Biological Conservation 197 (2016) 80–89
Contents lists available at ScienceDirect
Biological Conservation journal homepage: www.elsevier.com/locate/bioc
Prioritizing management actions for the conservation of freshwater biodiversity under changing climate and land-cover Chrystal S. Mantyka-Pringle a,b,c,d,⁎, Tara G. Martin b,c, David B. Moffatt e, James Udy f, Jon Olley g, Nina Saxton g, Fran Sheldon g, Stuart E. Bunn g, Jonathan R. Rhodes a,b a
The University of Queensland, School of Geography, Planning and Environmental Management, Brisbane, Qld 4072, Australia Australian Research Council Centre of Excellence for Environmental Decisions, The University of Queensland, Brisbane, Qld 4072, Australia Commonwealth Scientific and Industrial Research Organization, GPO Box 2583, Brisbane, Qld 4102, Australia d The University of Saskatchewan, Global Institute for Water Security, School of Environment and Sustainability, Saskatoon, SK S7N 5B3, Canada e Environmental Monitoring & Assessment Science, Department of Science, Information Technology, Innovation and the Arts, GPO Box 5078, Brisbane, Qld 4001, Australia f Healthy Waterways, PO Box 13086, Brisbane, Qld 4003, Australia g Australian Rivers Institute, Griffith University, Nathan, Qld 4111, Australia b c
a r t i c l e
i n f o
Article history: Received 1 January 2016 Accepted 29 February 2016 Available online xxxx Keywords: Freshwater conservation planning Management actions Costs Bayesian decision network Climate change Land-cover change
a b s t r a c t Freshwater ecosystems are declining under climate change and land-use change. To maximize the return on investment in freshwater conservation with limited financial resources, managers must prioritize management actions that are most cost-effective. However, little is known about what these priorities may be under the combined effects of climate and land-cover change. We present a novel decision-making framework for prioritizing conservation resources to different management actions for the conservation of freshwater biodiversity. The approach is novel in that it has the ability to model interactions, rank management options for dealing with conservation threats from climate and land-cover change, and integrate empirical data with expert knowledge. We illustrate the approach using a case study in South East Queensland (SEQ), Australia under climate change, land-cover change and their combined effects. Our results show that the explicit inclusion of multiple threats and costs results in quite different priorities than when costs and interactions are ignored. When costs are not considered, stream and riparian restoration, as a single management strategy, provides the greatest overall protection of macroinvertebrate and fish richness in rural and urban areas of SEQ in response to climate change and/or urban growth. Whereas, when costs are considered, farm/land management with stream and riparian restoration are the most cost-effective strategies for macroinvertebrate and fish conservation. Our findings support riparian restoration as the most effective adaptation strategy to climate change and urban development, but because it is expensive it may often not be the most cost-efficient strategy. Our approach allows for these decisions to be evaluated explicitly. © 2016 Elsevier Ltd. All rights reserved.
1. Introduction Pressures on ecosystems worldwide outpace current resources available for their management. As a result, prioritization of resources is necessary to maximize the benefits of conservation outcomes (Bottrill et al., 2008; Joseph et al., 2009). A common mistake in priority setting for conservation is the focus on prioritizing species, habitats, or locations rather than management actions (Game et al., 2013). However, it is the actions aimed at saving species and habitats that use the resources of conservation agencies, not the species or habitats themselves. Encouragingly, progress is being made in the development of prioritization assessments that account for the costs of actions, their benefits, and likelihood of success (Carwardine et al., 2012; Hermoso et al., 2012; ⁎ Corresponding author. E-mail address:
[email protected] (C.S. Mantyka-Pringle).
http://dx.doi.org/10.1016/j.biocon.2016.02.033 0006-3207/© 2016 Elsevier Ltd. All rights reserved.
Turak and Linke, 2010; Wilson et al., 2007). Another shortcoming of current prioritizations is that few studies account for multiple interacting threats when prioritizing management actions for conservation (Evans et al., 2011; Fuentes et al., 2014; Klein et al., 2010). The consequence of ignoring interacting threats and future change is that we may under or overestimate the benefits of abating a single threat (Auerbach et al., 2015; Evans et al., 2011; Mantyka-Pringle et al., 2015). Furthermore, the cost of abating more than one threat at a time may not be additive (Crain et al., 2008). Given the varying costs and benefits of different management actions, this has important implications for deciding upon the most appropriate actions to take. The prioritization of management actions and resources for the conservation of biodiversity within freshwater ecosystems has never been so essential. Due to intensive human use freshwaters are among the most seriously threatened and modified environments on the planet (Vörösmarty et al., 2010). Key disturbances, such as water extraction,
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
dams, invasive species, over-harvesting of fish, pollution, and modifications to riparian and in-stream habitats have heavily altered freshwater ecosystems and continue unabated across the globe (Vörösmarty et al., 2010). In addition, there is evidence that climate change interacts with land-use change to affect runoff, river flow regimes, water temperature, evaporation rates, and in turn biodiversity (e.g. Anteau, 2012; Nelson et al., 2009; Peterson and Kwak, 1999; Porter et al., 2013). Yet, conservation planning efforts in freshwater environments have been few compared to terrestrial and marine environments with only a handful of studies prioritizing actions for the conservation of freshwater biodiversity whilst minimizing costs (Januchowski-Hartley et al., 2011; Moilanen et al., 2011; Stewart-Koster et al., 2010; Ticehurst et al., 2007). Further, no freshwater decision framework has accounted for interactions between stressors. One approach for dealing with interactions between stressors is Bayesian Networks (BNs). BNs are probabilistic models that represent conditional dependencies between nodes in a directed acyclic graph (Kjaerulff and Madsen, 2008). The nodes represent variables that affect some outcome of interest and the links represent interactions between the nodes (Jensen, 1996). Underlying each dependent node is a conditional probability table (CPT) that specifies the probability of each state conditional on other variables (Marcot et al., 2006). BNs are better suited than other techniques/models for situations where considerable