香京julia种子在线播放

The impact of humans on natural environments occurs in a wide variety of the world’s ecosystems, mainly derived from changes in land use, pollution, and exploitation of natural resources (Sanderson et al., 2002; Foley et al., 2005; Ellis et al., 2010). These human alterations have led to widespread habitat loss, changes in biodiversity, species invasions and extinctions, and declines in ecosystem condition (Butchart et al., 2010; Ceballos et al., 2015). Consequently, shifts in the functioning of ecosystems have been detected and, in some cases, changes in some ecosystems services that are of value for human societies (Dobson et al., 2006; Cardinale et al., 2012).

Some ecosystems are so degraded that natural recovery is almost impossible in short and medium terms, even when the impacts leading to their collapse have been removed. Human-mediated restoration initiatives are often applied, becoming one of the only tools capable of reverting the loss of an ecosystem (Dobson et al., 1997; Holl & Aide, 2011). Ecological restoration has emerged as a critical measure for the management of heavily degraded ecosystems and is now a global priority (Aronson & Alexander, 2013; Suding et al., 2015). Supporting this idea, the Convention on Biological Diversity has set restoration targets (Secretariat of the Convention on Biological Diversity, 2014), and in 2019 the “United Nation Decade of ecosystem restoration 2021-2030” was declared. In fact, the number of undertaken restoration actions has increased notably over the past decades (Wortley et al., 2013; Carlucci et al., 2020), evolving from mitigation – removing human disturbances to allow for unassisted recovery – to active restoration – direct interventions in an effort to intentionally influence the trajectory of recovery (Simenstad et al., 2006; Elliott et al., 2007).

Some ecosystems have traditionally been more actively restored than others (e.g., terrestrial forests and grasslands) and several years of practice have determined what characterizes successful restoration. In general, vegetation structure, species diversity, and ecosystem functions have been suggested as the three key attributes to assess restoration outcomes (Ruiz-Jaen & Aide, 2005; Wortley et al., 2013). However, while the trajectory of vegetation structure (e.g., vegetation cover, density, and biomass) and species diversity (e.g., richness and abundance) following restoration is well understood, knowledge regarding the recovery effect on ecosystem functions is still limited (Brudvig, 2011; Kollmann et al., 2016). Many evaluations of restoration success first focus on the development of the target species because it is broadly assumed that following its recovery it will add structure to the habitat (Geist & Hawkins, 2016). In doing so, other associated organisms should benefit from the increase in structural complexity and the ecosystem will eventually recover its processes and functions (Geist & Hawkins, 2016). However, recent studies have demonstrated that in some cases either species composition or ecosystem processes have not been fully re-established, pointing to the need to empirically test the relationship between the recovery of a target – usually foundation species – and the processes and functions of the ecosystem (Rey Benayas et al., 2009; Moreno-Mateos et al., 2012; Crouzeilles et al., 2016). In this line, assessment of ecosystem functional recovery should include not only target species viability but also quantifiable properties to describe the community in terms of structure alongside ecosystem functions (Montoya et al., 2012).

Trait-based ecology enables the analysis of the influence of species identity and composition on ecosystem functions and processes (Garnier et al., 2004; McGill et al., 2006). Functional traits are determined by morphological, physiological, and biological characteristics of the species and are considered relevant to influence the environment, affecting ecosystem properties and services (Violle et al., 2007; Díaz et al., 2013). Thus, functional diversity indices based on trait information improve the understanding of the functioning of an ecosystem (Mouillot et al., 2013), and when used together with traditional taxonomic-based indices they can provide a comprehensive evaluation of restoration projects (Cadotte et al., 2011; Montoya et al., 2012; Laughlin, 2014). Indeed, the application of functional trait approaches to ecological restoration is receiving growing attention worldwide (Carlucci et al., 2020), with some examples mainly focused on vegetation (e.g., Piqueray et al., 2015; Engst et al., 2016; Zirbel et al., 2017). To achieve in-depth and thorough assessments of functional recovery, studies should include as many species as possible and a broad combination of traits encompassing different facets of ecological processes. To this effect, comprehensive assessments are still needed for many ecosystems, probably because the required surveys to gather information about different species groups and species traits is challenging and time consuming.

