Figures
Abstract
Management authorities seldom have the capacity to comprehensively address the full suite of anthropogenic stressors, particularly in the coastal zone where numerous threats can act simultaneously to impact reefs and other ecosystems. This situation requires tools to prioritise management interventions that result in optimum ecological outcomes under a set of constraints. Here we develop one such tool, introducing a Bayesian Belief Network to model the ecological condition of inshore coral reefs in Moreton Bay (Australia) under a range of management actions. Empirical field data was used to model a suite of possible ecological responses of coral reef assemblages to five key management actions both in the sea (e.g. expansion of reserves, mangrove & seagrass restoration, fishing restrictions) and on land (e.g. lower inputs of sediment and sewage from treatment plants). Models show that expanding marine reserves (a ‘marine action’) and reducing sediment inputs from the catchments (a ‘land action’) were the most effective investments to achieve a better status of reefs in the Bay, with both having been included in >58% of scenarios with positive outcomes, and >98% of the most effective (5th percentile) scenarios. Heightened fishing restrictions, restoring habitats, and reducing nutrient discharges from wastewater treatment plants have additional, albeit smaller effects. There was no evidence that combining individual management actions would consistently produce sizeable synergistic until after maximum investment on both marine reserves (i.e. increasing reserve extent from 31 to 62% of reefs) and sediments (i.e. rehabilitating 6350 km of waterways within catchments to reduce sediment loads by 50%) were implemented. The method presented here provides a useful tool to prioritize environmental actions in situations where multiple competing management interventions exist for coral reefs and in other systems subjected to multiple stressor from the land and the sea.
Citation: Gilby BL, Olds AD, Connolly RM, Stevens T, Henderson CJ, Maxwell PS, et al. (2016) Optimising Land-Sea Management for Inshore Coral Reefs. PLoS ONE 11(10): e0164934. https://doi.org/10.1371/journal.pone.0164934
Editor: Fabiano L. Thompson, Universidade Federal do Rio de Janeiro, BRAZIL
Received: June 14, 2016; Accepted: October 4, 2016; Published: October 20, 2016
Copyright: © 2016 Gilby et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data files are available from the Zenodo database (10.5281/zenodo.159508) at https://zenodo.org/record/159508#.V_cdtdywLRc.
Funding: Funded by Australian Rivers Institute and School of Environment at Griffith University. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
In coastal marine environments, urbanisation, habitat loss, fishing, sediments, and pollutant inputs (e.g. wastewater) degrade ecosystem condition [1, 2]. Judicious and efficient conservation strategies are, therefore, needed to maintain the condition and functioning of such systems [3], especially as funding limitations often restrict the capacity of management agencies to comprehensively address all threats. Achieving maximum benefit from management interventions requires that management actions be prioritised, preferably in a quantitative manner, with the management actions having greatest ecological benefit implemented first. Ideally, effective prioritisation of actions will result in net benefits that are greater than the sum of individual actions (i.e. synergistic effects) [4–6]. The likelihood and scale of any such synergistic outcomes are, however, rarely quantified [7].
Marine spatial planning integrates multiple forms of management interventions that are done both in the sea (e.g. fishing restrictions or habitat restoration), and on land (e.g. reducing catchment erosion and other runoff) (e.g. [8]). For example, it is well established that no-take marine reserves work best when they are implemented together with other management interventions that aim to reduce other external impacts (e.g. eutrophication, sediments). For coral reefs, assessments of the efficacy of management actions have been mostly conducted in reef systems with relatively lower impacts from terrestrial sources [9, 10]. By contrast, inshore reefs situated within coastal embayments or estuaries are subjected to several threats from the adjacent land and catchments. Consequently, these are excellent model systems for testing how different management actions on land and in the nearshore zone will benefit inshore marine systems because: 1) the response of reefs to stressors (especially sediments, nutrients and fishing) is generally well understood, 2) management actions to mitigate the effects of stressors are widely implemented (e.g. fishing restrictions, wastewater treatment), and 3) although effects might vary between ecosystems, key threats and their ecological effects are broadly comparable to those affecting other nearshore ecosystems (e.g. oyster reefs, kelp forests, seagrass beds), imparting generality within a broader environmental management context [2, 11].
In this study, we use a Bayesian belief network (BBN) incorporating empirical ecological data of ecosystem components and processes to determine which combinations of different interventions might have the greatest influence on the ecological condition of coral reefs within broader Moreton Bay, central eastern Australia (as opposed to the scale of individual reefs). BBNs are directed graphical models that illustrate causative links between nodes (i.e. variables) via arcs (i.e. arrows). BBNs are mathematically simple, and easy to interpret, thereby allowing increased easy of interpretation, greater scrutiny and simpler interpretation by practitioners [6].
Our principal objective is to develop a tool that is useful to prioritise management actions by modelling the potential ecological outcomes resulting from different combinations in the number and type of management actions. We show that synergistic benefits of multiple interventions are difficult to achieve, so management should focus on prioritising the implementation of actions that maximise combined ecological benefits, and consider these as separate, additive improvements to ecosystem condition. We suggest that similar approaches would be valuable for a range of different systems, but particularly inshore coral reef systems that are affected by a diversity of human threats.