uncertainty exists because of the relative ease of combining qualitative (or subjective) and quantitative data (e.g. Ban et al., 2014; Smith et al., 2007). Data feeding into BNs can be based on expert judgment through an expert elicitation process and/or empirical or modeled data about the relationships of interest (Martin et al., 2015). When the BN structure has been fully specified, and the CPTs are parameterized, the model can be used for predictive reasoning about uncertain systems (e.g. Ban et al., 2015; Mantyka-Pringle et al., 2014). For decision-making processes, BNs can also be modified to incorporate the relative costs and benefits of management actions. Such models are known as Bayesian decision networks (BDNs) and are used to model the most appropriate decision given estimated costs and benefits (e.g. Ticehurst et al., 2007). Integration of this information with interactions into decision making is a key gap that needs to be addressed for successful and effective conservation. To address this important issue we develop an approach for examining how climate change and land-use change determine conservation priorities for conserving freshwater biodiversity under future climate and land-cover change scenarios. We use a BDN to illustrate how to prioritize freshwater conservation and rehabilitation management actions for protecting freshwater macroinvertebrates and fish richness (i.e. stream and riparian restoration, farm/land management, restoration of natural flow, and best practice erosion control) and apply it to rural and urban areas of South East Queensland (SEQ), Australia. While BNs are generally used to depict causal reasoning, they are also often used to represent the correlative structure between variables. The BDN model in this paper contains a mix of both causal and correlative structures. BDNs are particularly useful for dealing with interacting stressors and this is the first use of BDNs for prioritizing freshwater management actions whilst accounting for multiple interacting stressors and future global change. 2. Materials and methods 2.1. South East Queensland study region SEQ covers an area of nearly 23,000 km2 with 15 major river catchments and numerous sub-catchments (= watersheds) (Abal et al., 2005). The rivers and streams of SEQ are under increasing pressure from agricultural activities and intensive urbanization, which places pressure on the receiving waters of Moreton Bay — an area of high conservation value (Abal et al., 2005). SEQ is the fastest growing region in Australia, with 754,000 new dwellings expected to be developed by 2031 to accommodate population growth (OUM, 2009). SEQ has only
81
25% of its native vegetation remaining and predicted increases in the number of dwellings are therefore likely to cause further impacts on native habitat and the ecological health of its waterways. In 2002, the Queensland Government established an Ecosystem Health Monitoring Program (EHMP) in SEQ to assess the effectiveness of management and planning activities aimed at improving SEQ's waterways in the face of global change (Bunn et al., 2010). Up until 2014, the EHMP involved the bi-annual assessment of 135 freshwater sites throughout SEQ, classified by stream order (one to eighth orders), stream type (upland, coastal, lowland) and land-use (e.g. urban, cropping) (see Bunn et al., 2010) (Fig. 1), and reported on five ecological indicators encompassing eighteen separate indices (EHMP, 2012). Data on water quality (including nutrients), aquatic macroinvertebrates and fish from the 135 EHMP sites were used in deriving the BDN presented here so our outputs are directly relevant to real decision-making. 2.2. Management actions We reviewed local, regional and state management plans and scientific literature to build an understanding of the management actions that could be used as rehabilitation and adaptation strategies for freshwater biodiversity conservation in SEQ under a changing climate. The BDN framework and potential actions, developed during the review phase, were presented to key stakeholders in SEQ for discussion. Stakeholders provided feedback on the management actions to ensure the actions and outputs investigated were realistic and appropriate for local management needs. The decision nodes described in Table 1 represent a summary of the most practical management actions that could be investigated in SEQ. 2.3. Bayesian decision network We used a validated BN that identified the major causal links between climate (i.e. air temperature, precipitation and rainfall variability) and land-cover (i.e. the amount of hard impervious surfaces and the amount of riparian vegetation) on freshwater biodiversity (i.e. macroinvertebrate taxa richness and native fish species richness) in SEQ as a baseline (Mantyka-Pringle et al., 2014; see Fig. 2 for a conceptual model). The BN included nitrogen, phosphorus, volume of runoff and water temperature as variables in the model, because they are among the most important drivers of freshwater biodiversity loss (linked to land-cover change) identified in the literature (see Appendix A for a review) and represent some of the greatest environmental changes expected to occur in the study region (e.g. increased urban development, vegetation clearing and rising temperatures). A ‘nutrient’ variable was included to represent the effect between higher nitrogen, phosphorus, runoff and rainfall variability caused by climate and land-use change (i.e. nutrient load; as rainfall becomes more variable the nutrient load is greater, see Appendix A). Elevation was also included because it is an important natural determinant of macroinvertebrate and fish distributions in SEQ and elsewhere (see Appendix A for an overview of the conceptual model and the relationships/links between the nodes). Macroinvertebrate taxa richness and fish species richness were chosen as indices based on their statistically strong association with the disturbance gradient in this study region (Bunn et al., 2010) and because they are generally sensitive to multiple stressors (e.g. Statzner and Bêche, 2010; Stendera et al., 2012). The spatial resolution of the network is the site and the extent is the SEQ region. With 75% of the 135 EHMP sites, the BN was updated to learn from the data while the remaining 25% of the 135 sites were used to test and validate the model. Prior to parameterization, all variables in the BN were categorized into states (classes) using the 33rd and 66th percentile values of each dataset and/or via consultation with freshwater scientists and managers who were familiar with the study region (see Appendix B for more details on these datasets). The BN was modified into a BDN by incorporating available management
82
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
Fig. 1. Map of South East Queensland, Australia, showing boundaries of the river catchments and distribution of the 135 Ecosystem Health Monitoring Program survey sites (urban sites = triangles; rural sites = circles).
