**4. Discussion**

### *4.1. Predicting the Expected Nekton Changes after Salinity Reduction*

This study developed a model to predict the expected shifts in nekton assemblage resulting from the restoration of a freshwater input in an inner portion of the Venice lagoon, taking into account temporal, geographic, environmental and habitat variability found along natural salinity gradients in the basin. The model calibration phase revealed a significant influence of the sampling season and year on overall assemblage and species biomass, confirming the high variability of temporal dynamics in nekton communities in the Venice lagoon [41,67,70,80]. Reproduction, recruitment and migrations are among the major factors causing seasonal variations in community structure in transitional water ecosystems [69,81]. In addition, seasonal patterns may show marked differences among years due to interannual changes in weather, abiotic conditions and trophic status [80,82,83]. Among the physico-chemical and habitat factors taken into account in the present analysis, only three variables significantly affected biomass of the nekton assemblage as a whole, namely water temperature, water salinity and habitat location within the saltmarsh. This confirms previous observations made on fish communities in the Venice lagoon [67,84,85] and in other Mediterranean transitional water ecosystems [8,82]. As many studies previously pointed out, nekton response to physico-chemical and habitat variables would probably be more complex, and show a significant influence of additional factors (e.g., type of substratum, trophic conditions, tidal cycles and weather) at the species level [80,83,86–89]. Exploring all the factors driving nekton changes at the species level was, however, beyond the scope of this work.

The primary effect on nekton fauna of reducing salinity in the lagoon area investigated would be an overall increase in total assemblage biomass compared to the current conditions, mostly due to the increase in biomass of five species. Among them, the solely resident *Pomatoschistus canestrinii* and the marine estuarine-dependent *Platichthys flesus* and *Chelon ramada* are known to prefer low-salinity, saltmarsh-dominated areas in the inner portions of the Venice lagoon, which they use as recruitment and feeding grounds and, for *P. canestrinii*, as elective reproductive habitats [80,87,90]. Also the estuarine residents *Atherina boyeri* and *Palaemon elegans*, widespread species that can tolerate a wide range of salinity levels in transitional water habitats [7,69], contributed to the expected increase in nekton biomass. The positive response to salinity decrease shown by these species could be related to broader changes in the restoration area, including an increase in trophic status. In the Venice lagoon, trophic levels indeed show marked changes along the salinity gradients, with higher nutrient and chlorophyll concentrations recorded in the more confined portions of the basin [91,92]. Common estuarine species, integrating complex ecological processes such as trophic interactions in transitional water ecosystems, could therefore play a key role as indicators of such changes in the inner areas of the lagoon [41,93–95].

This work emphasises the importance of adopting a functional perspective when employing predictive models in restoration ecology. Species traits, or a functional classification like the guild approach, could provide a better understanding of the processes involved in community colonisation of restored habitats and in overall ecosystem functioning [35,96]. The expected nekton response at the functional level further supports the hypothesis that the restoration of a salinity gradient in the Venice lagoon may cause relevant changes to the food web of the area, with biomass of hyperbenthivores-zooplanctivores, hyperbenthivores-piscivores and detritivores benefiting from the decrease in salinity. The newly-created freshwater input could have a major influence on distribution and community structure of hyperbenthic and planktonic prey due to the salinity decrease, as pointed out by Das et al. [28]. More generally, the nutrient loads from the drainage basin, although bu ffered by the restoration of vegetation, could enhance the productivity of the system, hence the overall availability of food items for nekton species in the area [8,67,97]. Future studies following the restoration of the salinity gradient should therefore take into account the potential changes in trophic status, which could explain a significant portion of variability in functional structure of nekton assemblages in the area.

The model's accuracy proved to be good in predicting the assemblage as a whole, as cross-validation highlighted. While predictive performances varied markedly among species, most of them were

predicted with high accuracy, and could then be included in the set of species to be predicted in the second phase of the work. In particular, this approach identifies a pool of species that best explains the expected assemblage shifts under changing environmental conditions. Overall, the ecological and trophic structure of such assemblage largely corresponds to the typical nekton community composition found in the inner areas of the Venice lagoon and other Mediterranean transitional water ecosystems [11,69,90,98].

As Scapin et al. [36] highlighted, the methodology presented in this work may serve to develop concise tools guiding the managemen<sup>t</sup> and conservation of biodiversity and ecosystem functions. In a perspective of reinstatement of a salinity gradient in the Venice lagoon, for instance, predicting the expected outcomes in terms of nekton structure and comparing them to the observed conditions after the restoration will provide a mean to quantitatively estimate the degree of restoration success. Applying the same protocol in a suitable time frame will also allow the tracking of restoration trajectories towards the endpoint conditions [36]. Similarly, the present approach may also be employed in the framework of WFD, to evaluate the efficacy of managemen<sup>t</sup> responses aiming to enhance the ecological status of hydrologically impaired transitional water bodies.

