F., Elgar, M.A., 1999. Effects of stock grazing on the ground invertebrate fauna of woodland remnants. Aust. J. Ecol. 24, 199–207. Brooks, A.J., Haeusler, T., 2016. Invertebrate responses to flow: trait-velocity relationships during low and moderate flows. Hydrobiologia 773, 23–34. CSIRO Division of Entomology, 1991. 2nd edition. The Insects of Australia: A Textbook for Students and Research Workers Vols. 1 and 2 Melbourne University Press, Carlton, Victoria. Cade, B.S., Noon, B.R., 2003. A gentle introduction to quantile regression for ecologists. Front. Ecol. Environ. 1, 412–420. Chessman, B.C., 1995. Rapid assessment of rivers using macroinvertebrates: a procedure based on habitat-specific sampling, family level identification and a biotic index. Aust. J. Ecol. 20, 122–129. Chiu, M.C., Leigh, C., Mazor, R., Cid, N., Resh, V., 2017. Chapter 5.1: Anthropogenic threats to intermittent rivers and ephemeral streams. In: Datry, T., Bonada, N., Boulton, A. (Eds.), Intermittent Rivers and Ephemeral Streams: Ecology and Management. Elsevier. Choquenot, D., McIlroy, J., Korn, T., 1996. Managing Vertebrate Pests: Feral Pigs. Australian Government Printing Service, Canberra. Cluzeau, D., Binet, F., Vertes, F., Simon, J.C., Riviere, J.M., Trehen, P., 1992. Effects of intensive cattle trampling on soil-plant-earthworms system in two grassland types. Soil Biol. Biochem. 24, 1661–1665. Corti, R., Datry, T., 2016. Terrestrial and aquatic invertebrates in the riverbed of an intermittent river: parallels and contrasts in community organisation. Freshw. Biol. 61, 1308–1320. Corti, R., Larned, S.T., Datry, T., 2013. A comparison of pitfall-trap and quadrat methods for sampling ground-dwelling invertebrates in dry riverbeds. Hydrobiologia 717, 13–26. DSITIA (Department of Science, Information Technology, Innovation and the Arts), 2013. Operations Support Method: Addressing Reference Condition Selection Criteria (AEMA006). Science Delivery, Queensland Government, Brisbane (version 1.0). Datry, T., Larned, S.T., Tockner, K., 2014. Intermittent rivers: a challenge for freshwater ecology. Bioscience 64, 229–235. Dobbie, M.J., Stevens Jr., D.L., 2006. Data requirements for design of the SEAP central province pilot survey. CSIRO Mathematical and Information Sciences Commercial-inConfidence Technical Report 06/181. Dobbie, M.J., Negus, P., 2013. Addressing statistical and operational challenges in designing large-scale stream condition surveys. Environ. Monit. Assess. 185, 7231–7243. Eyre, M.D., Luff, M.L., 2002. The use of ground beetles (Coleoptera: carabidae) in conservation assessments of exposed riverine sediment habitats in Scotland and Northern England. J. Insect Conserv. 6, 25–38. Field, S.A., O’Connor, P.J., Tyre, A.J., Possingham, H.P., 2007. Making monitoring meaningful. Aust. Ecol. 32, 485–491. Fleischner, T.L., 1994. Ecological costs of livestock grazing in western North America. Conserv. Biol. 8, 629–644. Gentle, M., Pople, A., 2013. Effectiveness of commercial harvesting in controlling feralpig populations. Wildlife Res. 40, 459–469. Green, D.M., Kauffman, J.B., 1995. Succession and livestock grazing in a northeastern Oregon riparian ecosystem. J. Range Manage. 48, 307–313. Greenslade, P., 2007. The potential of Collembola to act as indicators of landscape stress in Australia. Anim. Prod. Sci. 47, 424–434. Guo, F., Bunn, S.E., Brett, M.T., Kainz, M.J., 2017. Polyunsaturated fatty acids in stream food webs–high dissimilarity among producers and consumers. Freshw. Biol. 62, 1325–1334. http://dx.doi.org/10.1111/fwb.12956. Hashimoto, K., Hayashi, F., 2012. Structure and function of the large pronotal horn of the sand-living anthicid beetle Mecynotarsus tenuipes. Entomol. Sci. 15, 274–279. Hawkins, C.P., Olson, J.R., Hill, R.A., 2010. The reference condition: predicting benchmarks for ecological and water-quality assessments. J. North Am. Benthol. Soc. 29, 312–343. Hoffmann, B.D., Andersen, A.N., 2003. Responses of ants to disturbance in Australia, with particular reference to functional groups. Aust. Ecol. 28, 444–464. Kauffman, J.B., Krueger, W.C., 1984. Livestock impacts on riparian ecosystems and streamside management implicationsl: a review. J. Range Manag. 430–438. Kauffman, J.B., Krueger, W.C., Vavra, M., 1983a. Impacts of cattle on streambanks in northeastern Oregon. J. Range Manag. 683–685. Kauffman, J.B., Krueger, W.C., Vavra, M., 1983b. Effects of late season cattle grazing on riparian plant communities. J. Range Manag. 685–691. Kennard, M.J., Pusey, B.J., Olden, J.D., MacKay, S.J., Stein, J.L., Marsh, N., 2010. Classification of natural flow regimes in Australia to support environmental flow management. Freshw. Biol. 55, 171–193. King, J.R., Andersen, A.N., Cutter, A.D., 1998. Ants as bioindicators of habitat disturbance: validation of the functional group model for Australia’s humid tropics.