actions (i.e. decision nodes), the cost of available actions (cost functions), and the benefit of the management outcomes (utility functions) using Netica Software (Norsys Software, 2008) (Fig. 2). We populated the CPTs of the BDN with informative priors (i.e. variables/nodes were generated using a case file, an output from a model or expert knowledge) as described in Mantyka-Pringle et al. (2014), with a few changes specified here. For riparian cover (% of riparian zone covered in vegetation), runoff (mean annual accumulated soil water surplus) and nutrients (nutrient load), stakeholders identified these three variables as being directly linked to the proposed management actions. We, therefore, populated the CPTs of these nodes with informative priors using a two-stage expert elicitation procedure as described by Martin et al. (2012). This is a common approach and best practice when empirical data are absent (e.g. Ban et al., 2014; Martin et al., 2015). First, we asked representatives from stakeholder groups to attend a workshop in May 2012 facilitated by co-authors, T.G.M. and C.M.-P. Four experts in freshwater ecology and management attended
(co-authors D.B.M., J.U., N.S., and F.S.). A facilitated group discussion about the management actions and the combination of variables in question was undertaken. Experts were asked to consider the effect of each management action on their dependent nodes considering all stated sub-actions as listed in Table 1 are implemented within a 500 m radius of each site. Experts were then asked to make independent estimates of the conditional probabilities for these three parameters (riparian cover, runoff and nutrients) for rural and urban areas in SEQ. The BDN was split into two networks, rural versus urban areas, because the stakeholders identified that the appropriateness and effectiveness of different management strategies would depend on the location within the catchment. The group then reconvened and discussed these independent judgements. Finally each expert was invited to revise their initial estimates (if desired) after viewing the estimates of the other experts. This method captures the benefits of group judgment while at the same time maintains independence (Martin et al., 2012). Revised individual estimates were then combined using statistical means with
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89 Table 1 Description of the decision nodes (i.e. management actions) selected for South East Queensland, Australia. Decision node
Description
Stream and riparian restoration
Fence off dams and riparian zones from livestock Sediment erosion control through stabilization of banks Undertake riparian rehabilitation/restoration through removal of weeds and replanting native vegetation Expand and rebuild floodplain areas in strategic locations Rotate livestock grazing to reduce the amount of bare earth exposed in paddocks/pastures Train landowners in safe burning practices to reduce surface runoff, erosion, and sediment transport Management of pesticides, fertilizers, and nitrogen and phosphorus inputs Expand and rebuild wetland areas Integration of water sensitive urban design at all scales (e.g. through policy development) Invest in new technology or add features to older systems (e.g. retrofitting storm water outlets) Enforce best practice in construction works/development
Farm/land management
Restore natural flow
Best practice erosion control
variance providing a measure of uncertainty around the estimates (Martin et al., 2012). Two additional experts were unable to attend the workshop (co-authors S.E.B. and J.O.), but participated in the survey externally in the presence of a facilitator (C.M.-P.). There were no large differences in the level of expertise among experts, so each expert was given an equal weighting in the calculation of a group mean. Cain's (2001) CPT calculator was then used to generate the full CPTs using the experts' averaged elicited probabilities. The CPT calculator works by reducing the number of scenarios in a CPT to key anchoring points which are then interpolated to complete the entire table. All other CPTs in the BDN were calculated based on empirical data using 75% of the EHMP sites (n = 100) and averaged with experts' elicited probabilities. The best-fit model was produced when the empirical CPTs were
83
given 75% weighting and the expert CPTs 25% (see Mantyka-Pringle et al., 2014). 2.4. Costs and utility estimates Experts on management costs (see Acknowledgments) provided relative costs for each management action on a scale of 0 to 1 (0 = no costs; 1 = high costs); the relative cost of implementing each management action at a single EHMP site (=500 m radius) in rural areas and then urban areas. Experts were asked to include costs of labor, materials, maintenance and any other management expenses required over the time period it would take to see an improvement in the overall condition of the freshwater ecosystem at a site. Prior to the elicitation, a checklist including training, labor, materials, maintenance and assessment examples were discussed to ensure that each expert considered the same cost components in their estimates. Using a relative scale provided a benchmark and helped avoid inconsistencies in logic as the elicitation progressed (Appendix C). Each expert was given an equal weighting in the calculation of mean costs with variance. Relative utility was also assigned on a standardized scale (0–1) to reflect the desirability of the management outcomes, and in doing so we assumed that the presence of relatively high levels of macroinvertebrate and native fish taxa richness are desirable, and low levels are detrimental, to ecosystem structure and function. Utility scores were derived so as to be linearly related to number of species (high utility ≥ 6 fish species and N24 macroinvertebrate taxa; low utility ≤ 4 fish species and b18 macroinvertebrate taxa). Equal weight was given to each fish species and macroinvertebrate taxa. 2.5. Analysis Prior to running the BDN, the 135 EHMP sites were divided into rural vs urban sites (Fig. 1) based on the proportion of their grid cells comprising urban land. We calculated the percent cover of urban land using an impervious cover (% of area that is urban or a road in 2009)
Fig. 2. A conceptual diagram of the Bayesian decision network (BDN) and the seven types of nodes used to evaluate alternative management actions in SEQ. This BDN has been modified from a Bayesian Belief Network presented in Mantyka-Pringle et al. (2014), to incorporate available management actions, the cost of available actions, and the benefit of the management outcomes. A solid arrow indicates a positive effect or link, whereas a closed circle indicates a negative effect or link. Links are based on the relationships as explained in Appendix A.
84
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
GIS layer (Stein et al., 2014) including foundation layers for the Australian Hydrological Geospatial Fabric (http://www.bom.gov.au/ water/geofabric/index.shtml) and extracted the proportion of impervious cover (1:250,000 polygon layer rescaled to a 1 km2 grid) at each EHMP site using bilinear interpolation. We classified ‘urban sites’ as N10% urban land (n = 35) and ‘rural sites’ as b 1% (n = 91). For the eight sites that fell in between ≤10% and ≥1% urban land, we inspected each site separately in Google Earth (http://www.google.com/earth) to decide which sites are likely to behave as rural (e.g. urban land is far removed from site) (n = 8) and those more likely to behave as urban (e.g. high concentration of urban land close to the stream network near the site) (n = 0). We compared our classifications with detailed maps of the region and in Google Earth alongside experts to make sure that our calculations were accurate and representative of the expert's views. Current conditions and six climate and land-cover scenarios were run through the BDN to evaluate the expected influence of the proposed management actions now and in 2070. The scenarios consisted of two climate impact scenarios based on moderate and high emissions (moderate and high climate), two urban growth scenarios (moderate and high growth) and two combination scenarios (moderate climate + moderate growth; high climate + high growth). Climate change scenarios were simulated (25 × 25 km2 resolution) using outputs from the CSIRO Mk3.5 General Circulation Model (A1B and A1FI 2070 emissions scenarios; www.csiro.au/ozclim). Urban growth scenarios were simulated (500 × 500 m2 resolution) using the large scale urban model to predict the growth distribution of households across SEQ (2.14% and 2.2% increase by 2071; Stimson et al., 2012). For more details on these datasets and scenarios used, see Appendices B and D of the online Supplementary material. To apply the BDN to all rural and urban sites using different scenarios and management actions, the model was run once for each combination of scenario, climate, action and site. We evaluated our BDN by conducting a scenario-based evaluation with experts through verbal consent using macroinvertebrate and native fish species richness per proposed management action and scenario as probes to check that model outputs (see Appendix E) were consistent with expert expectations. For each scenario, we selected the sites that were negatively impacted by future climate and/or land-cover change (i.e. sites that declined in fish species richness, macroinvertebrate taxa richness, or both over time without management) as cases where mitigation is necessary to prevent loss of freshwater biodiversity. We used Friedman tests and Friedman a posteriori multiple comparison tests on results for those sites to test differences in the mean change in macroinvertebrate and fish richness over time (the difference between richness under future scenarios with proposed management actions and richness under current climatic and/or land-cover conditions) within rural and urban areas. We then prioritized management actions according to their associated cost and utility (return on investment) by dividing the benefit of each management action (the difference between future or current richness of fish and/or macroinvertebrates with proposed management actions and the future or current richness of fish and/or macroinvertebrates without management) by the relative cost of that management action in response to climate change, urban growth or climate change and urban growth together. 3. Results 3.1. Costs of management actions Experts identified stream and riparian restoration as the most expensive management option in SEQ, having the greatest relative cost of 1.00 in urban sites and 0.75 in rural sites (Fig. 3). The least expensive management option identified, apart from doing nothing, was farm/ land management, with the smallest relative cost of 0.08 and 0.12 in urban and rural sites respectively. The relative costs of restoring natural
Fig. 3. Cost of management options elicited from managers. Relative costs are standardized from 0 to 1 and per site (=500 m radius). For an explanation of each management action see Table 1. The differences in mean costs of management actions were not statistically analyzed because they were based on relative estimates and elicited from a small number of experts (n = 3).
