Spatially Well Structured Mangroves Fish Communities of the Persian Gulf; a Functional Perspective

Functional diversity is one of the most important aspects of biodiversity studies. In this work, the functional diversity of two mangrove-associated fish communities in the Persian Gulf is evaluated, including 34 species from the estuary community and 23 species from the creek/mud community. For each community, tree-based functional diversity indices were calculated, namely, FD (Functional Diversity, representing functional richness), MPD (Mean Pairwise Distance, representing dispersion), MNTD (Mean Nearest Taxon Distance, representing evenness). Also, multidimensional indices were calculated, including FRic (functional richness), FEve (functional evenness), FDiv (functional divergence), and FDis (functional dispersion) The comparison between the observed values and the expected values indicates the low functional diversity of the estuary community, whereas, the functional diversity of the species inhabiting creeks and mud flats was not significantly lower or higher than the expected functional diversity. Moreover, the dispersion of functional traits in creek/mud species was higher and the evenness was lower than the estuaries community. This can be related to the more homogenous habitat of the estuary environment, in comparison with the creeks and mudflats. Also, we can attribute the low functional diversity of the estuary community to both the homogenous structure and the preliminary situation of the Persian Gulf. Dendrogram-based analyses show four functional guilds in both communities and were used to the recognition of the most important species for the functional diversity in each community. We found that the most important functional traits that explain the functional diversity in the estuary community were the general traits of size and position in the vertical column in water, however, in creek/mud communities more specialized traits such as the feeding habit and the migratory status explain most of the diversity.