4.4. Future monitoring and assessment of dry riverbeds It is time to introduce the assessment of dry sites into current and future river health monitoring and assessment programs (Steward et al., 2012; Acuña et al., 2014). Dry and wet indicators can then be combined to report on the health of an entire river network. We have provided an example of how this can be done. This can easily be adopted by other programs in intermittent rivers around the world. A key way forward is to incorporate dry riverbeds into ecosystem monitoring and assessment programs through government policy and legislation. We need to recognise dry riverbeds as important elements of intermittent rivers – as habitats in their own right. Acknowledgements Bill Senior, Dean Holloway, Delwyn Hansen, Glenn McGregor, James Fawcett, Jaye Lobegeiger, Joanna Blessing, John Bowlen, Penny Rogers, Ryan Woods and Tess Mullins from the Queensland Government and Ric Newson from South West Natural Resource Management (NRM) Ltd. are acknowledged for field work assistance, and we also acknowledge the landholders for permission to sample on their properties. We would like to thank Farah Zavahir for assistance with laboratory processing of samples, and Chris Burwell from the Queensland Museum for verifying some of the beetle and ant identifications. Funding for this project was provided by the Department of Science, Information Technology and Innovation in the Queensland Government, Australia. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ecolind.2017.10.053. References Acuña, V., Datry, T., Marshall, J., Barceló, D., Dahm, C.N., Ginebreda, A., McGregor, G., Sabater, S., Tockner, K., Palmer, M.A., 2014. Why should we care about temporary waterways? Science 343, 1080–1081. Agouridis, C.T., Edwards, D.R., Workman, S.R., Bicudo, J.R., Koostra, B.K., Vanzant, E.S., Taraba, J.L., 2005. Streambank erosion associated with grazing practices in the humid region. Trans. ASAE 48, 181–190. Andersen, A.N., Majer, J.D., 2004. Ants show the way Down Under: invertebrates as bioindicators in land management. Front. Ecol. Environ. 2, 291–298. Andersen, A.N., Hoffmann, B.D., Müller, W.J., Griffiths, A.D., 2002. Using ants as bioindicators in land management: simplifying assessment of ant community responses. J. Appl. Ecol. 39, 8–17. Andersen, A.N., Fisher, A., Hoffmann, B.D., Read, J.L., Richards, R., 2004. Use of terrestrial invertebrates for biodiversity monitoring in Australian rangelands, with particular reference to ants. Aust. Ecol. 29, 87–92. Andersen, A.N., 1997. Functional groups and patterns of organization in North American ant communities: a comparison with Australia. J. Biogeogr. 24, 433–460. Andersen, A.N., 1999. My bioindicator or yours? Making the selection. J. Insect Conserv. 3, 61–64. Apitz, S.E., Elliott, M., Fountain, M., Galloway, T.S., 2006. European environmental management: moving to an ecosystem approach. Integr. Environ. Assess. Manag. 2, 80–85. Barrios-Garcia, M.N., Classen, A.T., Simberloff, D., 2014. Disparate responses of aboveand belowground properties to soil disturbance by an invasive mammal. Ecosphere 5, 1–13. Barson, M., 2013. Land management practice trends in Queensland’s beef cattle/sheep industries. Caring for Our Country Sustainable Practices Fact Sheet 17. Department of Agriculture, Fisheries and Forestry, Canberra (Last Accessed 07 December 2016).