flow and best practice erosion control both had intermediate costs, ranging between 0.16 and 0.56 for rural and urban sites. Relative costs for restoring urban sites in SEQ were generally higher than rural sites. 3.2. Expected benefits of management actions Similar patterns were found in the richness of macroinvertebrates and fish between the two climate impact scenarios, between the two urban growth scenarios, and between the two combination scenarios (Friedman test; Appendix D). All subsequent analyses were therefore based on only the moderate climate, moderate growth, and moderate climate + moderate growth impact scenarios. Our model predicted a significant decline in the richness of fish and macroinvertebrates in response to climate change, urban growth, and climate and urban growth together at 6–13% (for fish) and 45–52% (for macroinvertebrates) of the 99 rural sites (Friedman tests; Fig. 4a, c). At negatively impacted rural sites, we found no significant differences in the future richness of fish or macroinvertebrates with or without best practice erosion control, restoration of natural flow or farm/ land management under all three scenarios. Under the climate change scenario and the climate change + urban growth scenario, we found fish and macroinvertebrate richness to significantly increase with stream and riparian restoration. Under the urban growth scenario alone, we found only macroinvertebrate richness to significantly increase with stream and riparian restoration. The future richness of fish and macroinvertebrates with stream and riparian restoration did not significantly differ from that of farm/land management + stream and riparian restoration or when all four management actions are applied together in response to the three scenarios. Our model predicted a significant decline in the richness of macroinvertebrates at 46–51% of the 35 urban sites in response to climate change, urban growth, and climate and urban growth together (Friedman test; Fig. 4d). Apart from best practice erosion control under the climate change scenario, the future richness of macroinvertebrates at negatively impacted urban sites did not significantly differ with or without best practice erosion control, the restoration of natural flow or farm/ land management. Under all three scenarios, macroinvertebrate richness significantly increased with stream and riparian restoration, and increased even more with farm/land management + stream and riparian restoration and when all four management actions are applied
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
85
Fig. 4. Mean change in fish species richness (a, b) and macroinvertebrate taxa richness (c, d) over time in rural (a, c) and urban sites (b, d) of SEQ in response to climate change (black), urban growth (light gray), and climate change + urban growth (dark gray). Bars represent mean ± 1SE (based on only those sites that declined in richness with no management action) and were calculated as the difference between richness under future scenarios with proposed management actions and richness under current climatic and/or land-cover conditions. Analysis is by Friedman test followed by Friedman multiple comparisons test. Letters indicate homogenous subgroups. There are no error bars for fish in urban sites under the moderate climate change scenario (b) because only one study site declined in richness over time.
together. For fish in urban sites (Fig. 4b), there were too few sample sites (3–11%) to detect any significant differences among management actions or scenarios (Friedman test; Fig. 4b). 3.3. Integrating costs and benefits When we prioritized management actions according to their associated costs and benefits, we found that the ranking of management actions was dependent upon the climate and land-cover change scenario considered and location of the sites (rural vs urban areas) (Fig. 5). For fish (Fig. 5a, d) and fish and macroinvertebrates together (Fig. 5c, f), most management actions were more cost-effective under current conditions than under any of the climate change and/or urban growth scenarios. At rural sites under current conditions, climate change, and climate change + urban growth, farm/land management was ranked as the most cost-effective management action for the conservation of fish, and fish and macroinvertebrates together (Fig. 5a, c). Best practice
erosion control followed closely behind. Under urban growth, best practice erosion control was ranked highest for fish, and farm/land management + stream and riparian restoration was ranked highest for fish and macroinvertebrates together, but was only slightly higher than stream and riparian restoration alone. For the conservation of macroinvertebrates under climate change + urban growth, farm/land management was the highest ranked management action, but under all other scenarios, farm/land management + stream and riparian restoration was ranked highest (Fig. 5b). At urban sites under current conditions, farm/land management was ranked as the most cost-effective management action for the conservation of fish, macroinvertebrates, and fish and macroinvertebrates together (Fig. 5d–f). Under climate change, farm/land management was ranked as the most cost-effective management action for the conservation of macroinvertebrates and best practice erosion control was ranked highest for fish and macroinvertebrates together. Under urban growth, farm/land management + stream and riparian restoration was ranked
86
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
Fig. 5. Return on investment for each management action in response to current conditions (black), climate change (light gray), urban growth (dark gray) and climate change + urban growth (white) in rural (a–c) and urban sites (d–f) of SEQ. Bars represent the benefit of improved richness of fish (a, d), macroinvertebrates (b, e), or fish and macroinvertebrates (c, f) divided by the relative cost of each management action. * indicates the highest return on investment for each scenario. * is absent for fish in urban sites in response to climate change because each management action provided a negative return on investment. The return on investments for fish in urban sites in general (d) should be treated with caution due to their small sample size (see Fig. 4b).
highest for fish, macroinvertebrates, and fish and macroinvertebrates together, but again was only slightly higher than stream and riparian restoration alone. Under climate change + urban growth, farm/land
management was ranked highest for macroinvertebrates and farm/ land management + stream and riparian restoration was ranked highest for fish and macroinvertebrates together.
