Restoration promotes ecological functioning through greater complementarity

Restoration projects are increasingly widespread and many promote habitat succession and the diversity and abundance of faunal communities. These positive effects on biodiversity and abundance may extend to enhancing the ecological functioning and resilience of previously degraded ecosystems, but this is rarely quantified. This study surveyed a 200‐ha restoring coastal wetland and three control wetlands in the Maroochy River, eastern Australia to compare the effects of wetland restoration on the consumption of carrion and the biodiversity, abundance, and functional diversity of functionally important fish and crustaceans. Carrion consumption by fish and crustaceans was measured every 6 months from spring 2017 until spring 2021 for nine events using a combination of baited cameras and scavenging assays. We found restoration improved rates of carrion consumption and the biodiversity, functional diversity, and abundance of scavenger species. Despite positive effects on the diversity of scavengers and carrion consumption, the abundance of two species, longfin eels (Anguilla reinhardtii) and mud crabs (Scylla serrata), was the most important predictors of carrion consumption rates. The spatial distribution of carrion consumption was concentrated in areas with high saltmarsh extent, moderate to high mangrove extent, and high salinity, which also resembled the distribution of both longfin eels and mud crabs. We show that restoration can promote the rates of key ecological functions but that increases to functions are likely to be characterized by low functional redundancy and greater complementarity. Therefore, maintaining or increasing the abundance of functionally important species should become a key objective in future restoration projects.


Introduction
The capacity of ecosystems to resist and recover from disturbance is contingent upon the maintenance of multiple ecological functions that underpin resilience (Risser 1995;Decker et al. 2017;Manning et al. 2018).The rates of most ecological functions, including herbivory, predation, and pollination, is dependent on the presence, diversity, and distribution of functionally important species (Brose & Hillebrand 2016;Martin et al. 2018).Many ecosystems have, however, been degraded by the actions of multiple human stressors, and this has impacted negatively on both populations of functionally important species and the ecological functions they support (Vitousek et al. 1997;Sévêque et al. 2020).The combined effects of urbanization, overharvesting, and pollution have led in reductions to the biomass, abundance, and diversity of significant herbivores, predators, and scavengers in many regions, leading to losses in the key ecological functions they perform in marine, freshwater, and terrestrial landscapes (Vaughn 2010;Estes et al. 2011;Reich et al. 2012).Consequently, restoring key ecological functions to degraded ecosystems has become a central goal of natural resource management initiatives (Ripple & Beschta 2007;Montoya et al. 2012;Kormann et al. 2016).The potential for restoration projects to deliver on these objectives is determined by social and economic factors and also by the combined effects of spatial and physical processes that affect the recovery trajectories of functionally important species and shape the characteristics of ecological niche space (Pander & Geist 2013;Kaiser-Bunbury et al. 2017).Interactions between these processes and restoration performance are, however, rarely considered when optimizing project design and monitoring programs, and this likely detracts from the capacity of many initiatives to deliver wider benefits for ecological functioning (Falk et al. 2006;Jones & Davidson 2016;Palmer et al. 2016).
The spatial context, condition, and diversity of recovering habitats at restoration sites modifies how they are colonized by flora and fauna, and this in turn can structure the trophic composition of animal assemblages and determine their capacity to support pivotal ecological functions (Kemp et al. 1999;Paillex et al. 2013;Noh et al. 2017).For example, restoring vegetation at sites nearer to large, mature forests can boost the abundance and diversity of visiting frugivores and nectarivores and augment the ecological services these species deliver (Reid et al. 2014;Freeman et al. 2015).Similarly, restoring larger oyster reefs close to adjacent mangroves, saltmarshes, or seagrasses can promote the recruitment and diversity of invertebrates and fishes and enhance functions such as predation and herbivory (Grabowski et al. 2005;Gilby et al. 2019).These effects of restoration on ecological functions might be linked to the recovery of local biodiversity, or to increases in the abundance of functionally important species (Grabowski & Peterson 2007;Catterall 2018).Rates of ecological functions might increase with species and functional diversity when a range of colonizing species partition niche space to share resources and do so by performing distinct functions at elevated levels (i.e.high diversity, high functional redundancy) (Bellwood et al. 2003;Loreau 2004).By contrast, ecological functions might increase with the abundance of particular species, when functional diversity is low and these dominant taxa maintain certain functions at a high level (i.e.low diversity, limited functional redundancy) (Bellwood et al. 2004;Lohbeck et al. 2016;Henderson et al. 2020).Alternatively, species that use similar resources might co-exist by accessing these at different times or locations, resulting in elevated rates of particular functions in different contexts (i.e.high functional complementarily) (Cardinale et al. 2012;Loreau & de Mazancourt 2013;Brown et al. 2015).The effects of restoration on ecological functions are, therefore, likely influenced by interactions between diversity and dominance and by the level of redundancy and complementarity in animal assemblages in recovering assemblages (Palmer et al. 1997;Carlucci et al. 2020).
Low redundancy and high complementarity are thought to be common features in the food-webs of many coastal seascapes, where a relatively small pool of species often dominates a particular ecological function (Bellwood et al. 2003;Micheli & Halpern 2005;Gilby et al. 2017).Well-known examples include the functions of herbivory and predation in coastal wetlands, seagrass meadows, oyster reefs, kelp forests, and coral reefs, which are frequently done by just one or two species (Worm et al. 2006;Micheli et al. 2014;Duncan et al. 2019).High complementarity and dominance in these functions might reflect the trophic flexibility and broad habitat requirements of many coastal species, particularly taxa that migrate among ecosystems to consume a diversity of prey items (Mouillot et al. 2007;Henderson et al. 2022;Mosman et al. 2023b).It is not yet clear, however, whether and to what extent the recovery of species, functional diversity, and species abundance contribute to the delivery of ecological functions in restoring ecosystems.
This study tested for potential effects of ecological restoration on ecological functioning by measuring relationships among species richness, species abundance, and functional richness and the ecological function of carrion consumption in a restoring coastal wetland.The consumption of carrion (i.e.dead animal carcasses) was selected as the focal ecological function because a diversity of animals scavenge carrion in estuarine food webs, and the rate of carrion consumption is therefore likely linked to increases in species diversity and functional complementarity with wetland recovery (Wilson & Wolkovich 2011;Olds et al. 2018;Goodridge Gaines et al. 2020).Scavengers control necromass decomposition and recycle nutrients in food webs, thereby playing an important role in restoring wetlands where the changes to inundation frequency and salinity levels that promote wetland recovery might also inadvertently lead to some mortality (Wilson & Wolkovich 2011;Barton et al. 2013;Porter & Scanes 2015).Fish and crustaceans were used as model scavengers in this experiment because they dominate the consumption of carrion in estuaries, respond rapidly to wetland restoration, and the rate at which they scavenge animal carcasses has been linked to changes in water and habitat quality and impacts from fishing and urbanization (Boys & Williams 2012;Porter & Scanes 2015;Rummell et al. 2023b).We hypothesized that the rate of carrion consumption would increase at the restoration site over time and that these increases would be linked to the spatial distribution of species richness and functional richness across the recovering wetland.