Introduction
Biodiversity studies are one the most important analyses for understanding the ecological structure of marine ecosystems and their conservation status.Numerous studies introduced several diversity concepts based on species richness (Scott et al. 1987), genetic and phylogenetic relationships between species (Coates et al. 2018;Faith 2018), and functional traits (Petchey and Gaston 2002).Functional diversity analyses are based on sets of traits in understudied organisms that have ecological impacts on the environment (Loreau and Hector 2001;Cornelissen et al. 2003;Hooper et al. 2005) and ecological service of a region (Díaz et al. 2007).Currently, function-based approaches have an increasing role in the study of different systems such as microbial communities (Bhattacharyya et al. 2023), fungi (Plett et al. 2023), plants (Wang et al. 2023), invertebrates (Head et al. 2018;Zhang et al. 2022), and vertebrates (Ausprey et al. 2022;Mao et al. 2021;Walsh et al. 2022;Whitfield et al. 2022).Integrating functional diversity with other concepts in biodiversity (for instance, phylogenetic diversity and species richness) can produce novel conclusions about the history and assemblage of understudied communities (Flynn et al. 2011;Tucker et al. 2018;Lin et al. 2021).In particular, the relationship between functional and phylogenetic diversity is controversial (Saito et al. 2015;Owen et al. 2019;Molina-Venegas et al. 2021).In addition, functional diversity studies can be 79 Page 2 of 15 used to evaluate the ecological processes, ecological service, and the conservation status of a region (Cadotte et al. 2011;Shah Esmaeili et al. 2022).Numerous indices have been defined to calculate different facets of functional diversity (Schleuter et al. 2010;Pla et al. 2012a).Among them, dendrogram-based indices are one of the oldest and still one of the most used indices in functional ecology studies.The first introduced dendrogram-based index in these studies is simply called FD (Functional Diversity), and is the sum of the branch length of the trait dendrogram of the community (Petchey and Gaston 2006;Petchey et al. 2007).Other indices comprise convex hull, which is a hyper-volume of trait ranges and the position of each species in the trait space (Barber et al. 1996;Cornwell et al. 2006), whereas, other concepts have been formulated in functional diversity studies, based on richness, evenness, and divergence (Villéger et al. 2008).Nevertheless, when we want to choose appropriate indices we should be cautious, for among the diverse functional diversity indices, many of them are strongly correlated with each other and are basically redundant (Swenson 2014b).
Connecting to the Gulf of Oman, by the Strait of Hormoz, the Persian Gulf is a recently formed shallow basin with a maximum depth of 100m and an average of 35m (Barth and Khan 2008).This biogeographical and physical setting causes high salinity and temporal fluctuations and results in a considerably harsh and undesirable environment for its native fauna and flora (Bayani 2016;Sheppard et al. 2010).However, studies on the species composition and richness of several taxa in this area suggest that there is a noticeable taxonomic distinctness in this region, surprisingly with various endemic species (i.e.Price and Izsak 2005;Naderloo 2017).Mangrove forests of this region are young ecosystems with two tree species, Avicennia marina and Rhizophora mucronata, which have been both naturally grown and artificially planted along the coastlines of the Persian Gulf and the Gulf of Oman (Zahed et al. 2010).Avicennia marina is the most dominant and widely distributed where accounting around 97% of the coverage, while R. mucronata is just naturally grown in Gaz mangroves.These forests are the subject of negative human activity in these areas, however, sustainable development and management of mangrove ecosystems with works on their biodiversity and conservational status have been fruitful and increased these mangrove forests in recent years (Milani 2018).Recently, increasing attention is paid to the assessment of biodiversity concepts (other than species richness) in the mangrove ecosystems of the Persian Gulf.The focus is mostly on the functional diversity of macrobenthic communities (Delfan et al. 2021;Hajializadeh et al. 2020Hajializadeh et al. , 2022;;Nozarpour et al. 2023) and phylogenetic diversity of invertebrates such as brachyuran crabs (Raeisbahrami and Naderloo 2023).Also, mangrove vertebrates have been the subject of biodiversity studies, though these works are mainly limited to the faunistic surveys with little attention to the preliminary analyses of species richness and abundance, not functional or phylogenetic diversity assessments (Braulik and Sharif Ranjbar 2009;Ghasemi et al. 2012;Behrouz 2014;).
Mangrove forests are stable habitats for many marine and terrestrial species (Nagelkerken et al. 2008).Fish are one of the most important taxa among the vertebrate communities of the mangrove forests.They use mangrove ecosystems as nursery beds ( Sheridan and Hays 2003;Nagelkerken 2009) and have a major role in the sustainability of the food web in this region (Shahraki et al. 2014;Shahraki 2015).Moreover, ecosystem service and fisheries have a strong linkage in mangrove ecosystems (Manson et al. 2005).In addition, fish community markedly contributes to the movement of external organic material sources from adjacent habitats to the mangrove forests (Kruitwagen et al. 2010;Igulu et al. 2013).A recent checklist reported 743 fish species from 131 families in the Persian Gulf (Eagderi et al. 2019).Biodiversity studies of the fish species in mangrove forests of the Persian Gulf were conducted in Jask, Khalasi, Gabrik (Kamrani et al. 2016;Hashemi, unpublished MSc thesis), as well as Gaz creek (Zahedi et al. 2019), Qeshm, Khamir (Hashemi, unpublished PhD thesis), Tiab, Kolahi, Koleghan (Hashemi, unpublished MSc thesis), and Gwatr and Bahookalat (Danehkar 2006).193 species from 69 families are reported from these mangrove ecosystems in the Persian Gulf.There are two communities of mangrove-associated fish in the Persian Gulf including species inhabiting shallow and narrow creeks or mudflats and other species living in the outer estuaries of the mangrove forests.Noteworthy, there are three mudskipper species (Periophthalmus waltoni, Boleophthalmus dussumieri, and Scartelaos tenuis) in the first community that are burrower in mud, therefore, have a noticeable impact on physical, chemical, and ecological structure of the ecosystem (Colombini et al. 1996;Clayton and Snowden 2000;Ravi 2013).The second community is relatively more homogenous, with more species than the former group.
In this study, we examined the functional diversity of both creek/mud and estuary fish species of the mangrove forests of the Persian Gulf.We calculated the alpha diversity of each community and the beta diversity between them.Moreover, we investigated the functional structure of these two groups to understand the role of each species in the functional composition of each community.Also, we determined the most important functions and species responsible for the stability of the ecosystem and its functional diversity.Our goal is to explain the difference of estuaries and creek/mud fish communities in respect of their functional structure and composition.