Methods
Study System
The inshore coral reefs of Moreton Bay in subtropical eastern Australia (27°18’S; 153°17’E, Fig 1) provide an ideal location to model the relative effectiveness of, and interaction between, multiple different approaches for ecosystem management. Moreton Bay is managed as part of a high-use marine park. It lies adjacent to the city of Brisbane (population ~2 million people increasing ~2% annually; [12]), is bounded by three barriers islands, receives input from several estuaries that drain highly-modified catchments and transport large sediment and nutrient loads to the Bay [13]. Importantly, the ecological effects of human impacts on Moreton Bay are well understood; there have been multiple, long-term (>10 years), quantitative studies that have examined how human actions alter ecological assemblages and ecosystem functions in the bay (reviewed by [14, 15, 16]).
Habitat layers courtesy Queensland Herbarium.
Moreton Bay’s reefs are positioned within a shallow (<15 m), heterogeneous seascape of mangrove forests, seagrasses, and sandy and muddy seafloor [17, 18] (Fig 1). Reefs are generally dominated by massive corals, especially favids and gonioporids, and can have up to 50% cover of macroalgae [19]. Previous studies have shown that reef fish assemblages [17, 20], ecological functions [21] and MPA effectiveness [22] are influenced more by the spatial attributes of habitats within broader seascapes than by indices of water quality. Conversely, water clarity and water column nutrient concentrations regulate the cover of macroalgae on reefs and the effect of herbivory on algae is highly variable (both spatially and temporally) and low compared with tropical reefs [21, 23]. Consequently, water quality metrics (especially water column nutrient concentration and water clarity) have varied effects on reef health in Moreton Bay. These factors do not correlate with fish community structure, but correlate strongly with macroalgae, and therefore, benthic community structure.
Relatively nutrient poor and clear oceanic water enters the bay via two passages formed between barrier islands in the north and east (Fig 1). In the west, several estuaries drain a total catchment area of 22,700 km2 (Fig 1). Catchments contain large areas of grazing pastures (35%) and urban areas (7%), resulting in significant levels of catchment-derived sediment and nutrient entering the bay from channel erosion [13, 24]. Variable riverine runoff causes significant alterations to both benthic and pelagic community structure throughout the region over seasonal scales [20, 25].
Since the late 1990s, wastewater treatment plants have been upgraded to biological nutrient removal and this has resulted in significant reductions of effluent releases to the bay [26–28]. Treated sewage effluent does, however, still contribute 90% of point-source derived nutrients to the Bay [29].
Currently, 31% of Moreton Bay’s coral reefs are protected by no-take marine reserves, which achieves the 30% representative habitat protection targets recommended by the World Parks Congress [30]. Some studies, however, recommend up to 50% of total protection for the marine environment, particularly for systems and habitats of significance like the subtropical reefs of Moreton Bay [31, 32].
Development of Conceptual Diagrams
We constructed a conceptual model based on our current understanding of the causative relationships between management actions (i.e. a management technique that can be altered in scope or focus) and ecosystem components within Moreton Bay (Fig 2). Our conceptual framework can be divided into five levels: 1) management actions; 2) specified and measurable outcomes of management actions (target nodes); 3) quantifiable components of ecosystems, including ecological functions (field measure nodes); 4) the ecosystem components of the fish or benthic assemblages (component nodes); and 5) coral reef condition, a measure of ecosystem health, where good coral reef health is defined as a reef dominated by scleractinian corals, with low macroalgal cover (as has been the case historically in Moreton Bay [33]) and a high abundance of a diverse array of coral reef fishes (output node). S1 Table provides justifications and sources for all links between nodes.
The model comprises five levels: 1) management interventions (input/management nodes); 2) measurable outcomes of management (target nodes); 3) quantifiable components of ecosystems (field measure nodes); 4) fish or benthic assemblages (component nodes); and 5) overall coral reef condition (output node) for inshore reefs in Moreton Bay, Australia (cf. S1 Table for justifications of links between nodes for this systems).
Construction and Testing of the Bayesian Belief Network Model
The conceptual framework was tested using a Bayesian Belief Network (BBN) in Netica v5.12 [34]. For detailed descriptions of BBN analyses and theories see [35]. Relationships between nodes for target, field measure, component and output nodes were calculated in Netica using published ecological data for 11 reefs in Moreton Bay over two seasons (22 total cases; see S2 Table), with each node containing high and low states. Given that the network was calculated using information from multiple cases across Moreton Bay, conclusions made are generalised to the scale of all reefs across the bay, rather than to specific reef sites. An added benefit of this approach is that our model incorporates all likely environmental and biological conditions present on reefs within Moreton Bay within a year, making conclusions broadly applicable at the scale of whole-of-system management.
The five management nodes that we tested were: 1) fishing restrictions in the form of bag and size limits; 2) spatial extent of no-take marine reserves; 3) restoration of habitats within the bay; 4) levels of catchment-derived sediments; and 5) discharge levels from sewage treatment plants (Table 1; for further justifications see S1 Appendix). We implement a suite of possible management scenarios for each management node that are calculated relative to current and historic levels of management. Consequently, we implemented four actions for each management/input node through: 1) maintaining current management levels; increasing the scale of management actions (i.e. interventions) either 2) levels that are likely to be politically and socially agreeable (henceforth ‘intervention 1’), or 3) to the full levels of management scope and intensity of management interventions that, based on current scientific evidence for the bay, are likely to have the greatest benefits (henceforth ‘intervention 2’; see Table 1); or 4) reducing management actions (henceforth ‘reductions’) (Fig 2; Table 1). Management reductions were modelled to quantify effects of possible limitations to the operating budgets of management agencies, and to assess potentially worsening conditions for certain management nodes (e.g. ongoing habitat loss or increasing effluent releases with increasing population). Full justifications and explanations of underlying data for management scenarios are available in S1 Appendix. We explicitly emphasize that the range of management interventions modelled here represent a subset of all potential management interventions that could possibly be implemented and our results should be considered in this context. Notwithstanding these constraints, the scenarios presented are reasonable and plausible combinations of management actions that may occur.