Study Area and Restoration Site
We examined whether, and how, restoration modifies ecological functioning by contrasting patterns in carrion consumption and the functional composition of scavenging assemblages at one restoration site (i.e.Yandina Creek Wetlands) with three control sites (i.e.Coolum Creek, Rocky Creek, and Twin Ridges Creek) in central eastern Australia (Fig. 1).The Yandina Creek Wetlands is a 190-ha restoring coastal wetland complex, that was previously drained and disconnected from the wider Maroochy River by a matrix of tidal gates and drainage channels in the 1950s to support sugarcane and cattle farming (Brown et al. 2018;Iram et al. 2022).These farming practices ended in the early 2000s, and the site was abandoned until purchased by Unitywater in the early 2010s.Passive restoration actions commenced via the opening of tidal gates in 2018 (Rummell et al. 2023a(Rummell et al. , 2023b)).All four wetlands were located within the Maroochy River catchment, between 10 and 20 km upstream of the river mouth, and each experience semi-diurnal tidal regimes (with a maximum tidal height of 1.41 m), variable salinity (0-30 ppt), and modest annual rainfall (annual mean 1483 mm) (Bureau of Meteorology 2021; Maritime Safety Queensland 2021; EHMP 2022).The three control sites represent alternative potential end-point targets for restoration of the Yandina Creek wetlands; one is a designated conservation area (Coolum Creek), one is a local environment reserve (Twin Ridges Creek), and one abuts agricultural land (Rocky Creek), which makes them suitable controls for indexing possible pathways post restoration.
We used a bi-annual sampling approach to measure carrion consumption and quantify the functional composition of fish and crustacean assemblages, with surveys and experiments conducted in autumn and spring each year from 2017 to 2021.Two monitoring events were completed prior to the start of restoration at Yandina Creek (i.e.time: À6 and 0 months), and seven additional monitoring events were completed after restoration commenced (i.e.time: 6 to 42 months) (n = 9).All survey events were conducted during monthly spring tides to standardize for both the extent and duration of wetland inundation.On each event we quantified carrion consumption rates and scavenger assemblages with baited cameras; measured the context, area, and succession of vegetation communities (i.e.mangrove, saltmarsh, and common reed) using remote sensing; quantified connectivity with the Maroochy River; and described changes in physicochemical water quality (i.e.salinity, turbidity, pH, dissolved oxygen, and temperature) using a YSI water quality probe.