In marine coastal regions, brown seaweeds can create complex ecosystems often described as macroalgal forests, where macroalgae are the major contributors to primary production, providing food and shelter for other organisms, and consequently hosting an important biodiversity hotspot (Steneck et al., 2002; Smale et al., 2013). In the Mediterranean Sea, macroalgal forests are characterized by canopy-forming species of the genus Cystoseira sensu lato (including the genera Cystoseira C. Agardh, Gongolaria Boehmer and Ericaria Stackhouse; Molinari & Guiry, 2020), representing the highest level of complexity. However, several Mediterranean macroalgal forests have been declining over the last decades, such that previously widespread Cystoseira s.l. species have become locally extinct or their populations have been severely reduced and fragmented (e.g., Thibaut et al., 2005; Thibaut et al., 2015; Mariani et al., 2019). The first example of a self-sustaining macroalgal population after restoration in the Mediterranean Sea was reported by Verdura et al. (2018). In March 2011, a single restoring action allowed the recovery of a previously extinct population of the macroalga Gongolaria barbata in the Bay of Maó, Menorca (NW Mediterranean) ( Figure 1 ). The species thrived in the area one hundred years ago, but it became locally extinct following a reduction in water quality related to waste water effluents and harbour operations (Sales et al., 2011). Even when water quality was improved due to management measures involving stopping wastewater dumping, G. barbata populations were unable to recover. The success of the restoring action was first assessed considering the density and size structure distribution of the target species, which showed comparable values to those obtained in the reference population (Verdura et al., 2018) ( Figure 1 ). The presence of fertile individuals and new recruits confirmed that the restored population was self-maintained. More recently, a high-resolution cartography was carried out to describe the expansion patterns of the population, finding an increase of three orders of magnitude in the area occupied by G. barbata since the restoring action took place (Gran et al., 2022).

The lack of recovery evidence is even more severe in marine ecosystems, since marine restoration science is very young in comparison to its terrestrial counterpart (Elliott et al., 2007; Rey Benayas et al., 2009; Kollmann et al., 2016, but see Saunders et al., 2020), representing only 7% of all restoring actions available in the literature (Blignaut et al., 2013). In the context of macroalgal forests, the few worldwide successful examples of active restoration are focused on the recovery of canopy-forming species (Whitaker et al., 2010; Verdura et al., 2018; Fredriksen et al., 2020; Layton et al., 2020; Cebrian et al., 2021). To our knowledge, only two studies have introduced the measurement of the re-established diversity, addressed by species composition and their abundance (Ling, 2008; Marzinelli et al., 2016). Further, macroalgal restoration projects are in general maintained over time spans of less than 2 years (Eger et al., 2022).

Here, we assess for the first time the recovery of ecosystem functions in a ten-year restored macroalgal forest dominated by Gongolaria barbata. Using both traditional taxonomical indices and a functional trait approach, we test the assumption that after the reintroduction of the primary and structural species the associated biodiversity will recover and, in turn, the associated ecosystem functions and processes. We also hypothesised that the time since the presence of the structural species will influence the outcomes, with longer periods showing better recoveries. To do so, we considered the entire algal species composition and included 14 functional traits related to major aspects of the macroalgal species ecology. The main aim of the present study is to provide evidence of macroalgal taxonomical and functional recovery by describing a valid and solid approach which can be easily replicated and compared among marine and terrestrial ecosystems.