Here, Intervention 1 relates to management levels that are likely to be politically and socially agreeable, and Intervention 2 relates to the full levels of management scope and intensity of management interventions that, based on current scientific evidence, are likely to have the greatest benefits. Detailed information on the selection of these levels and the data underlying the chosen percentiles is provided S1 Appendix.
Management decisions are rarely made on the basis of scientific decisions alone, and always incorporate social and economic constraints. Importantly, this model was not designed to incorporate all conceivable components of ecosystem management (i.e. all ecological factors as well as social and economic factors), but rather to evaluate the potential ecological benefits of likely management interventions, thereby providing managers with a set of criteria or targets to aim for.
BBNs condition calculations on assumed response ratios, so results are necessarily contingent on the response ratios selected. Altering the response ratios of the effects of management nodes on target nodes (i.e. does an x % change in a management node’s influence result in a 1:1 change in target measurements in the bay?) would have an impact on the outcome for modelled reef condition. However, these changes are mathematically simple: a 1:2 change would result in a doubling of the effect of the management node on the target node, whilst a 2:1 change would result in a halving of that effect, and so on. These effects are trivial to calculate.
To test the model, we first conducted sensitivity analyses on the BBN to determine how sensitive findings at the coral reef condition output node were to findings at all other nodes. We then tested all possible combinations of management input nodes (n = 1024) to determine their relative influence on the likelihood of obtaining a good coral reef condition.
Results
Increasing Marine Reserves and Reducing Sedimentation Are Most Important for Coral Reef Health
Increasing the area of marine reserves had the greatest positive effect on coral reef condition (Fig 3A). Positive effects of increasing the area of marine reserves ranked highly in most of our model outputs (Fig 3A): they were included in 63% of all outputs showing positive net benefits, in 87% of outputs where net benefits were greater than the median, and in all of the top 5% of model outputs (Fig 3B–3D). Decreasing catchment-derived sediments was the next most important management intervention for increasing the modelled condition of coral reefs (Fig 3A). The positive effects of lower sediment inputs occurred in: 58% of all scenarios with positive outcomes, 60% of above median outcomes and 98% of the highest 5% of outcomes (Fig 3B–3D). The three remaining management interventions (i.e. lower releases of treated sewage, stricter bag and length limits for recreational fishers and mangrove and seagrass restoration) were comparatively less influential compared with larger effects modelled for increasing the area of marine reserves or decreasing the volume of sediment inputs (Fig 3).
Performance of management actions: (A) likelihood of a change in state when only a single action is implemented); (B) the proportion of outcomes that resulted in positive effects on coral reef condition; (C) the proportion of outcomes that resulted in above median effects on coral reef condition; and (D) the proportion of outcomes that resulted in the highest 5% of positive effects on coral reef condition. See Table 1 for full details of each management option.
Management reductions were included in some of the combinations of management interventions that were considered to be of high ecological benefit (i.e. those in Fig 3D). These reductions, however, were only ever included for the lower ranked interventions of habitat restoration and fishing restrictions (i.e. red bars in Fig 3D). Overall, the highest-ranking suite of management interventions that contained at least one management reduction was 25% less effective than the highest modelled outcome.
The influence of individual management interventions on coral reef condition varied, but doubling the overall level of management interventions (i.e. implementing Intervention 2 instead of Intervention 1) always resulted in at least double the ecological benefits for coral reefs, irrespective of the particular focus of management (Fig 4; for further detail on management interventions, see Table 1). In general, when management interventions were implemented sequentially, from highest to lowest influence, their combined ecological effects on coral reef condition were mostly additive (Fig 4). Our models did not show sizeable synergistic effects on the likelihood of good coral reef condition until marine reserve extent was doubled, and sediments were halved (Fig 4). We did, however, find that synergistic effects were present in many combinations of management interventions, however, these effects were consistently below 2% (Fig 5). Models suggest that greater benefits can be potentially be achieved by implemented two management actions (reductions in sediment inputs and larger marine reserves) at greater intensity (intervention level 2) rather than implementing more management actions at a lower level of intensity (intervention 1) (Fig 4).
No-take marine reserves, which prohibit fishing and increase fish biomass, are the most effective action for improving reef condition. Targeted catchment management, which reduces erosion and sediment loads, was the second-most effective action. These two actions always accounted for > 50% of improvements in reef condition, but additional gains could also be made through: improvements to wastewater treatment plants, which reduce nitrogen and phosphorus loads; restoration of other marine habitats (e.g. mangroves), which provide ecological functions that improve reef health; and focussed fishing restrictions through bag and size limits. Synergistic effects of these five management actions can result in additional improvements to reef condition, but this does not occur until the highest recommended management interventions by science have been implemented. Outcomes are displayed for two recommended management strategies (refer Table 1).
Figure represents all combinations of management actions that resulted in both positive outcomes for coral reef condition, and that resulted in synergistic additions to coral reef condition (i.e. above and beyond the sum of the values of each action). Outcomes are ordered from left to right from lowest to highest modelled state of good reef condition for each number of interventions implemented. Here, the coloured ‘base' effects (green = two interventions, light blue = three interventions, dark blue = four interventions, grey = five interventions) indicates the sum of the individual interventions in isolation, and 'Synergistic Effect' indicates the additional effect of these interventions when combined.