Data Collection
Based on mangrove fish checklists of the Persian Gulf, abundance tables, and sampling sessions conducted in shallow creeks, we provided a list of 193 species from 12 mangrove forests scattered in the Iranian coastline of the Persian Gulf.These mangrove forests are located in Khamir, Qeshm, Koleghan, Tiab, Kolahi, Gaz, Jask, Khalasi, Gabrik, Gwatr, and Bahookalat (Fig. 1).
For the creek/mud community we included all the observed and sampled species for the analysis.Regarding the estuary community, as an efficient and common practice, we selected either the most abundant species that compromise 80% of the total abundance (Grime 1998) or species that was reported in three or more than three stations.Eventually, 34 species from the estuary community and 23 species from the shallow creek/mud were selected with 6 common species (Table 1).For a better analysis of null models and random estimation, 38 additional species were included.These added species were recorded from at least 2 stations.However, we did not calculate functional diversity indices and the results of other analyses for these additional species.

Functional Traits
Functional traits were chosen based on their contribution and importance for ecosystem service (Pla et al. 2012b).These traits represent multitude aspects of the ecological influence of each species on its environment.We defined five sets of traits (size, feeding habit, migration habit, vertical position in column water, and reproduction pattern) with at least two modalities for each trait to cover the life cycle, position in the food web, and effect on the abiotic environment of the ecosystem.Ultimately, we examined the diversity and the structure of 15 modalities for the communities under consideration (Table 2).
We used a mixed coding approach to assign modalities to each species.For 13 trait modalities we used the fuzzy coding approach, as species can display two or more modalities of a functional trait simultaneously, with various intensities (Chevene et al. 1994).In this approach, we assigned a score, ranging from 0 to 3, to each trait and species (Supplementary Table 1).If a species shows a trait during its whole life, we assigned the score of 3.And if there is no affinity, the score of 0 was assigned.Intermediate states were represented by 1 or 2 according to the affinity of the species to the considered modality.Only for the reproduction type we used binary coding (representing oviparity and viviparity).Traits and modality data were collected from FishBase online database (Froese and Pauly 2000), peer reviewed literatures and personal expert consultations.

Dendrogram Based, Multidimensional, and Distance Matrix Based Indices
Trait dendrograms were constructed based on our trait data matrix, for each community and total species (Fig. 2).As we had mixed trait variables in our data set, Gower distance was used to assemble a trait distance matrix (Gower 1971).UPGMA method was used to construct trait dendrograms for the species.To edit and visualize the dendrogram, we used FigTree v1.4.4 (FigTree 2016).The constructed tree was used for the calculation of FD as one of the most basic and common tree-based indices of functional diversity (Petchey et al. 2007).This index represents the range of functional traits in our community; therefore, it can be interpreted as the functional richness of the community (Pla et al. 2012a).FD can be calculated using the below equation: In this equation, l i is the branch length of species i in the trait dendrogram of a community composed of n species.
Other indices of functional diversity that were calculated based on the Gower distance matrix were MPD (Mean Pairwise Distance) and MNTD (Mean Nearest Taxon Distance), which respectively can be used as the proxies for functional dispersion and functional evenness of a community (Pla et al. 2012a).MPD is calculated through the following equation: In this equation, δ is the trait distance between species i and j in a community composed of n species.
Similarly, the MNTD index can be calculated by the below equation: In this equation, min δ represents the minimum trait distance of species j and other species in a community composed of n species.

MNTD =
∑ n i min n randomization options both in trait matrix and functional dendrograms.

Beta Diversity Analysis
For the functional Beta diversity between the two aforementioned communities, we used D pw (pairwise trait distance) and D nn (Nearest neighbor trait distance) measures (Swenson 2014d).Both indices are the modified versions of Beta diversity concepts introduced in Rao (1982).D pw index measures the dissimilarity of functional diversity between each species of the two communities.Calculating D pw is based on the below equation:  In this equation n(c 1 ) and n(c 2 ) are the number species in the communities c 1 and c 2. And δ(ij) stands for the trait distance between species i in the c 1 community and species j in c 2 community.D nn index measures the difference of the most similar species between two communities.D nn was calculated based on the following equation: Here, minδ(ic2) is the minimum trait distance between species i in the c 1 community and all the species of c 2 community.Likewise, minδ(ic2) is the minimum trait distance between species j in the c 2 community and all the species of c 1 community.