Measuring Carrion Consumption and Censusing Scavenger Assemblages
Scavenging assays were used to measure changes in rates of carrion consumption and index ecological functioning at the restoration site and the three control wetlands.Experimental assays have been used widely to quantify a diversity of ecological functions in coastal seascapes, including herbivory with fleshy algae deployments (Yabsley et al. 2016), predation with tethered fish deployments (Martin et al. 2018) and carrion consumption with the deployment of fish carcasses (Porter & Scanes 2015).This study followed the approach of previous studies that have deployed fish carcasses to index the carrion consumption by estuarine scavengers (Olds et al. 2018;Henderson et al. 2020).Here, we used baited remote underwater video station (BRUVS) to survey both scavenger assemblages and scavenging rates.BRUVS are a standard technique for surveying fish and crustacean assemblages in coastal seascapes (Dorman et al. 2012;Santana-Garcon et al. 2014;Duncan et al. 2019) and have been used to measure rates of carrion consumption in a diversity of estuaries (Webley 2008;Porter & Scanes 2015;Olds et al. 2018;Henderson et al. 2020).Two pilchards of known weight were deployed on the upper surface of each BRUVS bait bag.Pilchards were then re-weighed immediately after BRUVS retrieval, and carrion consumption was indexed as the change in weight during each one-hour deployment using digital scales with AE0.01 g accuracy.The contribution of each scavenger species to carrion consumption was determined through analysis of BRUVS footage by measuring the proportion of each carcass consumed (g/hour) by each scavenger species visually.
BRUVS were deployed in a stratified manner at 16 locations in the Yandina Creek wetlands to capture change across the entire drainage channel and tidal gate network, and at 10 random locations in each control wetlands over nine monitoring events (n = 414).Each BRUVS was positioned simultaneously at least 200 m apart and comprised of a high-definition GoPro Hero camera, mounted on a 5 kg weight, and with a 0.5-m arm supporting a bait bag filled with 500 g of pilchards (Sardinops sagax) (Gladstone 2012;Borland et al. 2022a).All deployments lasted for 1 hour and were done within 2 hours of high spring tides to maximize visibility and the extent of wetland inundation and standardize for effects of tidal variation.Scavenger richness and abundance were quantified from the BRUVS footage using the standard MaxN statistic (i.e. the maximum count of each species observed at one time), which was recorded for all species that were observed to feed on deployed pilchard carcasses (Murphy & Jenkins 2010).
The functional richness of scavenging fish and crustacean assemblages was used to index scavenger functional diversity (Villéger et al. 2008;Mouillot et al. 2013).Functional richness describes the overall fullness of functional trait space and is calculated as the amount of functional trait space that is occupied by an assemblage, with higher functional richness indicating a greater diversity of functional traits and a higher potential to deliver a range of ecological functions (Gagic et al. 2015).Functional richness for fish and crustaceans species were indexed by obtaining data on the traits of all scavenger fish and crustacean species identified in BRUVS footage via fishbase and sealifebase (Boettiger et al. 2012;Palomares & Pauly 2023).Data were extracted to quantify variation in functional traits that correlate with lifecycle, diet, functional groups, and morphology, including functional group, body shape, trophic level (i.e.2-3.98), lifecycle group, body length, body depth, head length, pre-orbital length, and eye diameter (Elliott et al. 2007;Henderson et al. 2019;Borland et al. 2022b).However, only four of these traits were available for crustaceans; these traits included functional group, trophic level, lifecycle group, and body length.Data on scavenger fish and crustacean traits and the composition of scavenger fish and crustacean assemblages (extracted from BRUVS surveys) were then combined using the fundiv and FD packages in R (Petchey & Gaston 2002;Mason et al. 2005).