Bayesian Network Less Sensitive to Lower Node Positions
In our network, modifying the values of component and field measure nodes typically had a greater influence on coral reef condition than changes to target or input/management nodes (Table 2). There were, however, several important exceptions to this pattern; varying the level of herbivory and coral recruitment (both field measure nodes) had a greater effect on benthic assemblages (a component node) and subsequent coral reef condition than did changes to fish-related component or field measure nodes. For example, the benthic component node had almost three times more influence on coral reef condition than did the fish component node, despite being located one arc higher in the network. In turn, both coral and macroalgae cover (field measure nodes) had a greater influence on coral reef condition than their counterpart fish nodes (i.e. of piscivores, herbivores and carnivores).
Nodes with a greater entropy reduction value have a greater influence on the outcome of the output node.
Discussion
Most ecosystems experience a range of impacts, but the capacity of managers to address all stressors is usually constrained by available resources. Consequently, it is critical that management interventions are prioritised according to their likely ecological benefits [40, 41]. Using empirical data collected from field studies of inshore coral reefs in Moreton Bay, Australia, we developed an approach for prioritising management interventions, which also identified opportunities where multiple interventions may have synergistic effects on the ecological condition of coral reefs. Our model shows that marine spatial planning will have the greatest ecological benefits for coral reefs in Moreton Bay where managers focus their investment on increasing the extent of marine reserves and simultaneously decreasing sediment loads that enter the Bay from adjacent river catchments. It is not intended to assess the efficiency, in monetary terms, of actual investment decisions. Optimising land-sea management for inshore coral reefs, therefore, requires management strategies that address impacts in both marine and terrestrial realms [41–43]. We show that synergistic benefits of multiple management interventions are unlikely to be achieved unless best recommendations from science are employed fully. The modelling approach that we developed in this study can be used to prioritize investment across the land-sea interface and ensure that impacts are addressed in the most cost effective sequence possible.
When management interventions where implemented in order of importance (i.e. from those that had the greatest to lowest influence) the ecological benefits for coral reef condition were typically additive (Fig 4). Management interventions did not have measurable synergistic effects on coral reef condition until the highest interventions recommended by science were employed for the cover of marine reserves and reductions in sediments (see Intervention 2, step 2; Fig 4). The synergistic benefits of management actions on both marine reserves and sediments were, however, consistently small (<2% improvement in modelled reef condition which is likely no greater than natural variation), compared to the significant costs required for their implementation [13, 32, 44]. In our model, synergistic benefits for coral reefs in in Moreton Bay require increasing the spatial extent of marine reserves from 31 to 62% [32], and rehabilitating 6,350 km of riparian land in adjoining catchments to reduce sediment loads (Table 1) [13]. This level of investment is probably not feasible, given other social and ecological planning considerations in the region. Nevertheless, this finding accentuates the importance of prioritising management actions to explicitly address quantitative targets and to optimise likely return on investment [7, 41]. To maximise cost effectiveness, managers might, therefore, adopt a strategy that seeks to: 1) implement interventions with the highest benefits on ecological condition up to intervention one (in this case increase marine reserves to 46% protection and reduce sediments by 25%), thereby increasing ecological outcomes for investments, 2) spend remaining moneys across the other interventions, thereby maximising the joint ecological benefits all interventions (i.e. move towards the right of Fig 5); and 3) once intervention one has been reached for all actions, seek to increase investment up to intervention two for the major actions, thereby increasing the likelihood of synergistic effects. Importantly, under such approach, it is vital for changes in management actions to be made in concert, rather than in separate pieces of legislation under different jurisdictions or organisations. This is a simple and effective approach for optimising land-sea management for coastal ecosystems, and one which might overcome much of the uncertainty that can be associated with investments in coastal conservation and ecosystem management [45–47].
Increasing the spatial extent of marine reserves and reducing sediments were the two management interventions that resulted in the greatest ecological benefits for reefs in Moreton Bay. These management interventions are important, but they can also be expensive and must be balanced against other social, economic and political considerations. It is, therefore, critical that any changes to the marine reserves or catchment rehabilitation programs are done to maximise cost-effectiveness [48–50]. Given our current understanding of factors affecting performance, it is clear that any new marine reserves should be: 1) located to maximise positive effects of connectivity (i.e. with seagrasses and mangroves) on fish assemblages and ecological functions [51]; 2) big enough to protect species with large home ranges [52]; and 3) placed at reefs that are impacted less frequently by the chronic effects of flooding, sediments and eutrophication [19, 53, 54]. Furthermore, spending money improving the condition of the catchment is important because terrestrial-based stressors can override marine-based protections (e.g. [19, 55]). Reducing catchment-born sediments is a significant challenge, both within Moreton Bay [44, 56], and more broadly [57, 58], requiring managers to focus on 1) revegetating catchment verge vegetation and rehabilitation [13, 59], 2) maintaining current levels of remnant vegetation in the catchment at the highest possible levels [60, 61], and 3) implement ideas of intelligent design of urban water runoff systems [62, 63]. Given the significant levels of catchment revegetation required to reduce loads by the scientifically recommended 50% [13], it would seem that this intervention, and associated synergistic effects between marine reserves and sediments are unlikely. Such findings surrounding the importance of controlling harvesting through marine reserves and reducing the influence of catchment-borne sediments agree with many previous articles [2, 19, 52].