546

Ecological Indicators 85 (2018) 537–547

A.L. Steward et al.

2201–2220. Orr, H.K., 1960. Soil porosity and bulk density on grazed and protected Kentucky bluegrass range in the Black Hills. J. Range Manag. 80–86. Paetzold, A., Yoshimura, C., Tockner, K., 2008. Riparian arthropod responses to flow regulation and river channelization. J. Appl. Ecol. 45, 894–903. Palmer, M.A., Reidy Liermann, C.A., Nilsson, C., Flörke, M., Alcamo, J., Lake, P.S., Bond, N., 2008. Climate change and the world's river basins: anticipating management options. Front. Ecol. Environ. 6, 81–89. Peck, S.L., McQuaid, B., Campbell, C.L., 1998. Using ant species (Hymenoptera: formicidae) as a biological indicator of agroecosystem condition. Environ. Entomol. 27, 1102–1110. Pettit, N.E., 2002. Riparian vegetation of a permanent waterhole on cooper creek, southwest Queensland. R. Soc. Queensland 110, 15–25. Popolizio, C.A., Goetz, H., Chapman, P.L., 1994. Short-term response of riparian vegetation to 4 grazing treatments. J. Range Manag. 47, 48–53. R Core Team, 2013. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria(http://www.R-project.org /Accessed 3 November 2016). Rainio, J., Niemelä, J., 2003. Ground beetles (Coleoptera: carabidae) as bioindicators. Biodiv. Conserv. 12, 487–506. Reynoldson, T.B., Bailey, R., Day, K., Norris, R., 1995. Biological guidelines for freshwater sediment based on BEnthic Assessment of SedimenT (the BEAST) using a multivariate approach for predicting biological state. Aust. J. Ecol. 20, 198–219. Richards, C., Bacon, K.L., 1994. Influence of fine sediment on macroinvertebrate colonization of surface and hyporheic stream substrates. Western North Am. Natur. 54, 106–113. Robertson, A.I., Rowling, R.W., 2000. Effects of livestock on riparian zone vegetation in an Australian dryland river. Regul. Rivers: Res. Manag. 16, 527–541. Robson, B.J., Chester, E.T., Austin, C.M., 2011. Why life history information matters: drought refuges and macroinvertebrate persistence in non-perennial streams subject to a drier climate. Mar. Freshw. Res. 62, 801–810. Rogers, W., 1992. Quantile regression standard errors. Stata Tech. Bull. 9, 16–19. Sánchez-Montoya, M.M., Moleón, M., Sánchez‐Zapata, J.A., Tockner, K., 2016. Dry riverbeds: corridors for terrestrial vertebrates. Ecosphere 7. http://dx.doi.org/10.1002/ ecs2.1508. Schälchli, U., 1992. The clogging of coarse gravel river beds by fine sediment. Hydrobiologia 235, 189–197. Schepers, J.S., Francis, D.D., 1982. Chemical water quality of runoff from grazing land in Nebraska: I : influence of grazing livestock. J. Environ. Qual. 11, 351–354. Schon, N.L., Mackay, A.D., Yeates, G.W., Minor, M.A., 2010. Separating the effects of defoliation and dairy cow treading pressure on the abundance and diversity of soil invertebrates in pastures. Appl. Soil Ecol. 46, 209–221. Schulz, T.T., Leininger, W.C., 1990. Differences in riparian vegetation structure between grazed areas and exclosures. J. Range Manage. 43, 295–299. Sheldon, F., Bunn, S.E., Hughes, J.M., Arthington, A.H., Balcombe, S.R., Fellows, C.S., 2010. Ecological roles and threats to aquatic refugia in arid landscapes: dryland river waterholes. Mar. Freshw. Res. 61, 885–895. Sheldon, F., 2005. Incorporating natural variability into the assessment of ecological health in Australian dryland rivers. Hydrobiologia 552, 45–56. Steward, A.L., Marshall, J.C., Sheldon, F., Harch, B., Choy, S., Bunn, S.E., Tockner, K., 2011. Terrestrial invertebrates of dry river beds are not simply subsets of riparian assemblages. Aquat. Sci. 73, 551–566. Steward, A.L., von Schiller, D., Tockner, K., Marshall, J.C., Bunn, S.E., 2012. When the river runs dry: human and ecological values of dry riverbeds. Front. Ecol. Environ. 