Introduction
With the global population expected to reach nine billion inhabitants by 2050, global food demand is expected to increase by 35–56% by 2050 from 2010 levels, consequently exacerbating tensions between land use for agriculture and habitat for biodiversity (van Dijk et al.
2021). Agricultural expansion and intensification can result in the loss of ecosystems and biodiversity, the eutrophication and sedimentation of aquatic ecosystems, increased greenhouse gas emissions, and altered hydrology (Tilman et al.
2001,
2002; Awuchi et al.
2020; Gaugler et al.
2020). Wetlands, including rivers and estuaries following RAMSAR convention (Ramsar Convention Secretariat
2016), are among the ecosystems most impacted by agricultural activities given their vulnerability to nutrient, sediment, and pesticide runoff (Buck et al.
2012; Ostrowski et al.
2021), water abstraction (Acreman et al.
2000), and to drainage and reclamation (Coleman et al.
2008). Global floodplain wetland loss is estimated at 95 km
2/yr, though this loss is unlikely to slow or reduce under global human population predictions over the coming few decades (Coleman et al.
2008; Davidson
2014), with climate change set to exacerbate wetland loss and fragmentation across the globe (Segan et al.
2016). Despite this, approximately 10% of global animal biodiversity is found in freshwater ecosystems, which occupy <1% of Earth’s surface (Dudgeon
2019), with wetlands providing many ecosystem services (Barbier
2019; Davidson et al.
2019; Xu et al.
2020). There is, however, growing recognition of the need for the transformation of agricultural landscapes with both wetlands restoration and the application of agroecological principles, such as regenerative agricultural practices (LaCanne and Lundgren
2018; van Coppenolle and Temmerman
2019; Gliessman
2020). Large-scale wetlands construction and restoration are critical to advancing progress towards the UN Sustainable Development Goals (SDGs), including the universal provision of clean water access, the sustainable management of land and water ecosystems, and climate action (United Nations General Assembly
2015). Agroecology uses ecological principles to design and manage sustainable food systems, seeking to reconcile economic, environmental, and social dimensions (Gliessman
2020).
Like the rest of the world, Australia faces a legacy of degraded freshwater ecosystems, despite a small population and a relatively short 200 years of colonial urban, industrial, and agricultural development (Creighton et al.
2016). For example, between 2001 and 2017, the Great Barrier Reef (GBR) catchment experienced net loss of 740 ha (0.55% of 2001 extent) of coastal floodplain wetland. While contemporary loss has been relatively small, since European arrival, approximately 77,500 ha of palustrine wetland has been lost (~21.2% loss of pre-clear extent), much of which is from freshwater floodplain systems. Artificial/highly modified wetlands, however, have increased substantially between 2001 and 2017, increasing by 21,690 ha, much of which was created through the construction of tidal barrages (8299 ha) (Environmental Protection Agency
2005; Department of Environment and Science
2019). This loss of natural habitat, and the degradation of remaining habitat, is also reducing the GBR’s resilience to pressures from on-going pollutant runoff (Waterhouse et al.
2016; MacNeil et al.
2019; Adame et al.
2019a), and reducing habitat availability for species with freshwater life stages (Arthington et al.
2015; Adame et al.
2019a). Not only do coastal freshwater floodplain wetlands support diverse biological communities, but they also form part of important, broader, connected ecosystems, providing habitat for migratory species, flow regulation, and reduce sedimentation and nitrogen runoff (Bainbridge et al.
2009; Brodie and Waterhouse
2012; Waterhouse et al.
2016). While artificial/highly modified wetlands have been largely constructed for agricultural benefit, they still provide habitat for many wetland-dependent species and may be mitigating impacts of natural wetland loss (Canning and Waltham
2021).
Funding wetlands restoration and ensuring that it does not come at the expense of food production are two of the biggest barriers to large-scale wetland restoration (Waltham et al.
2020; Canning et al.
2021). While wetlands have been shown to improve food security directly through providing harvestable resources (e.g., fish, shellfish, and plants) (Cunningham
2015), food production can be improved indirectly through increased water security for irrigation and improved catchment drainage (Shennan and Bode
2002; Chen and Wong
2016). Leveraging on situations where wetland restorations also provide agricultural benefits (i.e., win-win scenarios) will be essential for realizing large-scale wetland restoration. In many situations, the financial benefits arising from wetland restoration need to outweigh the costs of restoring and maintaining the wetlands, and lost income. Where wetlands do not result in sufficient on-farm profit returns, financial incentive schemes, such as those paying for the provision of ecosystem services in the form or carbon sequestration, nutrient removal, or habitat provision, may help to ensure they are profitable at the farm scale (Banerjee et al.
2013; Sapkota and White
2020; Canning et al.
2021).