Management interventions that reduced eutrophication by limiting treated sewage releases were clearly placed third to reserves and sedimentation in importance to reef health. Limiting fishing effects (through restrictions on take) and enhanced the restoration of ecologically linked habitats (e.g. mangroves and seagrass) were less important to coral reef condition than either marine reserves, sediments or sewage. However, we show that implementing management interventions which address any of these key impacts resulted in positive and quantifiable ecological benefits for coral reefs in Moreton Bay. This is crucial, as managers should not underestimate the capacity of these management actions to assist in incremental improvements in overall ecosystem condition. Further improvements will be made to sewage releases as technology advances, however, this may be offset by significant population growth within the region in future years [12], meaning that reductions in total effluent release (in terms of kg/y) are of low likelihood; a problem of global significance [64, 65]. Increasing the strictness of fishing size and bag limits is likely to be one of the more financially viable interventions discussed here, as such regulations are already policed [38]; however, such changes are likely to be politically and socially difficult to implement, and there is some uncertainty surrounding the capacity for these rules to deter fishers from reducing overall catch [66, 67] and the degree to which potential changes result in positive ecological outcomes [68, 69]. Finally, restoring up to 1850 ha and 2580 ha of seagrass and mangroves, respectively, is a significant challenge for such a small return (~2.3%), meaning that such interventions which aim only to improve coral reefs are likely financially limited. Marine habitat restoration has proven successful in many systems, but must be optimised and prioritised in much the same way that marine reserve placement is [7, 70].
Global changes (i.e. warming, sea level rise) have the potential to change and override local stressors and management [71]. Although the current study did not incorporate such effects, the long-term influences of these changes should always be considered [72]. Further, there has been extensive research conducted on the importance of microbial communities for coral reef health (e.g. [73]). Whilst microbial data is not currently available for the study system, we acknowledge that the inclusion of such data into the model might have been beneficial. In any case, the effects of such factors are likely to be lower than the impacts of our top three most includes factors.
Conclusions
In this study, we provide model estimates of ecological benefits of management actions using empirical ecological data to show that optimising land-sea management for inshore coral reefs requires management strategies that address impacts in both marine (marine reserves) and terrestrial (sediments) realms. In our study, interventions did not have synergistic effects on coral reef condition until after maximum investment to increase the cover of marine reserves and decrease sediments; both of which are likely to be difficult to achieve both financially and politically. Synergistic benefits were also very small compared to the significant costs required for their implementation. In combination, these findings indicate that synergistic effects of multiple management actions on ecosystem condition are unlikely within our study system. Therefore, to maximise cost effectiveness in other similar systems, which have been heavily degraded by multiple human impacts, we suggest that managers should consider adopting a strategy that seeks to; 1) maximise the ecological benefits of joint ecological effects by implementing actions with the highest ecological outcomes up to intervention one first, 2) spend further moneys, up the the total managerial budget, evenly across remaining actions, and then 3) seek increasing the likelihood of synergistic effects by increasing investment up to intervention in the same order as step 1. This is a simple and effective approach for prioritizing investment across the land-sea interface for inshore coral reefs, and other coastal ecosystems that are similarly afflicted by multiple human impacts.
Supporting Information
S1 Appendix. Description and justification of modelled management interventions.
https://doi.org/10.1371/journal.pone.0164934.s001
(DOCX)
S1 Table. Justification of all arcs between nodes in the conceptual framework for management of Moreton Bay’s reefs.
https://doi.org/10.1371/journal.pone.0164934.s002
(DOCX)
S2 Table. Data sources for the Bayesian belief network.
https://doi.org/10.1371/journal.pone.0164934.s003
(DOCX)
Acknowledgments
The authors acknowledge all field volunteers and the staff of the Moreton Bay Research Station, North Stradbroke Island, Australia. This project was funded in part by the Australian Rivers Institute and School of Environment at Griffith University.
Author Contributions
- Conceptualization: BLG ADO TS CJH IRT PSM TAS.
- Formal analysis: BLG ADO PSM TAS DSS DR.
- Funding acquisition: BLG TS TAS.
- Investigation: BLG TS IRT CJH ADO.
- Methodology: BLG TS IRT TAS PSM ADO RMC.
- Project administration: TS IRT TAS.
- Resources: BLG TS.
- Software: BLG PSM DSS.
- Supervision: TS IT TS.
- Visualization: BLG ADO.
- Writing – original draft: BLG ADO TAS RMC.
- Writing – review & editing: BLG ADO RMC TS CJH PSM IRT DSS DR TAS.
References
- 1. Kennish MJ. Environmental threats and environmental future of estuaries. Environmental Conservation. 2002;29(01):78–107.
- 2. Halpern BS, Walbridge S, Selkoe KA, Kappel CV, Micheli F, D'Agrosa C, et al. A global map of human impact on marine ecosystems. Science. 2008;319(5865):948–52. Epub 2008/02/16. pmid:18276889.
- 3. Mitchell SB, Jennerjahn TC, Vizzini S, Zhang WG. Changes to processes in estuaries and coastal waters due to intense multiple pressures—An introduction and synthesis. Estuarine Coastal and Shelf Science. 2015;156:1–6. pmid:WOS:000354151700001.
- 4. Brown CJ, Saunders MI, Possingham HP, Richardson AJ. Interactions between global and local stressors of ecosystems determine management effectiveness in cumulative impact mapping. Diversity and Distributions. 2014;20(5):538–46. pmid:WOS:000333926300005.