In this study, we provide the first example of recovery of ecosystem functions, under a functional-trait framework, after an active restoration in a macroalgal forest. The patterns in species and functional diversity reported here suggest that restoration is an important strategy in mitigating biodiversity losses and reinstating part of the lost functionality. Due to the limited restoration examples, marine restoration success is often based on short-term periods (1-2 years on average, Bayraktarov et al., 2016; Kollmann et al., 2016, but see Saunders et al., 2020). In terrestrial projects, however, the time elapsed since restoration started has been a key factor to reach similar conditions as those found in reference systems (Martin et al., 2013; Crouzeilles et al., 2016). Considering that the recovery of many marine ecosystems can take up to 15-25 years (Jones & Schmitz, 2009; Borja et al., 2010), the current time span used in most marine restoration activities might be insufficient to reliably evaluate the outputs. To this effect, where short-lived and high-turnover species are the target biological components of an ecosystem, full recovery could be achieved within shorter timescales (see examples in Borja et al., 2010). However, this is not the case of the structural components of macroalgal forests, where some of the foundational species are considered mid-/long-lived seaweeds (see examples in Schiel & Foster, 2006; Smale et al., 2013). Knowledge regarding the dynamics of the target species (e.g., growth rates, turnover, and longevity) should guide restoration efforts, pointing to the importance of long-term monitoring assessments for assemblages constituted of species with long life spans, such as some Mediterranean macroalgal forests (e.g., Ballesteros et al., 2002; Ballesteros et al., 2009; Capdevila et al., 2016).

Also derived from the short-term evaluations, especially in marine ecosystems, restoration success often considers parameters related to the target species’ survival, with common measures of growth or density of the individuals (Bayraktarov et al., 2016; Jacob et al., 2018; Cebrian et al., 2021). In terrestrial ecological restoration, vegetation structure is also the most common quantitative indicator (Ruiz-Jaen & Aide, 2005; Gómez-Aparicio, 2009) together with community measures such as the species richness, the biodiversity indices and the species composition (Gatica-Saavedra et al., 2017; Hughes et al., 2017). In this study, species biomass, mean number of species, and species diversity (Shannon-Weiner index) of the restored locality reached the values of one reference locality (“Reference 1”). The restored locality even surpassed mean values of the other reference locality (“Reference 2”), supporting restoration success. Variations in biodiversity of the two reference localities could be explained by differences in local factors, since a priori similar communities might be subjected to different environmental conditions, resulting in species composition variations (Bernhardt-Römermann et al., 2015; and see examples in Matthews et al., 2009; da Costa et al., 2019). Thus, local environmental factors (e.g., type of substrate or sediment loading) and biotic factors (e.g., grazing pressure) might determine species composition and diversity. For this reason, the inclusion of more than one reference locality in restoration assessments is highly recommendable to account for the variability of natural communities and to describe different, but possible, restoration outcomes (Gann et al., 2019; Prach et al., 2019).

The macroalgal species assemblage of the restored and expansion localities differed from the non-restored locality, and were also different when compared with the macroalgal forests of reference ( Figure 8 ). This trend has also been found in other restoration projects, where restored systems do not strictly follow the expected direction toward references (e.g., Sluis, 2002; Matthews & Spyreas, 2010; Suganuma et al., 2014). Time will determine whether the restored locality will eventually mirror a reference macroalgal forests or it will remain structurally and functionally dissimilar. Recent restoring frameworks and directives suggest that ecosystem restoration should first aim to recover past ecosystems where viable, but where not, a shift in objectives targeting novel but still improved ones should be adopted (Jackson & Hobbs, 2009; Gann et al., 2019). Restored ecosystems may have new combinations of species that have not occurred previously, but that could ensure the maintenance of the structure and functions of the system (Suding & Gross, 2006; Choi et al., 2008). Even so, historical knowledge is highly valuable and should always be used as a guide which accepts multiple trajectories, considers future environmental changes, and seeks for the recovery of ecosystem services (Higgs et al., 2014; Martin, 2017; Gann et al., 2019).