Quantifying Variation in Wetland Connectivity, Habitat Conditions, and Water Quality
To examine how changes in spatial context, habitat conditions, and water quality influenced ecological functioning, we measured the level of connectivity with the adjoining estuarine seascape, spatial extent of estuarine habitats, three-dimensional properties of wetland terrain and physicochemical water quality conditions at each location where fish and crustanceans were surveyed (Table S1).Connectivity between locations on the Yandina Creek wetlands and the adjoining estuarine seascape was quantified using least cost path accumulation (LCPA) analysis calculated using a resistance surface (RS) raster in QGIS (Weber & Allen 2010;Dutta et al. 2016).LCPA was calculated as the accumulative cost of moving through each cell of the RS from each sampling location at the Yandina Creek wetland to the nearest creek outside of the restoration site (see Supplement S1).The spatial extent of estuarine habitats at each site were quantified by adopting a highly accurate approach to measure the change in the area of mangrove, saltmarsh, and reed communities that include a high-resolution (0.5-2 m) Worldview 2 satellite imagery, object-based image analysis, and a random forest classifier (sensu Rummell et al. 2022).The three-dimensional properties of each site were also quantified using digital elevation models (Geoscience Australia) by calculating the slope and elevation of each site (Table S2) (Borland et al. 2021).The area of estuarine habitats, as well as slope and elevation, was calculated in QGIS using 50 m buffers around each sampling location.This scale was chosen to minimize the potential for spatial autocorrelation and emphasize fine-scale effects of habitat recovery on fish and crustacean assemblages (Olds et al. 2012;Gilby et al. 2018;Duncan et al. 2019).Water quality was measured in situ at each site using a YSI multiparameter water quality probe to record variation in turbidity, dissolved oxygen, pH, temperature, and salinity (Table S2) (Dibble et al. 2015).

Data Analysis
Generalized linear models (GLMs) in R were used to test for the potential effects of wetland restoration on ecological functioning (indexed as the rate of carrion consumption), and the species richness, functional richness, and total abundance of scavengers assemblages.These GLMs tested for differences between site (fixed factor, four levels; Yandina Creek Wetlands, Coolum Creek Wetlands, Rocky Creek Wetlands, and Twin Ridges Creek Wetlands) and time (continuous variable; À6 to 42 months).
Generalized additive models (GAMs) were used to identify nonlinear relationships between scavenging species that dominated carrion consumption, and for which variation in abundance was positively linked to variation in consumption rates.GAMs were then used to test whether, and how, significant scavengers and key functional metrics (i.e.species richness, functional richness, and abundance of scavenger assemblages) were linked to increases rates of carrion consumption at the restoration site (Wood 2015;Barton 2019).Finally, GAMs were then also used to test for potential complementarity in the function of carrion composition, by examining how changes in the level of wetland connectivity, area of recovering wetland habitat, and physicochemical water quality conditions combined to modify the spatial distribution of significant scavengers and key functional metrics at the restoration site.We first tested for conconcurvity among the 11 spatial, physical, and biological properties of study sites at the restoration site (Tables S2 and S3).Highly correlated variables (>0.70 r), including elevation, slope, pH, and temperature were excluded from further analyses (Table S3).All models were then checked for the distribution of residuals, homogeneity of variance, and over-dispersion and were fitted with either Poisson, Negative Binomial (if overdispersion was present), or Gaussian distributions with log link functions in the MASS package (Ripley et al. 2013).Model overfitting was reduced by running all possible combinations of four or fewer factors and by restricting individual models to four or fewer knots (i.e.individual polynomial functions that combine to smooth GAMs (Wood 2006).Models were compared using the Akaike information criterion (AIC) corrected for small sample sizes using the MuMIn package, and best-fit models were selected as those with the lowest AIC value (Barton 2019).The relative importance of variables in each model was calculated by summing weighted AIC values across all models containing the variable of interest; values closer to one indicate a greater and more consistent association of a predictor and the response variable (Burnham & Anderson 2002).