- 5. Strain EMA, Thomson RJ, Micheli F, Mancuso FP, Airoldi L. Identifying the interacting roles of stressors in driving the global loss of canopy-forming to mat-forming algae in marine ecosystems. Global Change Biology. 2014;20(11):3300–12. pmid:WOS:000343762800002.
- 6. Maxwell PS, Pitt KA, Olds AD, Rissik D, Connolly RM. Identifying habitats at risk: simple models can reveal complex ecosystem dynamics. Ecological Applications. 2015;25(2):573–87. pmid:26263676
- 7. Margules CR, Pressey RL. Systematic conservation planning. Nature. 2000;405(6783):243–53. pmid:WOS:000087080100062.
- 8. Halpern BS, Ebert CM, Kappel CV, Madin EMP, Micheli F, Perry M, et al. Global priority areas for incorporating land-sea connections in marine conservation. Conservation Letters. 2009;2(4):189–96. pmid:WOS:000279223600006.
- 9. Klein CJ, Tulloch VJ, Halpern BS, Selkoe KA, Watts ME, Steinback C, et al. Tradeoffs in marine reserve design: habitat condition, representation, and socioeconomic costs. Conservation Letters. 2013;6(5):324–32. pmid:WOS:000325501800004.
- 10.
Alvarez-Romero JG, Pressey RL, Ban NC, Vance-Borland K, Willer C, Klein CJ, et al. Integrated Land-Sea Conservation Planning: The Missing Links. In: Futuyma DJ, Shaffer HB, Simberloff D, editors. Annual Review of Ecology, Evolution, and Systematics, Vol 42. Annual Review of Ecology Evolution and Systematics. 422011. p. 381–409.
- 11. Lotze HK, Lenihan HS, Bourque BJ, Bradbury RH, Cooke RG, Kay MC, et al. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science. 2006;312(5781):1806–9. pmid:WOS:000238452800061.
- 12.
Australian Bureau of Statistics, Canberra. Queensland Population Growth Canberra, Australia2012 [cited 2012 13-2-2012].
- 13. Olley J, Burton J, Hermoso V, Smolders K, McMahon J, Thomson B, et al. Remnant riparian vegetation, sediment and nutrient loads, and river rehabilitation in subtropical Australia. Hydrological Processes. 2015;29(10):2290–300. pmid:WOS:000353296900002.
- 14.
Tibbetts IR, Hall NJ, Dennison WC. Moreton Bay and Catchment. Brisbane, Australia: School of Marine Science, University of Queensland; 1998.
- 15. Gibbes B, Grinham A, Neil D, Olds A, Maxwell P, Connolly R, et al. Moreton Bay and its estuaries: A sub-tropical system under pressure from rapid population growth. 2014:203–22.
- 16. Leigh C, Burford MA, Connolly RM, Olley JM, Saeck E, Sheldon F, et al. Science to support management of receiving waters in an event-driven ecosystem: From land to river to sea. Water. 2013;5(2):780–97. pmid:WOS:000320769800026.
- 17. Olds AD, Connolly RM, Pitt KA, Maxwell PS. Primacy of seascape connectivity effects in structuring coral reef fish assemblages. Marine Ecology Progress Series. 2012;462:191–203.
- 18. Stevens T, Connolly RM. Local-scale mapping of benthic habitats to assess representation in a marine protected area. Marine and Freshwater Research. 2005;56:111–23.
- 19. Gilby BL, Maxwell PS, Tibbetts IR, Stevens T. Bottom-up factors for algal productivity outweigh no-fishing marine protected area effects in a marginal coral reef system. Ecosystems. 2015;18(6):1056–69.
- 20. Gilby BL, Tibbetts IR, Olds AD, Maxwell PS, Stevens T. Seascape context and predators override water quality effects on inshore coral reef fish communities. Coral Reefs. 2016; 35(3):979–990.
- 21. Yabsley NA, Olds AD, Connolly RM, Martin TSH, Gilby BL, Maxwell PS, et al. Resource type modifies the effects of reserves and connectivity on ecological functions. Journal of Animal Ecology. 2015;85(2):437–444. pmid:26476209
- 22. Olds AD, Pitt KA, Maxwell PS, Connolly RM. Synergistic effects of reserves and connectivity on ecological resilience. Journal of Applied Ecology. 2012;49(6):1195–203.
- 23.
Gilby BL, Tibbetts IR, Stevens T. Low functional redundancy and high variability in Sargassum browsing fish populations in a subtropical reef system. Marine and Freshwater Research. 2016;In Press.
- 24. Wallbrink PJ. Quantifying the erosion processes and landuses which dominate fine sediment supply to Moreton Bay, Southeast Queensland, Australia. Journal of Environmental Radioactivity. 2004;76:67–80. pmid:15245841
- 25. Saeck EA, Hadwen WL, Rissik D, O'Brien KR, Burford MA. Flow events drive patterns of phytoplankton distribution along a river-estuary-bay continuum. Marine and Freshwater Research. 2013;64(7):655–70. pmid:WOS:000321264800007.
- 26. Pitt KA, Connolly RM, Maxwell P. Redistribution of sewage-nitrogen in estuarine food webs following sewage treatment upgrades. Marine Pollution Bulletin. 2009;58(4):573–80. pmid:19138774.