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
4. Discussion and conclusions A novel and important contribution of this work is using an environmental model for multiple actions and for multiple interacting threats that also accounts for the cost-effectiveness of actions for freshwater biodiversity conservation. Further, our decision framework provides an innovative process to help managers prioritize the allocation of limited resources to help mitigate the impacts of climate change and land-use change on aquatic ecosystems. Throughout the world, there is a rapidly growing drive to restore freshwater environments that are either heavily degraded or faced with uncertainty through practices such as riparian buffer plantings, wetland creation, bank stabilization and riparian fencing. This is because such practices can reduce terrestrial runoff, filtering nutrients and sediment, offsetting maximum water temperatures, and providing important food sources and freshwater refuges for stream communities (e.g. Collins et al., 2013; Stranko et al., 2012). However, when multiple stressors interact, the effects are likely to vary considerably depending on the location and composition of the ecosystem, and management options need to be considered in this context (Ban et al., 2014; Mantyka-Pringle et al., 2015). Our research shows that the ranking of conservation strategies depends on whether multiple interacting threats (i.e. climate and land-cover change) are considered or not. In particular, when multiple threats and costs are considered, farm/land management + stream riparian restoration was found to be the most cost-effective management action for conserving freshwater macroinvertebrate and fish richness in SEQ. Whereas, when costs are not considered, stream and riparian restoration as a single management action can provide the greatest overall protection of macroinvertebrate and fish richness in SEQ in response to climate change and/or urban growth. Stream and riparian restoration is a practical response to climate and land-cover change (Mantyka-Pringle et al., 2014; Sheldon et al., 2012). The effect of stream and riparian restoration on macroinvertebrates and fish, however, has not always proved successful (e.g. Collins et al., 2013; Stranko et al., 2012). Some studies have found other factors such as in-stream barriers, dispersal abilities and the distance to potential source populations (Smith, 2009), the type of substrate (Collins et al., 2013), upstream land-use conditions (Lorenz and Feld, 2013), and the level of site degradation (Stranko et al., 2012) play an important role in determining diversity. These studies highlight the importance of understanding the characteristics of a site in order to maximize the benefits of restoration efforts for macroinvertebrate and fish conservation. The integration of land-use intensity, land/soil capability and the ecological and life history requirements of species into conservation prioritization methods and software will aid in this endeavor (e.g. Linke et al., 2007; Moilanen et al., 2008). Although stream and riparian restoration may provide greater overall benefits, it is both time consuming and expensive. Managers identified that the cost of fencing for stream and riparian restoration can vary from anywhere between AU$10,000–$15,000/km over most conditions. Revegetation can cost around AU$30,000–$40,000/ha. Stabilization of banks through bank/bed armoring and the construction of chutes/stepped weirs to convey runoff can also be very costly (e.g. a typical 10 m long, 6 m wide rock chute can cost AU$10,000–$15,000). The cost of farm/land management on the other hand, is cheap in comparison. In many Western countries, current government policy relies heavily on voluntary arrangements, education and information as the main policy instruments through which to persuade farmers to adopt better environmental farm management (Gunningham, 2007). In the last two decades, however, there has been a need for increased subsidies with greater focus on a more targeted approach, because commercial farmers have more readily adopted practices that are financially beneficial than those that have a positive environmental impact (Cary and Roberts, 2011). With a continuing and growing demand for agricultural and ecosystem services in the twenty-first century, farmers and conservationists need to
87
reconsider the relationship between agriculture and conservation of biodiversity. For example, many diversified, sustainable, organic and agro-ecological farming systems support substantially greater biodiversity in both freshwater and terrestrial systems in comparison to conventional farming practices (e.g. Rizo-Patrón et al., 2013). Well-managed agricultural landscapes can also provide better protection against climate change and extreme weather events via improved soil quality, carbon sequestration, and water-holding capacity in surface soils (e.g. Scherr and McNeely, 2008). Thus, this study supports the need to reconcile agricultural production and other farm/land management practices with healthy ecosystems to coordinate landscape and policy action (Acharya, 2006; Scherr and McNeely, 2008). We identified farm/land management to be more cost-effective than stream and riparian restoration in rural areas of SEQ, but we would expect to see only a modest improvement in freshwater biodiversity without investing in stream and riparian restoration. This is because changing land-use practices on farms in the broader catchment can improve some water quality parameters (e.g. nutrients/pesticides/ sediments), but overall these will have only a modest effect on biodiversity — especially if the riparian land is degraded (Bunn et al., 1999; Olley et al., 2013). For example, the Queensland Government in Australia has invested AU$250 million to improve the water quality reaching the World Heritage Listed Great Barrier Reef (GBR) — of which $175 million is targeted towards best management practice programs for the sugar cane and grazing industries located in the upper catchments of North Queensland (The State of Queensland, 2013). The result so far has been a very small improvement in the overall condition of the inshore marine environment (The State of Queensland, 2014). Although improvements in marine conditions will take time and are dependent on local weather events (The State of Queensland, 2014), we argue that if the goal is to see a minor/modest improvement in water quality and freshwater biodiversity, then spending money on farm/land management would be the most cost-effective, but sustained action may not result in continuing improvement in biodiversity or water quality. If the goal is to actually protect biodiversity (and restore streams), then the cost to do so by farm/land management alone would be ongoing as one cannot achieve this without stream and riparian restoration (Olley et al., 2013; Sheldon et al., 2012). In the case of SEQ, Olley et al. (2013) estimated that to reduce sediment and phosphorus loads to Moreton Bay by 50% would involve rehabilitating 6000 km of the 24,000 km of degraded channel network to the same condition as the remnant woodland areas. Although this management approach remains to be tested in the SEQ region, studies elsewhere indicate that sediment inputs to streams from bank erosion and channel instability can reduce quickly and dramatically in response to riparian restoration (McKergow et al., 2003). We therefore suggest that conservation efforts should first focus on protecting areas where the riparian cover is in relatively good condition and then on re-vegetating the channel network (Holl and Aide, 2011). The BDN results for urban areas were not as clear as those for rural areas. Some management actions showed a negative return on investment, which may be a result of the relatively small number of urban sites for which fish data were available. We therefore interpret these findings cautiously. Urban streams within Australia are generally located downstream of farmland, so altered farm/land management can improve water quality and freshwater biodiversity at a broad catchment scale, but at an urban site-specific scale, the practicality and impact of farm/land management declines. Instead, stream and riparian restoration via protection and revegetation of riparian areas may be more effective, but only possible if the infrastructure is completely removed at significant cost. High property values and finely subdivided land and dense human infrastructure (e.g. roads, sewer lines) limit the spatial extent of urban river restoration options, while stormwaters and the associated sediment and pollutant loads may limit the potential for restoration projects to reverse degradation (Bernhardt and Palmer, 2007; Suren and McMurtrie, 2005). Improving conditions of highly