Effects of Restoration on Ecological Functioning and Key Functional Metrics
The ecological function of carrion consumption increased following the start of restoration and continued this trend over the 4 years during which monitoring was conducted (Tables 1 &  S3; Fig. 2A).Three key metrics that describe the functional composition of the scavenger assemblage, scavenger richness, functional richness, and abundance, also increased at the restoration site over this same period (Tables 1 & S3; Fig. 2B-D).The rate and magnitude of increases in carrion consumption and the three functional metrics were typically greater at the restoration site than at the control wetlands, and after 4 years all metrics were highest at the restoration site (Table S4; Fig. 2).

Linking Changes in Ecological Functioning to Key Species and Functional Metrics
Nine species of fish and crustaceans, including yellowfin bream (Acanthopagrus australis), mud crab (Scylla serrata), longfin eel (Anguilla reinhardtii), blue catfish (Neoarius graffei), banded toadfish (Marilyna pleurosticta), crescent grunter (Terapon jarbua), long-clawed prawn (Macrobrachium novaehollandiae), estuary perchlet (Ambassis marianus), and empire gudgeon (Hypseleotris compressa) consumed carrion at the restoration site, with the majority being piscivores, zoobenthivores, or omnivores (Elliott et al. 2007;Webley 2008;Froese & Pauly 2018).Longfin eel and mud crab dominated scavenging and combined to consume 91% of all deployed carrion (Fig. 3A).Changes in the abundance of both longfin eel and mud crabs were the only significant predictors for increases in the rate of carrion consumption over time (Tables 2 & 3; Fig. 3B).Variation in the abundance of no other species was linked to temporal increases in the rate of carrion consumption.

Effects of Connectivity, Habitat, and Water Quality on Ecological Functioning
Changes in the ecological function of carrion consumption were linked to salinity variation at the restoration site (Table S5; Fig. 4A).The greatest increase in carrion consumption rates occurred in downstream areas of the site where salinity was consistently high (Table S5; Fig. 4A).The abundance of longfin eels increased over time, but the greatest gains occurred in locations with a moderate area of mangroves (1000-2000 m 2 ) and a large area of saltmarsh (>2000 m 2 ) (Table S5; Fig. 4B).Increases in the abundance of the two key scavenger species also occurred in downstream areas of the site.The abundance of mud crabs also increased at the restoration site over time, with the greatest gains occurring in areas where wetland connectivity was high (i.e.resistance values <2000) (Table S5; Fig. 4C).Increases in the species richness and functional richness of scavengers were also greatest in locations with a large area of mangroves (Table S5; Fig. S1).By contrast, increases in the abundance of scavengers were greatest in locations with high salinity, moderate to high connectivity and a large area of mangroves (Table S5; Fig. S1).