- 27. Costanzo SD, Udy J, Longstaff B, Jones A. Using nitrogen stable isotope ratios (d15N) of macroalgae to determine the effectiveness of sewage upgrades: changes in the extent of sewage plumes over four years in Moreton Bay, Australia. Marine Pollution Bulletin. 2005;51:212–7. pmid:15757722
- 28. Connolly RM, Gorman D, Hindell JS, Kildea TN, Schlacher TA. High congruence of isotope sewage signals in multiple marine taxa. Marine Pollution Bulletin. 2013;71(1–2):152–8. pmid:23602260
- 29.
HWMP. Healthy Waterways Monitoring Program. 2015. Available: http://www.healthywaterways.org/. Accessed 7 April 2015.
- 30.
World Parks Congress. The promise of Sydney. Sydney, Australia: 2014.
- 31. Dobbs K, Fernandes L, Slegers S, Jago B, Thompson L, Hall J, et al. Incorporating dugong habitats into the marine protected area design for the Great Barrier Reef Marine Park, Queensland, Australia. Ocean & Coastal Management. 2008;51(4):368–75.
- 32.
World Parks Congress. A strategy of innovative approaches and recommendations to reach conservation goals in the next decade. Sydney, Australia: 2014.
- 33. Lybolt M, Neil D, Zhao J, Feng Y, Yu K- F, Pandolfi J. Instability in a marginal coral reef: the shift from natural variability to a human-dominated seascape. Frontiers in Ecology and the Environment. 2011;9(3):154–60.
- 34.
Software Norsys. Netica User's Guide. Vancouver, Canada: Norsys Software; 1997.
- 35.
Jensen FV. Bayesian Networks and Decision Graphs. New York: Springer; 2001.
- 36. Manson FJ, Loneragan NR, Phinn SR. Spatial and temporal variation in distribution of mangroves in Moreton Bay, subtropical Australia: a comparison of pattern metrics and change detection analyses based on aerial photographs. Estuarine Coastal and Shelf Science. 2003;57(4):653–66. pmid:WOS:000184979400011.
- 37. Saunders MI, Leon J, Phinn SR, Callaghan DP, O'Brien KR, Roelfsema CM, et al. Coastal retreat and improved water quality mitigate losses of seagrass from sea level rise. Global Change Biology. 2013;19(8):2569–83. pmid:WOS:000328744900023.
- 38.
Webley J, McInnes K, Teixeira D, Lawson A, Quinn R. Statewide Recreational Fishing Survey 2013–14. Brisbane, Australia: Queensland Government, 2015.
- 39.
Campbell AB, O'Neill MF, Sumpton W, Kirkwood J, Wesche S. Stock assessment summary of the Queensland snapper fishery (Australia) and management strategies for improving sustainability. Brisbane, Queensland: Department of Employment, Economic Development and Innovation, 2009.
- 40. Brown CJ, Bode M, Venter O, Barnes MD, McGowan J, Runge CA, et al. Effective conservation requires clear objectives and prioritizing actions, not places or species. Proceedings of the National Academy of Sciences of the United States of America. 2015;112(32):E4342–E. pmid:WOS:000359285100003.
- 41. Klein CJ, Ban NC, Halpern BS, Beger M, Game ET, Grantham HS, et al. Prioritizing land and sea conservation investments to protect coral reefs. PLoS One. 2010;5(8). pmid:WOS:000281375500003.
- 42. Makino A, Beger M, Klein CJ, Jupiter SD, Possingham HP. Integrated planning for land-sea ecosystem connectivity to protect coral reefs. Biological Conservation. 2013;165:35–42. pmid:WOS:000323871700005.
- 43. Beger M, Grantham HS, Pressey RL, Wilson KA, Peterson EL, Dorfman D, et al. Conservation planning for connectivity across marine, freshwater, and terrestrial realms. Biological Conservation. 2010;143(3):565–75. pmid:WOS:000275946400006.
- 44. Saxton NE, Olley JM, Smith S, Ward DP, Rose CW. Gully erosion in sub-tropical south-east Queensland, Australia. Geomorphology. 2012;173:80–7. pmid:WOS:000309573900007.
- 45. Nicotra AB, Beever EA, Robertson AL, Hofmann GE, O'Leary J. Assessing the components of adaptive capacity to improve conservation and management efforts under global change. Conservation Biology. 2015;29(5):1268–78. pmid:WOS:000363729100003.
- 46. Johnson FA, Eaton MJ, McMahon G, Nilius R, Bryant MR, Case DJ, et al. Global change and conservation triage on National Wildlife Refuges. Ecology and Society. 2015;20(4). pmid:WOS:000367628900026.
- 47. Boersma PD, Parrish JK. Limiting abuse: marine protected areas, a limited solution. Ecological Economics. 1999;31(2):287–304. pmid:WOS:000084042500009.
- 48. Carwardine J, Wilson KA, Hajkowicz SA, Smith RJ, Klein CJ, Watts M, et al. Conservation planning when costs are uncertain. Conservation Biology. 2010;24(6):1529–37. pmid:WOS:000284172800012.
- 49. Scales IR. Paying for nature: what every conservationist should know about political economy. Oryx. 2015;49(2):226–31. pmid:WOS:000352628600013.
- 50. Firn J, Martin TG, Chades I, Walters B, Hayes J, Nicol S, et al. Priority threat management of non-native plants to maintain ecosystem integrity across heterogeneous landscapes. Journal of Applied Ecology. 2015;52(5):1135–44. pmid:WOS:000361186500004.