Statistical Analyses and Random Models
After excluding the reproduction type, due to its binary nature, four moments of the distribution of trait modalities (mean, standard deviation, skewness, and kurtosis) were calculated for each community (Supplementary Table 2).Test of normality was conducted for the distribution of each trait.Based on the normality test results, Mann-Whitney Test were performed to compare the distribution of traits in both communities.P-value of the differences are reported.
Also, a principal component analysis was conducted on the trait database to determine the contribution of each trait in the functional diversity of its respective community and the most important functions of each species composition.Eigenvalues of functional traits with each component were calculated to understand the functional composition of communities.Scree plots were produced for each community.Statistical tests and PCA were conducted through IBM SPSS Statistics (Version 26).
Additionally, using randomized functional dendrograms and trait matrices, we compared the observed functional diversity indices (FD, MPD, and MNTD) to the expected values in creek/mud and estuary fish communities (Swenson 2014c).Randomization of the functional trait dendrogram and matrix was performed with 10,000 iterations.And the expected mean and standard deviation values of FD, MPD, and MNTD were calculated for each species composition.Observed values of these indices in each community were calculated as well (Table 3a).The P-value of the difference between the observed and expected indices is reported.
Moreover, to determine the contribution of each species to the functional richness of both communities we eliminated the species in consideration from the trait dendrogram and calculated the updated FD metric of the new community.
Then we compared the new FD with both the observed and expected functional richness of the original community.
All the analyses concerning functional diversity indices were carried out through R software v.4.2.2 (R Development Core Team (2011).The calculation of FD, MPD, and MNTD was conducted through ape v.5.7-1 (Paradis and Schliep 2019) and picante v.1.8.2.(Kembel et al. 2010) packages.We used "comdist" and "comdistnt" functions in picante to calculate the D pw and D nn indices respectively.The elimination of species, for the sake of assessing the functional value, was accomplished by using "drop.tip"function in ape package.

Traits Moments of Distribution and PCA
Comparing the distribution of functional traits between the two communities indicates significant difference (p-value < 0.05) in some trait modalities (Supplementary Table 3).Migratory status is significantly higher in estuary species (p-value = 0.04).Also, carnivorous feeding habit is significantly lower in creek/mud community (p-value = 0.02).In addition, it can be seen that limnivores feeding habit and small size are more common in creek/mud community.
79 Page 8 of 15 Principal component analysis (PCA) was conducted for both communities.Six components were recognized in creek/mud species, explaining the 80% of the variance in data.Also, five principal components were recognized for estuary communities that explain 80% of variance in the data.In the creek/mud species (Supplementary Table 4), the most important traits based on their eigenvalues are limnivore feeding habit and mud dwelling for the first component, small body size and migration status for the second component, top dwelling (with positive association) and bottom dwelling (with negative contribution) for the third component, reproduction types for the fourth component, and herbivory and medium body size for the fifth component.In estuary species composition, traits have evener eigenvalues for each principal component (Supplementary Table 5).Nevertheless, the most important traits of the components are the reproduction types, migratory status, and very large body size for the first component, small body size and carnivorous and limnivorous feeding habits for the second component, large body size (negative), top dwelling habit for the third component, middle dwelling and for the fourth component, and, finally, middle and large size for the last component.Traits associated with feeding habit and habitat are more important in creek/mud species.On the other hand, position in water column, size, and migratory status are more important to explain the trait variance in estuary community.

Functional Diversity Metrics
To calculate and interpret FD, we considered polytomies in the resulted dendrogram tree similar to polytomies in phylogenetic trees (Swenson and Worthy 2018).FD was calculated for both creek/mud (FD = 3.18) and estuary (FD = 3.19) communities.Considering the different species numbers of these two communities, these similar FD values need a plausible explanation.Compared with creek/mud species, observed MPD and MNTD was lower in the estuary community.In comparison with null models, observed FD, MPD, and MNTD were all higher than their respective expected values in the creek/mud community (Figs. 3a,4a,5a).However, the differences were not statistically significant in these cases.On the other hand, the observed FD, MPD, and MNTD values of the estuary community were significantly lower (p-value < 0.05) than their expected values (Figs. 3b,4b,5b).
Results of the multidimensional analysis of the various functional diversity indices (FRic, FEve, FDiv, and FDis) support the FD, MPD, and MPD results (Table 3a).FRic value of the estuary community (0.0187) is higher than the FRic value of the creek/mud community (0.0131), though, as in FD values the difference is not significant.FEve (representing evenness) in concordance with MNTD, is higher in creek/mud community (= 0.82) than in estuary species (= 0.69).At last, FDis value in creek/mud species (= 3.90) is higher than that in the estuary species (= 3.13), which is completely in accordance with the MPD results for the both communities.