Discussion
Restoration initiatives commonly target the recovery of a focal habitats and associated biodiversity and ecological functions (Rey Benayas et al. 2009;Gann et al. 2019).Successful wetland restoration projects can enhance key ecological functions, but it is not clear whether, and how, these functional benefits of restoration are linked to changes in species diversity or dominance (Zhao et al. 2016;Lechene et al. 2018).The results of this study illustrate a rapid increase (i.e.within 4 years) in the rate of carrion consumption at the restoration site and highlight the significance of functional complementarity in wetland recovery.
Increases in the rate of carrion consumption were linked to positive changes in the functional composition of food webs at the restoration site (i.e.greater species richness, functional richness and abundance of scavenger species), but were primarily related to the distribution and abundance of two key scavengers.Longfin eels and mud crabs dominated carrion consumption, and changes in the rate of this function were positively correlated with increases in their abundance.Longfin eels were dominant scavengers in areas where the area of mangroves and saltmarsh increased over time, whereas mud crabs were the primary consumers of carrion in locations with high connectivity to adjoining wetland habitats (Rummell et al. 2022).These findings demonstrate the importance of functional complementarity and habitat diversity in recovering wetlands.Many restoration projects focus on supporting the recovery of a single target ecosystem (e.g.mangroves), whilst also hoping to have additional benefits for allied ecosystem services (Aerts & Honnay 2011;Wortley et al. 2013).We suggest that broadening the scope of restoration planning to focus on the joint recovery of a multitude of well-connected habitats is essential for the expansion of ecological niche space to support the restoration of multiple ecological functions.
The findings of this study highlight positive effects of wetland restoration on the ecological function of carrion consumption, and illustrate that these functional benefits of restoration were linked to high functional complementarity, with two dominant scavengers (i.e.longfin eels and mud crabs) consuming carrion in different contexts.The abundance of both functionally important species increased at the restoration site over time, but did so in different places.Longfin eels became more abundant in areas where mangroves and saltmarsh were greatest and mud crabs in locations with high connectivity to adjoining wetland habitats.Both species are highly aggressive, large-bodied omnivores that regularly consume carrion but also have unique functional traits that likely shape their efficiency as scavengers (Webley 2008;Arai 2016).For example, longfin eels are mobile swimming predators that typically reside in fresh and upper-estuarine waters, but also use marine and estuarine habitats, including mangroves and saltmarshes, as feeding sites (Cucherousset et al. 2007;Breber et al. 2008).They access these intertidal habitats to feed, and can derive a significant proportion of the carbon in their diet from saltmarsh plants, but must retreat with the receding tide to shelter in subtidal areas with high habitat complexity (Eberhardt et al. 2015;Lechêne et al. 2018).By contrast, mud crabs are benthic omnivores that inhabit mudflats and shallow muddy channels, where their burrowing is not obstructed by dense vegetation or debris.Because their burrows retain water they can remain in intertidal areas throughout the tidal cycle (Hill et al. 1982;Alberts-Hubatsch et al. 2016;Leoville et al. 2021).Both habitat types present ample opportunities for these scavengers as tidal and freshwater flows transport an abundance of carrion and detritus, originating either on land or from local vegetation communities that experience die-back with tidal inundation, to them for consumption (Kwak & Zedler 1997;Matthews & Endress 2010).It is likely that the habitat-specific, and complementary, distributions of these two scavengers has enabled them to partition the consumption of carrion in both space and time, supporting increases in this ecological Figure 3. Contribution of fish and crustacean species to carrion consumption rates at the Yandina Creek Wetlands site (A).Best-fit GAMs illustrating how carrion consumptions rates were related to the abundance of longfin eels and mud crabs (B) (see Table 3 for model details).
Table 2. Summary of best-fit generalized additive models (GAMs) testing for correlations between the rate of carrion consumption at the restoration site and the: abundance of all species that consumed carrion; total scavenger abundance; scavenger richness; and scavenger functional richness.Numbers in parentheses are important values and highlight the significance of each species or trait for the function of carrion consumption.Higher importance values indicate a greater contribution to the function.