- 51. Olds AD, Connolly RM, Pitt KA, Maxwell PS. Habitat connectivity improves reserve performance. Conservation Letters. 2012;5(1):56–63. pmid:WOS:000299468900007.
- 52. Edgar GJ, Stuart-Smith RD, Willis TJ, Kininmonth S, Baker SC, Banks S, et al. Global conservation outcomes depend on marine protected areas with five key features. Nature. 2014;506:216–20. Epub 2014/02/07. pmid:24499817.
- 53. Olds AD, Pitt KA, Maxwell PS, Babcock RC, Rissik D, Connolly RM. Marine reserves help coastal systems cope with extreme weather. Global Change Biology. 2014;20:3050–8. pmid:24849111
- 54. Maxwell PS, Pitt KA, Burfeind DD, Olds AD, Babcock RC, Connolly RM. Phenotypic plasticity promotes persistence following severe events: physiological and morphological responses of seagrass to flooding. Journal of Ecology. 2014;102(1):54–64.
- 55. Huijbers CM, Connolly RM, Pitt KA, Schoeman DS, Schlacher TA, Burfeind DD, et al. Conservation benefits of marine reserves are undiminished near coastal rivers and cities. Conservation Letters. 2015;8(5):312–319n/a-n/a.
- 56. Grinham A, Gale D, Udy J. Impact of sediment type, light and nutrient availability on benthic diatom communities of a large estuarine bay: Moreton Bay, Australia. Journal of Paleolimnology. 2011;46(4):511–23. pmid:WOS:000297593300002.
- 57. Fabricius KE. Effects of terrestrial runoff on the ecology of corals and coral reefs: review and synthesis. Marine Pollution Bulletin. 2005;50(2):125–46. pmid:WOS:000227762000014.
- 58. Bartley R, Bainbridge ZT, Lewis SE, Kroon FJ, Wilkinson SN, Brodie JE, et al. Relating sediment impacts on coral reefs to watershed sources, processes and management: A review. Science of the Total Environment. 2014;468:1138–53. pmid:WOS:000331776000120.
- 59. Jones BK, Neale AC, Nash MS, Van Remortel RD, Wickham JD, Riitters KH, et al. Predicting nutrient and sediment loadings to streams from landscape metrics: a multiple watershed study from the United States Mid-Atlantic Region. Landscape Ecology. 2001;16:301–12.
- 60. Allan JD, Erickson DL, Fay J. The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology. 1997;37(1):149–61. pmid:WOS:A1997WJ19200012.
- 61. Nearing MA, Jetten V, Baffaut C, Cerdan O, Couturier A, Hernandez M, et al. Modeling response of soil erosion and runoff to changes in precipitation and cover. Catena. 2005;61(2–3):131–54. pmid:WOS:000230297600004.
- 62. Bratieres K, Fletcher TD, Deletic A, Zinger Y. Nutrient and sediment removal by stormwater biofilters: A large-scale design optimisation study. Water Research. 2008;42(14):3930–40. pmid:WOS:000259753000038.
- 63. Vaze J, Chiew FHS. Nutrient loads associated with different sediment sizes in urban stormwater and surface pollutants. Journal of Environmental Engineering-Asce. 2004;130(4):391–6. pmid:WOS:000220531800005.
- 64. Gupta VK, Ali I, Saleh TA, Nayak A, Agarwal S. Chemical treatment technologies for waste-water recycling-an overview. RSC Advances. 2012;2(16):6380–8. pmid:WOS:000306415100001.
- 65. Rabalais NN, Diaz RJ, Levin LA, Turner RE, Gilbert D, Zhang J. Dynamics and distribution of natural and human-caused hypoxia. Biogeosciences. 2010;7(2):585–619. pmid:WOS:000274993900013.
- 66.
Cook MF, Goeman TJ, Radomski PJ, Younk JA, Jacobson PC. Creel limits in Minnesota: a proposal for change. Fisheries. 2001;Fisheries:19–26.
- 67. Radomski PJ, Grant GC, Jacobson PC, Cook MF. Visions for recreational fishing regulations. Fisheries. 2001;26:7–18.
- 68. Tetzlaff JC, Pine WE III, Allen MS, Ahrens RNM. Effectiveness of size limits and bag limits for managing recreational fisheries: a case study of the Gulf of Mexico recreational gag fishery. Bulletin of Marine Science. 2013;89(2):483–502. pmid:WOS:000318378500005.
- 69. van Poorten BT, Cox SP, Cooper AB. Efficacy of harvest and minimum size limit regulations for controlling short-term harvest in recreational fisheries. Fisheries Management and Ecology. 2013;20(2–3):258–67. pmid:WOS:000316125200015.
- 70. Perring MP, Standish RJ, Price JN, Craig MD, Erickson TE, Ruthrof KX, et al. Advances in restoration ecology: rising to the challenges of the coming decades. Ecosphere. 2015;6(8):art131.
- 71. Perry RI, Ommer RE. Scale issues in marine ecosystems and human interactions. Fisheries Oceanography. 2003;12(4–5):513–22. pmid:WOS:000185446000028.
- 72. Tribbia J, Moser SC. More than information: what coastal managers need to plan for climate change. Environmental Science & Policy. 2008;11(4):315–28. pmid:WOS:000257013400004.
- 73. McDole T, Nulton J, Barott KL, Felts B, Hand C, Hatay M, et al. Assessing coral reefs on a Pacific-wide scale using the microbialization score. PLoS One. 2012;7:e43233. pmid:22970122