88
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89
degraded urban sites via farm/land management and/or stream and riparian restoration may therefore be impractical and costly. Yet, we recommend these two management actions as first priority because urban streams with intact riparian buffers generally exhibit higher biodiversity levels than those without (e.g. Wasson et al., 2010). It is important to acknowledge that in some cases, stream and riparian restoration are unlikely to result in substantial improvements in instream ecological condition because it does not match the scale of the degrading process (Bernhardt and Palmer, 2007; Suren and McMurtrie, 2005). The ecological benefits of improving physical habitat and water quality may be tempered by the persistent effects of altered streamflow, storm runoff or even invasive species. In these cases, other effects of urban development may also need to be addressed for successful restoration of urban stream biodiversity (Konrad and Booth, 2005). Managers should therefore assess each individual site first to see which management action(s) are applicable. Comprehensive and controlled scientific studies at the urban site-specific scale are also needed to evaluate the ecological impacts of specific management actions under various levels of degradation. Nonetheless, our study has verified that the current suite of management actions that are being investigated or implemented by decision-makers and the scientific community for improving water quality and aquatic biodiversity in degraded catchments (i.e. farm/land management and stream and riparian restoration; e.g. Del Tánago et al., 2012; Kay et al., 2012) can also be used as strategies for mitigating the negative effects of future climate change and land-cover change on freshwater systems (Davies, 2010). The relative cost of management actions did not change among the different scenarios, therefore the difference in cost-effectiveness rankings between the scenarios can be attributed to the difference in effectiveness of management actions. For example, if we are concerned with the conservation of fish under urban growth in rural areas of SEQ, our results indicate that undertaking best practice erosion control would be the most cost-effective management action. Whereas, under scenarios including current conditions, climate change, or climate change and urban growth, investing in farm/land management would be most cost-effective. Similarly, if we were concerned with the conservation of macroinvertebrates in urban areas under urban growth, then we are best off investing in farm/land management + stream riparian restoration. Whereas, under alternative scenarios (current conditions, climate change, or climate change and urban growth), we would have a greater conservation success rate if we invested in farm/land management upstream. By considering the interactions among stressors when making decisions about where to manage and which actions to take, the expected return on investment in management changes (Auerbach et al., 2015; Mantyka-Pringle et al., 2015). When acting under the constraint of a limited budget, it is therefore essential that interactions between stressors are considered, as it will make a substantial difference to the effectiveness of management strategies. One possible reason why we did not predict a significant decline in fish richness at many of our urban sites is because the management actions in our study design were designated at a site-specific scale. Sitespecific management can be ineffective in improving conditions for mobile taxa such as fish (Bernhardt and Palmer, 2007). On the other hand, effects on fish may simply be small irrespective of scale. In an ideal world when management options are not constrained by costs or land, prioritization of management actions should be conducted at a landscape scale with upstream–downstream connectivity and terrestrial–freshwater interactions integrated into the assessment in order to improve conservation adequacy (e.g. Beger et al., 2010; Rivers-Moore et al., 2010). There are a number of caveats to bear in mind when interpreting our results. In selecting management actions for our BDN, we did not aim to address the cultural, ethical, socio-economic or spatial components necessary for an implementation plan. This study did not consider the effectiveness or feasibility of current or future management delivery models
(Carwardine et al., 2012). For some of the management actions, costs were uncertain and real costs may prove to be higher or lower than estimated. However, because cost effective priorities will depend most strongly on relative rather than absolute costs, we believe that this would not unduly influence the identified priorities. The costeffectiveness of actions for achieving increased richness and other benefits will also vary depending upon the values of the manager and other objectives for the region. Nonetheless, we believe that our analysis is robust in terms of the relative cost-effectiveness of the management actions and the relative benefits. Our BDN can also be updated as improved information on the costs and/or benefits of management actions becomes available. Incorporating multiple interacting threats into the selection of management actions for the conservation of biodiversity is important if we are to avoid poor conservation outcomes (Ban et al., 2014; MantykaPringle et al., 2015). Integrating costs also allows us to achieve more for our conservation dollar and consequently make greater gains for conservation overall (Naidoo et al., 2006; Wilson et al., 2006). While it is inappropriate to take our management action rankings too literally, they provide an indication of which actions are the best and most cost-effective in terms of conserving freshwater biodiversity in areas where the negative effects of climate change and/or land-cover change are likely to be high. Provided that the trade-offs of each management action are carefully evaluated, we are confident that farm/land management and stream and riparian restoration in combination can strengthen freshwater conservation efforts in rural and urban areas of SEQ and elsewhere, even in the face of climate change and land-cover change and with limited resources. Farm/land management and stream and riparian restoration can also contribute to a range of other benefits such as the conservation of native plants and terrestrial communities, improved carbon sequestration, improved recreational activities, and the achievement of more sustainable agricultural production. Acknowledgments We thank the representatives from Healthy Waterways, the Australian Rivers Institute (ARI), and the Queensland Department of Science, Information Technology, Innovation and the Arts for their active participation and assistance in the project. We thank C. Smith (University of Queensland) for his support in developing the BDN. Special thanks to T. McKew and T. Costantini from SEQ Catchments and J.O. for providing cost data on management actions in SEQ. This project was funded in part by a Queensland Government Smart Futures PhD Scholarship (C.M.-P.) and an Australian Government Postgraduate Award (C.M.-P.). This research was also conducted with the support of funding from the SEQ Climate Adaptation Research Initiative, the Australian Government's National Environmental Research Program and the Australian Research Council Centre of Excellence for Environmental Decisions. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.biocon.2016.02.033. References Abal, E.G., Bunn, S.E., Dennison, W.C., 2005. Healthy Waterways, Healthy Catchments: Making the Connection in South East Queensland. Moreton Bay and Catchments Partnership, Brisbane, Queensland. Acharya, K.P., 2006. Linking trees on farms with biodiversity conservation in subsistence farming systems in Nepal. Biodivers. Conserv. 15, 631–646. Anteau, M.J., 2012. Do interactions of land use and climate affect productivity of waterbirds and prairie-pothole wetlands? Wetlands 32, 1–9. Auerbach, N.A., Wilson, K.A., Tulloch, A.I., Rhodes, J.R., Hanson, J.O., Possingham, H.P., 2015. Effects of Threat Management Interactions on Conservation Priorities Conservation Biology.