Best Fit Models and Importance
R 2 df AICc Longfin eel (1.00) + Mud crab (0.96) 0.31 6 1310.5  Olds et al. 2018;Henderson et al. 2020).If populations of these two scavengers continue to expand, the level of competition between them for carrion might also increase.However, the potential for densitydependent interactions will likely be moderated by further expansion of mangrove and saltmarsh habitats and by the widening of channels and mudflats with continued tidal inundation.These results highlight the significance of promoting habitat diversity and connectivity in restoration and suggest that features might be significant for underpinning rates of ecological functioning in recovering land and seascapes (Escribano-Avila et al. 2014;Derhé et al. 2018).
The effects of wetland recovery on ecological functioning were related to increases in biodiversity at the restoration site, with greater carrion consumption occurring as the species richness, functional richness and abundance of scavengers grew over time.Relationships between biodiversity and an ecological function like this are widely cited as the reason natural resource managers should focus their efforts on preserving full complements of species, rather than particular taxa, in conservation (Cardinale et al. 2012;Mori et al. 2013;Mosman et al. 2023a).In this instance, however, the recovery of both species and functional diversity were only weakly linked to the function of carrion consumption because two scavengers together consumed 91% of all carrion deployed at the restoration site.The dominance of this small pool of scavengers might indicate that there has been little recovery in the overall niche space of local food webs.This is, however, unlikely given the concomitant increases in species and functional richness, or alternatively could reinforce the idea that limited functional redundancy might be a common feature in low-diversity ecosystems (Gamfeldt & Roger 2017;Manning et al. 2018;Olds et al. 2018).Nevertheless, this finding concurs with patterns recorded in the early stages of oyster reef recovery, and in some estuarine conservation areas, where increasing rates of ecological functioning (i.e.predation and scavenging) were linked to changes in the abundance of a single functionally important species (Duncan et al. 2019;Goodridge Gaines et al. 2020;Jones et al. 2020).Restoration might, therefore, be an effective strategy for enhancing rates of some ecological functions, but these functional benefits could also be short-lived if species dominance is common in recovering food webs (Bellwood et al. 2003;Gamfeldt & Roger 2017;Renzi et al. 2019;Henderson et al. 2020).When functional redundancy is low, impacts to populations of functionally dominant species, such as urbanization, declining water quality or harvesting, have the potential to cascade rapidly to ecosystem structure and function (Lohbeck et al. 2016;Olds et al. 2018;Goodridge Gaines et al. 2020).
The dominant scavengers in this study, longfin eels and mud crabs, are harvested in both recreational and commercial fisheries, and also migrate through tidal wetlands in response to high freshwater flow events to reproduce (Elliott et al. 2007;Webley 2008;Alberts-Hubatsch et al. 2016;Koster et al. 2021).Both species are easily captured in large numbers when they are moving across seascapes, and this indicates that the ecological function of carrion consumption might be particularly sensitive to fishing pressure or to any actions that impinge on fish passage across the restoration site (Webley 2008;Arai 2016).In situations like this, the capacity of restoration to support ecological functions might be contingent on the integration of restoration planning decisions with the broader spatial management of functionally important species.
Rates of ecological functioning at the restoration site surpassed those recorded at all three reference wetlands.These trends were also detected for fish and crustacean species richness and the abundance of fisheries targeted species at the same site and timeframe in Rummell et al. (2023a).It is likely that wetland restoration augmented these aforementioned responses by fish and crustacean assemblages due to the joint effects of high resource availability (i.e.mortality of pre-existing plant and animal communities), greater connectivity and improved accessibility to unoccupied ecological niche space (Duncan et al. 2019;Olson et al. 2019).Urban structures and terrain variation have previously elicited similar supplementation effects on the ecological functions performed by many generalist fish species in modified estuaries (Olds et al. 2018;Borland et al. 2022aBorland et al. , 2022b)).It is anticipated as biophysical (i.e.hydrology, geomorphology, settlement of habitat forming species) and ecological factors (i.e.interspecies competition, food web dynamics, changes in primary producers) progress, the rates of ecological functions performed by fish and crustaceans will stabilize and resemble natural reference sites.This is because, in the early phases of restoration typically resource availability is high but diversity is low, and overtime resources become less rich but more diverse (James et al. 2020).We recommend future research that investigates the ecological and functional dynamics of restoring ecosystems and determine what and how spatial drivers limit or promote the rates of these functions over large temporal scales (i.e.10+ years) (Renzi et al. 2019).
The recovery of ecological functions is a common objective in restoration, but there is an absence of empirical data to describe the where, and how, functional benefits arise from ecosystem repair in coastal seascapes (Blüthgen & Klein 2011;Cadier et al. 2020).Our results demonstrate that restoration actions can promote an ecological function in a degraded wetland and show that the functional benefits of restoration were linked to increases in abundance of functionally important species rather than effects of biodiversity.They also highlight the significance of functional complementarity and habitat diversity in supporting the recovery of ecological functions in restoring coastal wetlands.Restoration can positively affect rates of ecological functions by enhancing seascape connectivity and by promoting access to new habitats and resources, which broadens ecological niche space and might have consequences for a variety of ecological functions.Restoration projects seeking to deliver increases in the rates of ecological functions, as co-benefits from the repair a target ecosystem/s, might therefore experience greater success when their site selection, design, and monitoring phases prioritize the recovery of habitat connectivity and landscape heterogeneity.

Figure 1 .
Figure 1.The position of survey locations at the restoration site (Yandina Creek wetlands), and the three control wetlands (Coolum Creek, Twin Ridges Creek and Rocky Creek) in the Maroochy River catchment, eastern Australia.

Figure 2 .
Figure 2. GLMs illustrating recovery of ecological functioning, scavenger diversity, scavenger functional diversity and scavenger abundance, at the restoration site (Yandina Creek wetlands) relative to changes at multiple control sites.

Figure 4 .
Figure 4. Best-fit GAMs illustrating how effects of habitat context, connectivity and water quality combined to shape changes in the rate of carrion consumption (A), and the abundance of mud crabs (B) and longfin eels (C) at the restoration site (see TableS4for model details).

Table 1 .
Summary of generalized linear models (GLMs) testing for recovery of functional indicators (i.e.carrion consumption and the diversity, abundance, and functional richness of scavengers) at the restoration site (Yandina Creek wetlands), relative to changes at multiple control sites.

Table 3 .
Summary of generalized additive model testing for the effects of longfin eel and mud crab abundance overtime on carrion consumption rates.