### *4.2. Implications for Management and Conservation*

The present analysis highlighted that restoring the salinity gradient could have major implications for managemen<sup>t</sup> and conservation of the inner Venice lagoon. Most of the tools developed for the evaluation of ecological status of fish in transitional waters under the WFD, including the one employed in Italian transitional waters, incorporate metrics calculated on assemblage structure and composition, often based on functional guilds [79,99–102]. Given the expected increase in whole assemblage biomass and the shift in taxonomical and trophic structure, significant changes in the ecological status of fish may therefore occur after the restoration. As a result, nekton surveys in the area following the reinstatement of the freshwater input will be of critical importance, in order to measure the actual effect of creating a salinity gradient on the ecological status.

While restoration could benefit the overall ecological status of fish fauna, the consequences on species conservation may vary. The salinity decrease would contribute to creating suitable environmental conditions, particularly in areas characterised by lower salinity levels, for *Pomatoschistus canestrinii*, an endemic species of northern Adriatic coastal lagoons listed in the Annex II of the Habitats Directive. One of the major threats for this species is indeed the loss of estuarine and brackish habitats [80,103]. Other species protected at the European level, namely *Aphanius fasciatus* and *Knipowitschia panizzae*, could in turn be limited to areas that are less influenced by the salinity gradient, or gather in habitats different from saltmarsh edges, such as creeks. Small saltmarsh channels are often preferred by small estuarine resident species, by providing better shelter and higher food availability [80,90]. Overall, this emphasises the importance of preserving habitat diversity, even in a scenario of beneficial restoration, as already pointed out for the Venice lagoon by Cavraro et al. [90] and Scapin et al. [36,70].

The results of this work also indicate a positive influence of restoring a salinity gradient on the recruitment of *Chelon ramada* and *Platichthys flesus*, which exploit transitional water habitats during juvenile stages and represent important resources for local fisheries [41,87,104–109]. Transitional water ecosystems play a central role in supporting the populations of many marine migrant species of commercial value. Multiple environmental and geographical factors contribute to sustain this function, including water turbidity and sediment characteristics, trophic status and prey availability as well as the availability of structured habitats such as saltmarshes, and the degree of habitat connectivity [3,110–112]. In this light, salinity may not be the only factor regulating the entrance and growth of marine migrant juveniles in the Venice lagoon shallow waters and, while the restoration could contribute to support juveniles of *C. ramada* and *P. flesus*, the overall nursery role of the area should be evaluated by future ad hoc studies.

Transitional water habitats rely on the delicate balance between sediment accretion and erosion, as well as on quality and availability of freshwater inputs. The consequences of climate change

could therefore pose additional threats to their survival and quality in the next decades. For instance, the expected sea level rise would result in the loss of major intertidal and subtidal shallow water areas in northern Adriatic coastal lagoons due to erosion and submergence in this century [49,58,113]. Moreover, relevant alterations in temperature and precipitation patterns all over the world are already affecting river flows and overall freshwater availability [114]. In the Venice lagoon, climate change could lead to longer periods of high salinity levels and more marine-like conditions due to more severe drought episodes, resulting in further loss of transitional water features in the inner portions of the basin [115]. Restoration schemes aiming at enhancing the riverine influence on the inner lagoon, such as the Lagoon ReFresh project, may therefore represent a viable way to mitigate the negative effects of climate change on transitional water ecosystems, in particular by preserving suitable environmental conditions for nekton assemblages.

**Author Contributions:** Conceptualization, L.S. and P.F.; methodology, L.S., M.Z. and A.F.; validation, L.S., M.Z. and P.F.; formal analysis, L.S. and M.Z.; investigation, L.S., P.F.; resources, P.F.; writing—original draft preparation, L.S.; writing—review and editing, All the authors; visualization, L.S. and M.Z.; supervision, P.F. and A.S.; project administration, A.B. and R.B.B.; funding acquisition, A.B., R.B.B. and A.S.

**Funding:** This research was funded by European Union's LIFE+ financial instrument (grant LIFE16 NAT/IT/000663—LIFE LAGOON REFRESH, which contributes to the environmental recovery of a Natura 2000 site, SIC IT3250031—Northern Venice Lagoon). Field data used in this study were collected under various projects funded by the Italian Ministry of Education, Universities and Research (PRIN gran<sup>t</sup> 2009W2395), by Corila (Consorzio Ricerche Lagunari) and by European Union's LIFE+ financial instrument (grant LIFE12 NAT/IT/000331).

**Conflicts of Interest:** The authors declare no conflicts of interest.