10, 202–209. Steward, A.L., Langhans, S., Corti, R., Datry, T., 2017. Chapter 4.4: The biota of intermittent rivers and ephemeral streams – Terrestrial and semi-aquatic invertebrates. In: Datry, T., Bonada, N., Boulton, A. (Eds.), Intermittent Rivers and Ephemeral Streams: Ecology and Management. Elsevier. Steward, A., 2014. Mecynotarsus: a genus of odd-looking, but interesting beetles. News Bull. Queensland Entomol. Soc. 41, 177–178. Stoddard, J.L., Larsen, D.P., Hawkins, C.P., Johnson, R.K., Norris, R.H., 2006. Setting expectations for the ecological condition of streams: the concept of reference condition. Ecol. Appl. 16, 1267–1276. Tansey, M.R., Brock, T.D., 1972. The upper temperature limit for eukaryotic organisms. Proc. Natl. Acad. Sci. USA 69, 2426. Taylor, F.R., Gillman, L.A., Pedretti, J.W., 1989. Impact of cattle on two isolated fish populations in Pahranagat Valley Nevada. Great Basin Natur. 49, 491–495. Trimble, S.W., Mendel, A.C., 1995. The cow as a geomorphic agent—a critical review. Geomorphology 13, 233–253. Vtorov, I.P., 1993. Feral pig removal: effects on soil microarthropods in a Hawaiian rain forest. J. Wildl. Manag. 57, 875–880. Williams, G., 1959. The seasonal and diurnal activity of the fauna sampled by pitfall traps in different habitats. J. Anim. Ecol. 28, 1–13. Williams, W.D., 1988. Limnological imbalances: an antipodean viewpoint. Freshw. Biol. 20, 407–420. Wishart, M.J., 2000. The terrestrial invertebrate fauna of a temporary stream in southern Africa. African Zool. 35, 193–200. Wood, P.J., Armitage, P.D., 1997. Biological effects of fine sediment in the lotic environment. Environ. Manag. 21, 203–217.

Biodiv. Conserv. 7, 1627–1638. Kleinwächter, M., Rickfelder, T., 2007. Habitat models for a riparian carabid beetle: their validity and applicability in the evaluation of river bank management. Biodiv. Conserv. 16, 3067–3081. Koenker, R., 2016. Quantreg: Quantile Regression R Package Version 5.26. (Accessed 3/ 11/2016). https://CRAN.R-project.org/package=quantreg. Lalley, J.S., Viles, H.A., Henschel, J.R., Lalley, V., 2006. Lichen-dominated soil crusts as arthropod habitat in warm deserts. J. Arid Environ. 67, 579–593. Larned, S.T., Datry, T., Robinson, C.T., 2007. Invertebrate and microbial responses to inundation in an ephemeral river reach in New Zealand: effects of preceding dry periods. Aquat. Sci. 69, 554–567. Larned, S.T., Datry, T., Arscott, D.B., Tockner, K., 2010. Emerging concepts in temporaryriver ecology. Freshw. Biol. 55, 717–738. Leach, E., Nakamura, A., Turco, F., Burwell, C.J., Catterall, C.P., Kitching, R.L., 2013. Potential of ants and beetles as indicators of rainforest restoration: characterising pasture and rainforest remnants as reference habitats. Ecol. Manag. Restor. 14, 202–209. Leigh, C., Stubbington, R., Sheldon, F., Boulton, A.J., 2013. Hyporheic invertebrates as bioindicators of ecological health in temporary rivers: a meta-analysis. Ecol. Indic. 32, 62–73. Lindsay, E.A., Cunningham, S.A., 2009. Livestock grazing exclusion and microhabitat variation affect invertebrates and litter decomposition rates in woodland remnants. For. Ecol. Manag. 258, 178–187. Lowrance, R., Todd, R., Fail, J., Hendrickson, O., Leonard, R., Asmussen, L., 1984. Riparian forests as nutrient filters in agricultural watersheds. Bioscience 34, 374–377. Ludwig, J., Tongway, D., 2002. Clearing savannas for use as rangelands in Queensland: altered landscapes and water-erosion processes. Rang. J. 24, 83–95. Majer, J.D., 1983. Ants: bio-indicators of minesite rehabilitation, land-use, and land conservation. Environ. Manag. 