Instances where restored wetlands provide both agricultural and biodiversity benefits have primarily been due to the use of wetlands for growing rice, supporting livestock grazing, and improved water storage for irrigation (Verhoeven and Setter
2010; McIntyre et al.
2011; Peh et al.
2014). However, the benefits from using wetlands for drainage and flow regulation to improve both crop production and biodiversity simultaneously are rarely documented (Brander et al.
2013; Kadykalo and Findlay
2016).
On the floodplains of Australia’s wettest catchment (Tully-Murray), sugarcane farmers have grappled with the challenges of growing sugarcane on frequently flooded land, while supporting the health of the downstream Great Barrier Reef ecosystems. As part of Queensland’s $40 M (2001 AUD) Sugar Industry Infrastructure Program (hereafter ‘SIIP’), funded by the Queensland Government and the Australian Government, the Riversdale-Murray Valley Water Management Scheme (hereafter ‘the Riversdale-Murray Scheme’), completed in 2004, financially incentivised sugarcane farmers to construct wetlands across the Tully-Murray floodplain (Ernst and Young
2001). The Riversdale-Murray Scheme aimed to reduce cane inundation by flood waters and increase cane production across the Tully-Murray floodplain through the creation of arterial drainage and lagoon wetlands. Reduced inundation of mid-catchment cane land was achieved by directing water via preferential flow paths and draining it into ‘sump’ lagoons where the increased residence time then improved flow regulation and reduced flooding of downstream areas (Merrin; Karim et al.
2012,
2014). Lagoons were to have 100 m
2 of surface area per hectare of their contributing catchment and a depth of up to 3 m to provide adequate detention time. Spoil excavated from a lagoon could be spread on adjacent low-lying areas to improve their productivity. Lagoons were also designed to be steep-sided to reduce weed growth, vegetated, and well-connected to the downstream river to provide on-farm fish habitat. As a result, a series of lagoons were created via the Riversdale-Murray Scheme for the specific purpose of providing both agricultural and biodiversity benefits (Merrin).
With over 15 years of wetland maturation and stakeholder hindsight, the Riversdale-Murray Scheme provides an opportunity to evaluate the impacts of catchment-integrated wetlands, designed primarily for improved drainage and flow regulation, on sugarcane profitability, fish biodiversity, and an array of ecosystem services more broadly. In this study, we aimed to:
(1)
Quantify the fish biodiversity, and associated water quality, provided by scheme-funded wetlands;
(2)
Estimate the return on investment and the benefit to cost ratio from the perspective of a representative landholder investing in wetland construction via the Riversdale-Murray Scheme; and
(3)
Identify the range of final ecosystem services potentially provided by the Scheme-funded wetlands.
Discussion
Providing for the needs of agriculture and biodiversity in the same landscape is a significant challenge the world over, with heated debate over approaches for balancing competing demands. Here we observed that the integration of constructed lagoons throughout an intensive sugarcane dominated catchment in north Queensland to reducing flooding not only improved sugarcane profitability but provided habitat for freshwater biodiversity and potentially provide numerous other ecosystem services.
We found that the constructed lagoons provided sufficient habitat, water quality and connectivity to support high native fish diversity (Pearson et al.
2013; Karim et al.
2014), including commercially valuable species such as mangrove jack (
Lutjanus argentimaculatus) and barramundi (
Lates calcarifer), and iconic species such as the saltwater crocodile (
Crocodylus porosus). Mangrove jack, for example, spawn near the outer reef and continental shelf, then migrate as juveniles to the shoreline, inhabiting mangrove roots, snags and rocks, gradually moving upriver and into lagoons as they mature. Once mature, after 2–11 years, adults begin to migrate back to spawning areas where they may reside for up to 40 years (Waltham et al.
2019). Barramundi, a diadromous fish, will ingress into coastal freshwater wetlands during wet season floods, to access important nursery habitat and forage in wetlands. As the lagoons were designed to be close to, and well connected to, the mainstem of the Tully River (Karim et al.
2014), migratory species are able to access and use the created habitat. The lagoons were also designed to be steep-sided and at least 1 m deep, which has helped to reduce macrophyte weed growth and provide habitable dissolved oxygen concentrations (Butler and Burrows
2007). As nutrient and phytoplankton (indicated by chlorophyl
a) concentrations were high, reducing nutrient runoff into the lagoons may further improve the habitable condition and food web (Dodds and Smith
2016). Despite being artificial and draining intensive agricultural land, the wetlands support freshwater biodiversity and may provide alternative habitat to that lost from the draining of natural wetlands (Canning and Waltham
2021). Further work is required to examine the contribution of the lagoons towards providing habitat and food for aquatic birds and insects and delineating the habitats that support the greatest diversity for these groups.