We also demonstrated an increase of functional richness (Fric) with restoration ( Figure 8 ), indicating that the recovery of macroalgal species filling vacant functional space not occupied prior to restoration, effectively occurred. Interestingly, we observed that the species composition, the distribution and the abundance of FEs and the variability within traits in the area where G. barbata is naturally spreading (“Expansion Area”) is highly similar to those of the restored locality ( Figure 8 ). These similarities are due to the most dominant and abundant species in the restored locality being already present in the area of expansion. Despite this, functional diversity was lower since a high number of less abundant species are still lacking, showing that the recovery of some species and part of the biomass does not directly translate into a full recovery of all other species and their functionality ( Figure 8 ). These results highlight, again, the need of long periods to allow the recovery of the whole community after restoration.

In the restored locality, there were more functional entities (FEs) and they were more distanced in the functional space, deriving from a greater range of categories within each trait (that is, more varied traits). In particular, after restoration, the macroalgal assemblage shifted from having only two main taxa and a small number of trait categories (e.g., two morphological forms, only solitary species, and one category of height and width) to a more heterogeneous assemblage involving six morphologies, the presence of gregarious and colonial taxa, and varied sizes with some large species but also other smaller ones. In addition, the restored assemblage included species with longer life spans and later reproductive maturity. In the same way as terrestrial plants, an increase in species architecture and size suggests a superior ability to compete for light (Reich, 2014) and it is sometimes associated with higher accumulation of standing biomass (de Bello et al., 2010). Similarly, recovering long-lived species along with morphological complexity is likely to increase the creation of three-dimensional biogenic habitats, increasing, for instance, long-term nutrient storage (de Bello et al., 2010; Teixidó et al., 2011) ( Figure 8 ). Adding stable structure and complexity also offers shelter for other organisms and reinstates the nursery capacity of these communities (Ørberg et al., 2018; Fulton et al., 2019). Hence, restoration not only benefited the macroalgal species but also probably lead to further consequences for other groups (i.e., invertebrates and fishes) ( Figure 8 ). Nonetheless, it is important to consider that one process or function is not controlled by a single trait belonging to a specific trophic level (e.g., primary producers, herbivores or carnivores) but by some key traits related to various trophic levels, which are simultaneously involved in the control of multiple functions (de Bello et al., 2010; Pistón et al., 2019). Like the restored locality, both reference macroalgal forests were also characterised by a high variability in the different functional trait categories. A greater range of trait categories in these reference forests could increase the likelihood of some species responding differently to variable conditions and perturbations (e.g., extreme climatic events), which also contributes to assemblage stability and resilience and to the maintenance of long-term ecosystem functioning (Díaz & Cabido, 2001; Cadotte et al., 2011). Last, and of interest, is that the non-restored locality and one reference macroalgal locality (Reference 2) showed similar low diversity values through the application of taxonomical information, the reference macroalgal locality doubled the diversity results when using functional information, underlining the utility and complementarity of these metrics.

Predictable outcomes are a major goal of restoration ecology (Choi, 2007; Rey Benayas et al., 2009; Matthews & Spyreas, 2010), thus studies assessing ecosystem functioning through the lens of trait-based approaches can help to promote more generalizable outcomes (Suding et al., 2008). Even so, few studies have tested these ideas and, to our knowledge, they have never been applied in marine benthic ecosystems. We have demonstrated that if we are dealing with the structural species of an assemblage, a single restoring action can lead to the recovery of the associated biodiversity and, in turn, to the recovery of its functioning. Our study showed that the application of functional metrics together with traditional taxonomical indices provide valuable information for the understanding of macroalgal ecosystem functioning, and this approach has the potential to be extended and replicated in other restoration evaluations.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material . The R code used to perform all the statistical analyses can be found in Git Hub: https://github.com/cgalobart/Macroalgal-restoration-functional-traits.

EC conceived the ideas; CG and EC collected the data; EB and CG identified the species and defined the traits; CG, RG and EC performed the data analyses; CG and EC drafted the manuscript. All authors contributed to the article and approved the submitted version.