C.S. Mantyka-Pringle et al. / Biological Conservation 197 (2016) 80–89 Ban, S.S., Pressey, R.L., Graham, N.A.J., 2014. Assessing interactions of multiple stressors when data are limited: A Bayesian belief network applied to coral reefs. Glob. Environ. Chang. 27, 64–72. Ban, S.S., Pressey, R.L., Graham, N.A., 2015. Assessing the effectiveness of local management of coral reefs using expert opinion and spatial Bayesian modeling. PLoS One 10, e0135465. Beger, M., Grantham, H.S., Pressey, R.L., Wilson, K.A., Peterson, E.L., Dorfman, D., Mumby, P.J., Lourival, R., Brumbaugh, D.R., Possingham, H.P., 2010. Conservation planning for connectivity across marine, freshwater, and terrestrial realms. Biol. Conserv. 143, 565–575. Bernhardt, E.S., Palmer, M.A., 2007. Restoring streams in an urbanizing world. Freshw. Biol. 52, 738–751. Bottrill, M.C., Joseph, L.N., Carwardine, J., Bode, M., Cook, C., Game, E.T., Grantham, H., Kark, S., Linke, S., McDonald-Madden, E., Pressey, R.L., Walker, S., Wilson, K.A., Possingham, H.P., 2008. Is conservation triage just smart decision making? Trends Ecol. Evol. 23, 649–654. Bunn, S., Davies, P., Mosisch, T., 1999. Ecosystem measures of river health and their response to riparian and catchment degradation. Freshw. Biol. 41, 333–345. Bunn, S.E., Abal, E.G., Smith, M.J., Choy, S.C., Fellows, C.S., Harch, B.D., Kennard, M.J., Sheldon, F., 2010. Integration of science and monitoring of river ecosystem health to guide investments in catchment protection and rehabilitation. Freshw. Biol. 55, 223–240. Cain, J., 2001. Planning Improvements in Natural Resources Management: Guidelines for Using Bayesian Networks to Support the Planning and Management of Development Programmes in the Water Sector and Beyond. Centre for Ecology and Hydrology. Crowmarsh Gifford, Wallingford, UK. Carwardine, J., O'Connor, T., Legge, S., Mackey, B., Possingham, H.P., Martin, T.G., 2012. Prioritizing threat management for biodiversity conservation. Conserv. Lett. 5, 196–204. Cary, J., Roberts, A., 2011. The limitations of environmental management systems in Australian agriculture. J. Environ. Manag. 92, 878–885. Collins, K.E., Doscher, C., Rennie, H.G., Ross, J.G., 2013. The effectiveness of riparian ‘restoration’ on water quality — a case study of lowland streams in Canterbury, New Zealand. Restor. Ecol. 21, 40–48. Crain, C.M., Kroeker, K., Halpern, B.S., 2008. Interactive and cumulative effects of multiple human stressors in marine systems. Ecol. Lett. 11, 1304–1315. Davies, P.M., 2010. Climate change implications for river restoration in global biodiversity hotspots. Restor. Ecol. 18, 261–268. Del Tánago, M.G., De Jalón, D.G., Román, M., 2012. River restoration in Spain: theoretical and practical approach in the context of the European Water Framework Directive. Environ. Manag. 50, 123–139. EHMP, 2012. Ecosystem Health Monitoring Program 2011–12 Annual Technical Report. South East Queensland Healthy Waterways Partnership, Brisbane http://www.ehmp.org. Evans, M.C., Possingham, H.P., Wilson, K.A., 2011. What to do in the face of multiple threats? Incorporating dependencies within a return on investment framework for conservation. Divers. Distrib. 17, 437–450. Fuentes, M.M.P.B., Blackwood, J., Jones, B., Kim, M., Leis, B., Limpus, C.J., Marsh, H., Mitchell, J., Pouzols, F.M., Pressey, R.L., Visconti, P., 2014. A decision framework for prioritizing multiple management actions for threatened marine megafauna. Ecol. Appl. 25, 200–214. Game, E.T., Kareiva, P., Possingham, H.P., 2013. Six common mistakes in conservation priority setting. Conserv. Biol. 27, 480–485. Gunningham, N., 2007. Incentives to improve farm management: EMS, supply-chains and civil society. J. Environ. Manag. 82, 302–310. Hermoso, V., Pantus, F., Olley, J., Linke, S., Mugodo, J., Lea, P., 2012. Systematic planning for river rehabilitation: integrating multiple ecological and economic objectives in complex decisions. Freshw. Biol. 57, 1–9. Holl, K.D., Aide, T.M., 2011. When and where to actively restore ecosystems? For. Ecol. Manag. 261, 1558–1563. Januchowski-Hartley, S., Visconti, P., Pressey, R., 2011. A systematic approach for prioritizing multiple management actions for invasive species. Biol. Invasions 13, 1241–1253. Jensen, F.V., 1996. An Introduction to Bayesian Networks. UCL press, London. Joseph, L.N., Maloney, R.F., Possingham, H.P., 2009. Optimal allocation of resources among threatened species: a project prioritization protocol. Conserv. Biol. 23, 328–338. Kay, P., Grayson, R., Phillips, M., Stanley, K., Dodsworth, A., Hanson, A., Walker, A., Foulger, M., McDonnell, I., Taylor, S., 2012. The effectiveness of agricultural stewardship for improving water quality at the catchment scale: experiences from an NVZ and ECSFDI watershed. J. Hydrol. 422, 10–16. Kjaerulff, U.B., Madsen, A.L., 2008. Bayesian Networks and Influence Diagrams. Springer Science + Business Media 200, p. 114. Klein, C.J., Ban, N.C., Halpern, B.S., Beger, M., Game, E.T., Grantham, H.S., Green, A., Klein, T.J., Kininmonth, S., Treml, E., Wilson, K., Possingham, H.P., 2010. Prioritizing land and sea conservation investments to protect coral reefs. PLoS One 5, e12431. Konrad, C.P., Booth, D.B., 2005. Hydrologic changes in urban streams and their ecological significance. American Fisheries Society Symposium, pp. 157–177. Linke, S., Pressey, R.L., Bailey, R.C., Norris, R.H., 2007. Management options for river conservation planning: condition and conservation re-visited. Freshw. Biol. 52, 918–938. Lorenz, A.W., Feld, C.K., 2013. Upstream river morphology and riparian land use overrule local restoration effects on ecological status assessment. Hydrobiologia 704, 489–501. Mantyka-Pringle, C.S., Martin, T.G., Moffatt, D.B., Linke, S., Rhodes, J.R., 2014. Understanding and predicting the combined effects of climate change and land-use change on freshwater macroinvertebrates and fish. J. Appl. Ecol. 51, 572–581. Mantyka-Pringle, C.S., Visconti, P., Di Marco, M., Martin, T.G., Rondinini, C., Rhodes, J.R., 2015. Climate change modifies risk of global biodiversity loss due to land-cover change. Biol. Conserv. 187, 103–111. Marcot, B.G., Steventon, J.D., Sutherland, G.D., McCann, R.K., 2006. Guidelines for developing and updating Bayesian belief networks applied to ecological modeling and conservation. Can. J. For. Res. 36, 3063–3074. Martin, T.G., Burgman, M.A., Fidler, F., Kuhnert, P.M., Low-Choy, S., McBride, M., Mengersen, K., 2012. Eliciting expert knowledge in conservation science. Conserv. Biol. 26, 29–38.