In addition to the lagoons (and associated drains) supporting freshwater biodiversity, we estimated that they also improved on-farm profitability (from a landholder’s perspective) in two ways: (1) increased cane yield from improved drainage and flow regulation across a portion of the farm, and (2) increased cane yield from land that was elevated using wetland excavation spoil. Improved drainage and flow regulation reduced cane yield loss from waterlogging. Elevation of land with excavation spoil permitted two potential land use change scenarios to arise. The first was the conversion of cattle fattening land to production of sugarcane on a plant cane plus four ratoons cycle. The second was extension of the existing sugarcane production cycle from plant cane plus two ratoons to plant cane plus four ratoons.
It remains unexamined whether the scheme resulted in a net increase or decrease in nitrogen loading to the downstream Great Barrier Reef (GBR). An expansion and intensification of cane farming, along with increased drainage, may result in greater nitrogen leaching and runoff (Thorburn et al.
2011; Fraser et al.
2017). However, the constructed lagoon has the potential to denitrify runoff and offset any increased nitrogen losses (Land et al.
2016; Adame et al.
2019b; Wallace and Waltham
2021), while the increased cane growth, reduced tillage from longer cane cycles (increased ratooning) and reduced fertiliser loss (from flooding) may contribute to lower nitrogen losses (Webster et al.
2012; Skocaj et al.
2013; Thorburn et al.
2017). As the Great Barrier Reef requires substantial reductions in nitrogen loading to improve its ecological health (Brodie et al.
2012; Kroon et al.
2012; Wooldridge et al.
2015), if the approach resulted in a net increase in nitrogen loading to the reef, then decision-makers would need to consider a values trade-off. The health of the Great Barrier Reef, and the associated economic benefits from tourism and fisheries, would be pitted against improved sugarcane profitability and freshwater biodiversity. A net nitrogen reduction, however, would benefit the reef and could help attract funding that improves financial viability.
The extent to which nitrogen is removed by wetlands depends on many factors, including: the concentration and speciation nitrogen inflows; hydraulic loading rate, residence time and efficiency; temperature; wetland size and shape; composition of the ecological community (particularly the vegetation type and density); sediment type and composition; and oxygen concentrations and redox potential (Land et al.
2016; Alldred and Baines
2016; Vymazal
2017). Denitrification occurs in anaerobic conditions (negative oxidation-reduction (redox) potential) and when nitrate is used by denitrifying bacteria in respiration. Denitrifying bacteria are those with either the nirS, nirK, and nosZ genes, and use oxidised nitrogen compounds as a terminal electron acceptor in the absence of oxygen. Complete denitrification occurs in optimal conditions and released nitrogen gas as a by-product, whereas sub-optimal conditions lead to incomplete denitrification that releases N
2O (a potent greenhouse gas) or NO
2 as by-products (Burgin and Hamilton
2007; Martínez-Espinosa et al.
2021; Pinto et al.
2021). Complete denitrification is more probable when C:N ratios are high (e.g., >15–20) and anoxic conditions are persistent, if carbon becomes scarce or the environment becomes oxygenated then incomplete denitrification can occur (Klemedtsson et al.
2005). It is, therefore, essential that constructed wetlands with the goal of reducing nitrogen runoff are designed to have high hydrological residence time, persistent anoxic conditions and high carbon supply if it is to have low nitrous oxide emissions (Land et al.
2016; Oertel et al.
2016; Jahangir et al.
2016; Maucieri et al.
2017). Anoxic conditions can arise when soils are water-logged with minimal mixing, have minimal disturbance (such as mechanical ploughing and animal grazing) (Drewry et al.
2008), and when plant oxygenation rates are low as aquatic plants often oxygenate the soils proximal to their roots via aerenchyma transport (Oertel et al.
2016; Jahangir et al.
2016; Maucieri et al.
2017). Root oxygenation rates differ with functional guilds, growth stage, root density and depth, and the abundance of aerenchyma in tissue (Sorrell and Brix
2013; Alldred and Baines
2016). Given that the Tully-Murray catchment has substantial nitrogen runoff to the GBR, further research is recommended to evaluate the efficacy of the constructed wetlands in denitrifying runoff and whether improved design and management of the lagoons could lead to greater nutrient removal or retention. Given that denitrification requires systems to be anoxic, designing wetlands to support this function would require a values trade-off as anoxic conditions are not conducive to supporting biodiversity. It may be more appropriate for any future PES schemes to incentivise wetlands for denitrification in some instances and incentivise wetlands for biodiversity in other instances (Canning et al.
2021).