7, 375–383. Marc, P., Canard, A., Ysnel, F., 1999. Spiders (Araneae) useful for pest limitation and bioindication. Agriculture. Ecosyst. Environ. 74, 229–273. Martin, T.G., McIntyre, S., 2007. Impacts of livestock grazing and tree clearing on birds of woodland and riparian habitats. Conserv. Biol. 21, 504–514. Mazor, R.D., Stein, E.D., Ode, P.R., Schiff, K., 2014. Integrating intermittent streams into watershed assessments: applicability of an index of biotic integrity. Freshw. Sci. 33, 459–474. McClain, M.E., Boyer, E.W., Dent, C.L., Gergel, S.E., Grimm, N.B., Groffman, P.M., Hart, S.C., Harvey, J.W., Johnston, C.A., Mayorga, E., McDowell, W.H., 2003. Biogeochemical hot spots and hot moments at the interface of terrestrial and aquatic ecosystems. Ecosystems 6, 301–312. Meehl, G.A., Stocker, T.F., Collins, W.D., Friedlingstein, P., Gaye, A.T., Gregory, J.M., Kitoh, A., Knutti, R., Murphy, J.M., Noda, A., Raper, S.C.B., Watterson, I.G., Weaver, A.J., Zhao, Z.-C., 2007. Global climate projections. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press, Cambridge/New York, pp. 747–845. Michaels, K.F., 2007. Using staphylinid and tenebrionid beetles as indicators of sustainable landscape management in Australia: a review. Anim. Prod. Sci. 47, 435–449. Mills, J.R., Turner, E.J., Caltabiano, T., 1989. Land degradation in south-western Queensland. Project Report Q0 89008. Queensland Department of Primary Industries, Brisbane. Mitchell, J., Dorney, W., Mayer, R., McIlroy, J., 2007a. Ecological impacts of feral pig diggings in north Queensland rainforests. Wildl. Res. 34, 603–608. Mitchell, J., Dorney, W., Mayer, R., McIlroy, J., 2007b. Spatial and temporal patterns of feral pig diggings in rainforests of north Queensland. Wildl. Res. 34, 597–602. Muenz, T.K., Golladay, S.W., Vellidis, G., Smith, L.L., 2006. Stream buffer effectiveness in an agriculturally influenced area, southwestern Georgia: responses of water quality, macroinvertebrates, and amphibians. J. Environ. Qual. 35, 1924–1938. Murphy, G.E.P., Romanuk, T.N., 2014. A meta-analysis of declines in local species richness from human disturbances. Ecol. Evol. 4, 91–103. NLWA (National Land and Water Resources Audit), 2002. Australians and Natural Resource Management 2002. NLWA, Australian Capital Territory Canberra. Naeth, M.A., Bailey, A.W., Pluth, D.J., Chanasyk, D., Hardin, R.T., 1991. Grazing impacts on litter and soil organic matter in mixed prairie and fescue grassland ecosystems of Alberta. J. Range Manage. 44, 7–12. Naiman, R.J., Décamps, H., 1997. The ecology of interfaces: Riparian zones. Annu. Rev. Ecol. Syst. 28, 621–658. Negus, P., Blessing, J., Clifford, S., Steward, A., 2013. Riverine Assessment in Queensland's Lake Eyre and Bulloo Catchments: Stream and Estuary Assessment Program 2012. Department of Science, Information Technology, Innovation and the Arts Queensland Government, Brisbane. Negus, P., Blessing, J., Clifford, S., Steward, A., 2015. Riverine Assessment in the Warrego, Paroo, Bulloo and Nebine Catchments. Q–catchments: Technical Report 2012. Department of Science, Information Technology and Innovation Queensland Government, Brisbane. Norris, R.H., Linke, S., Prosser, I., Young, W.J., Liston, P., Bauer, N., Sloane, N., Dyer, F., Thoms, M., 2007. Very-broad-scale assessment of human impacts on river condition. Freshw. Biol. 52, 959–976. Obrist, M.K., Duelli, P., 2010. Rapid biodiversity assessment of arthropods for monitoring average local species richness and related ecosystem services. Biodiv. Conserv. 19,

547