89
Martin, T.G., Murphy, H., Liedloff, A., Thomas, C., Chadès, I., Cook, G., Fensham, R., McIvor, J., van Klinken, R.D., 2015. Buffel grass and climate change: a framework for projecting invasive species distributions when data are scarce. Biol. Invasions 1–14. McKergow, L.A., Weaver, D.M., Prosser, I.P., Grayson, R.B., Reed, A.E., 2003. Before and after riparian management: sediment and nutrient exports from a small agricultural catchment, Western Australia. J. Hydrol. 270, 253–272. Moilanen, A., Leathwick, J., Elith, J., 2008. A method for spatial freshwater conservation prioritization. Freshw. Biol. 53, 577–592. Moilanen, A., Leathwick, J.R., Quinn, J.M., 2011. Spatial prioritization of conservation management. Conserv. Lett. 4, 383–393. Naidoo, R., Balmford, A., Ferraro, P.J., Polasky, S., Ricketts, T.H., Rouget, M., 2006. Integrating economic costs into conservation planning. Trends Ecol. Evol. 21, 681–687. Nelson, K.C., Palmer, M.A., Pizzuto, J.E., Moglen, G.E., Angermeier, P.L., Hilderbrand, R.H., Dettinger, M., Hayhoe, K., 2009. Forecasting the combined effects of urbanization and climate change on stream ecosystems: from impacts to management options. J. Appl. Ecol. 46, 154–163. Olley, J., Burton, J., Hermoso, V., Smolders, K., McMahon, J., Thomson, B., Watkinson, A., 2013. Remnant vegetation, sediment and nutrient loads, and river rehabilitation in subtropical Australia. Hydrol. Process. 29, 2290–2300. Peterson, J.T., Kwak, T.J., 1999. Modeling the effects of land use and climate change on riverine smallmouth bass. Ecol. Appl. 9, 1391–1404. Porter, E.M., Bowman, W.D., Clark, C.M., Compton, J.E., Pardo, L.H., Soong, J.L., 2013. Interactive effects of anthropogenic nitrogen enrichment and climate change on terrestrial and aquatic biodiversity. Biogeochemistry 114, 93–120. Rivers-Moore, N.A., Goodman, P.S., Nel, J.L., 2010. Scale-based freshwater conservation planning: towards protecting freshwater biodiversity in KwaZulu-Natal, South Africa. Freshw. Biol. 56, 125–141. Rizo-Patrón, V.F., Kumar, A., McCoy Colton, M.B., Springer, M., Trama, F.A., 2013. Macroinvertebrate communities as bioindicators of water quality in conventional and organic irrigated rice fields in Guanacaste, Costa Rica. Ecol. Indic. 29, 68–78. Scherr, S.J., McNeely, J.A., 2008. Biodiversity conservation and agricultural sustainability: towards a new paradigm of ‘ecoagriculture’ landscapes. Philos. Trans. R. Soc. B 363, 477–494. Sheldon, F., Peterson, E.E., Boone, E.L., Sippel, S., Bunn, S.E., Harch, B.D., 2012. Identifying the spatial scale of land use that most strongly influences overall river ecosystem health score. Ecol. Appl. 22, 2188–2203. Smith, P.J., 2009. Genetic principles for freshwater restoration in New Zealand. N. Z. J. Mar. Freshw. Res. 43, 749–762. Smith, C.S., Howes, A.L., Price, B., McAlpine, C.A., 2007. Using a Bayesian belief network to predict suitable habitat of an endangered mammal — the Julia Creek dunnart (Sminthopsis douglasi). Biol. Conserv. 139, 333–347. Software, Norsys, 2008. Netica 4.08. Norsys Software, Vancouver, British Columbia. OUM (Office of Urban Management), 2009. South East Queensland Regional Plan 2009– 2031. OUM, Brisbane, Queensland, Australia. Statzner, B., Bêche, L.A., 2010. Can biological invertebrate traits resolve effects of multiple stressors on running water ecosystems? Freshw. Biol. 55, 80–119. Stein, J., Hutchinson, M., Stein, J., 2014. A new stream and nested catchment framework for Australia. Hydrol. Earth Syst. Sci. 18, 1917–1933. Stendera, S., Adrian, R., Bonada, N., Cañedo-Argüelles, M., Hugueny, B., Januschke, K., Pletterbauer, F., Hering, D., 2012. Drivers and stressors of freshwater biodiversity patterns across different ecosystems and scales: a review. Hydrobiologia 696, 1–28. Stewart-Koster, B., Bunn, S.E., MacKay, S.J., Poff, N.L., Naiman, R.J., Lake, P.S., 2010. The use of Bayesian networks to guide investments in flow and catchment restoration for impaired river ecosystems. Freshw. Biol. 55, 243–260. Stimson, R., Bell, M., Corcoran, J., Pullar, D., 2012. Using a large scale urban model to test planning scenarios in the Brisbane-South East Queensland region. Reg. Sci. Policy Pract. 4, 373–392. Stranko, S.A., Hilderbrand, R.H., Palmer, M.A., 2012. Comparing the fish and benthic macroinvertebrate diversity of restored urban streams to reference streams. Restor. Ecol. 20, 747–755. Suren, A.M., McMurtrie, S., 2005. Assessing the effectiveness of enhancement activities in urban streams: II. Responses of invertebrate communities. River Res. Appl. 21, 439–453. The State of Queensland, 2013. (Department of the Premier and Cabinet). Reef Water Quality Protection Plan Investment Strategy 2013–2018. Queensland Government, Brisbane, Australia. The State of Queensland, 2014. (Department of the Premier and Cabinet). Great Barrier Reef Report Card 2012 and 2013 Reef Water Quality Protection Plan. Queensland Government, Brisbane, Australia. Ticehurst, J.L., Newham, L.T.H., Rissik, D., Letcher, R.A., Jakeman, A.J., 2007. A Bayesian network approach for assessing the sustainability of coastal lakes in New South Wales, Australia. Environ. Model. Softw. 22, 1129–1139. Turak, E., Linke, S., 2010. Freshwater conservation planning: an introduction. Freshw. Biol. 56, 1–5. Vörösmarty, C.J., McIntyre, P., Gessner, M.O., Dudgeon, D., Prusevich, A., Green, P., Glidden, S., Bunn, S.E., Sullivan, C.A., Liermann, C.R., 2010. Global threats to human water security and river biodiversity. Nature 467, 555–561. Wasson, J.G., Villeneuve, B., Iital, A., Murray-Bligh, J., Dobiasova, M., Bacikova, S., Timm, H., Pella, H., Mengin, N., Chandesris, A., 2010. Large-scale relationships between basin and riparian land cover and the ecological status of European rivers. Freshw. Biol. 55, 1465–1482. Wilson, K.A., McBride, M.F., Bode, M., Possingham, H.P., 2006. Prioritizing global conservation efforts. Nature 440, 337–340. Wilson, K.A., Underwood, E.C., Morrison, S.A., Klausmeyer, K.R., Murdoch, W.W., Reyers, B., Wardell-Johnson, G., Marquet, P.A., Rundel, P.W., McBride, M.F., 2007. Conserving biodiversity efficiently: what to do, where, and when. PLoS Biol. 5, e223.