While these lagoons were profitable from a landholder’s perspective when a farm-wide benefit from reduced inundation and the Scheme’s 67% subsidy were included alongside the benefits from improved productivity on the elevated land adjacent to the lagoon, future schemes seeking to create wetlands in a similar way will also likely need to subsidize works. For lagoon creation to be profitable for a landholder, payments would likely need to ensure farmers achieve a benefit to cost ratio of at least 1:1 over the evaluation period (15 years in this study). To achieve a benefit to cost ratio of 1:1 over 15 years without the subsidy for lagoon construction and without including benefits from reduced inundation, but still accounting for the improved productivity on land elevated with excavated spoil, payments would need to be $8076/ha of wetland (in 2019 AUD$) when elevated land can be converted from cattle fattening to four-ratoon cane rotation, or $10,684/ha of wetland (in 2019 AUD$) when elevated land is converted from two-ratoon to four-ratoon cane rotation. Achieving a benefit to cost ratio of 1:1 over a shorter timeframe would be advantageous. It may be possible for future schemes to cover these costs by securing payments for ecosystem services, particularly if the payment scheme has sufficient flexibility to support multiple services, including those benefits that are non-rival and non-excludable (Canning et al.
2021). However, the annual payment rates required to achieve a benefit to cost ratio of 1:1 over a 15-year evaluation timeframe are much higher than the annual gross margins achieved from the land uses that preceded wetland conversion (by a factor of 14 with 2 ratoon cane as the prior land use, or by a factor of 54 with cattle fattening as the prior land use).
While we anticipate the lagoons providing, with various levels of confidence, at least 22 final ecosystem services (Haines-Young and Potschin
2012), not all services provide easily quantifiable benefits to clearly identifiable beneficiaries. Examples include the removal of nutrients and sediment, improvements to physical and mental health, and harvests from transient fisheries. Challenges in quantifying benefits and clearly attributing beneficiaries can make it difficult for these services to be recognized and rewarded through market-like schemes (Costanza et al.
2021). Having a scheme, that funds investment into wetland creation/restoration that does not rely heavily on benefit quantification, can accommodate multiple, bundled ecosystem services (including non-excludable and non-rival services), and is viable long-term to support ongoing maintenance, such as that facilitated by a common asset trust, may be the best option going forward (Canning et al.
2021; Costanza et al.
2021).
The results obtained here indicate a significant need to consider the retention and restoration of well-connected wetlands within the sustainable development of sugarcane landscapes. Wetlands have substantial ability to improve agricultural flood resilience, while providing wildlife habitat and ecosystem services. Recently, Saunders et al. (
2022) identified barriers to the uptake of nature-based solutions, such as wetlands restoration, within Australia, along with key actions to address barriers. Recommendations included: (1) developing fit-for-purpose permitting processes for ecological restoration; (2) improving integrated mapping and classification of coastal ecosystems; (3) conducting research into the effective and risks of using restoration as nature-based solutions; (4) developing national-scale restoration guidelines that can cascade to state and local levels, including guidelines to support climate-resilient restorations; (5) develop decision-support models to help inform which actions to take under what circumstances; and (6) adapt the Restoration Opportunities Assessment Methodology (ROAM) to inform a systematic approach towards prioritization of restoration (Saunders et al.
2022). With respect to restoring wetlands for flood control, further work would be required to identify and prioritize locations where wetlands could be restored and yield a positive return on investment. If returns included accounting for other ecosystem services, then there may also be opportunities for funding from payment for ecosystem service schemes (Canning et al.
2021).
While this study demonstrates the potential benefits of an integrated catchment-scale wetlands restoration scheme, benefits may not be readily transferrable to other catchments without further assessment. Future schemes should use catchment-scale multi-property hydrological modelling to determine the ideal wetland sizes and positions on low-value agricultural land for regulating flood flows. Multi-criteria analysis could then be used to inform site selection by weighting locations and sizes that support the provision of other ecosystem services, such as carbon sequestration, nitrogen removal and supporting biodiversity.
In addition to using hydrological modelling to inform wetland design and connectivity for flood regulation, consideration should also be given to the hydrological connectivity needs of native fish assemblages. Across the studied lagoons, the temporal variation in fish assemblages is significantly influenced by lagoon connectivity with the downstream river, distance from the coast and flood dynamics (Karim et al.
2014; Arthington et al.
2015; Godfrey et al.
2016). Future schemes, particularly those in the Wet Tropics region, should still ensure there is a maintenance of seasonal patterns of flow and connectivity. Furthermore, wetlands should be designed to prevent the growth of exotic ponded-pasture grasses (such as
Hymenachne and
Brachiaria mutica) as these habitats supported the lowest fish species richness compared with other lagoon habitats (Arthington et al.
2015). This would include ensuring depth is greater than the tolerance of ponded-pasture grasses, riparian zones are well shaded with vegetation, and frequent flushing flows are maintained.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.