Beta Functional Diversity
Despite assessing the functional diversity of each community and comparing the observed values with the expected ones, analyzing the differences between these two communities can provide valuable information.D pw (representing the overall similarity of two communities) in this study equals to 0.28, and D nn (representing the similarity between the most similar species in two communities) is 0.06.This result shows that the overall similarity between the two communities not considerably high.Also, there are species with distinct functional traits in both communities (Table 3b).

Functional Value
As a measure of functional richness, we considered the effect of species elimination on FD.MPD and MNTD were not considered in this analysis due to their insensitivity to species numbers (Swenson 2014c).The most important species for the functional richness of the creek/mud communities were Chanos chanos, Hemiramphus archipelagicus, Tylosurus crocodilus, Acentrogobius dayi, and Arius arius, in order of priority.It should be noted that the elimination of the three mudskipper species in this community, has a substantial influence on the functional richness of creek/mud species (decreasing from FD = 3.18 to FD = 2.69).For the estuary community, Maculabatis gerrardi, Gerres oyena, Epinephelus coioides, Plectorhinchus pictus, and Scomberoides commersonnianus were recognized as the most important species for functional diversity (Table 4).

Functional Structure
Considering the functional trait dendrogram of each community, we can determine four functional guilds in both communities (Fig. 1).However, Maculabatis gerrardi and Chanos chanos cannot be included in any definitive functional guild in estuary community.Functional values of these two species support this exclusion, as the elimination of Maculabatis gerrardi has the most negative influence on the FD value of this community, and Chanos chanos is among the most important species for functional richness.
In comparison with the estuary community, creek/mud functional guilds are more distinct from each other, with the distinguished group of mudskippers (Boleophthalmus dussumieri, Periophthalmus waltoni, and Scartelaos tenuis) plus the small and burrower species, Acentrogobius dayi.Due to the specific habit of burrowing in mud bottoms, this distinction is completely expected.In fact, as was noted before, the elimination of these four species has a huge detrimental impact on the functional richness of this community.Limnivory (FL) and mud dwelling (HI) represent this burrow making behavior and it can be seen that these modalities have a considerable contribution in principal component analysis for the creek/mud community.PCA shows that other than the burrow dwelling habit, migration type has a considerable contribution to the functional structure of the fish species inhabiting creeks and muds.The results support the importance of resource delivery from outside of forests to this ecosystem (Kruitwagen et al. 2010).Also, PCA results suggest that more general and variable traits such as body size modalities and vertical position are the most important functional traits in estuary fish species.In addition, the dispersion of traits and their contribution to each principal component in estuary community is more uniform in comparison to creek/mud species.This can be attributed to the more homogenous environment of the estuaries.Although species from the creek/mud community inhabit various habitats such as shallow creeks, deep creeks, near the tree roots, and mudflats, we cannot see this heterogeneity in the habitat of mangroveassociated species sampled from estuaries (Ellison 2019;Hongwiset et al. 2021).The lower FD index of functional richness for this species supports the explanation.

Functional Diversity Measures
Despite the larger species number of the estuary community (34 species in contrast with 23 creek/mud species) comparison of the FD index, as a measure of functional richness, between the two communities shows a slight difference.Two different hypotheses can explain this similarity, namely, 1) the high functional richness of creek/mud species, or 2) the low functional richness of the estuary community.Comparing the observed FD values of each community with its expected value indicates the second explanation.The observed FD for creek/mud species does not have a significant difference from the expected value.However, this is not the case for the estuary community, which its observed FD is significantly lower than the expected value.
In addition, the comparison of the observed values of other functional diversity indices (MPD and MNTD) with their respective expected values implies the low functional diversity of the estuary community.Recent studies show that the accumulation of functional diversity is non-identical with species diversity and mainly depends on the age of the area (Oliveira et al. 2016;Jarzyna and Stagge 2023).Knowing the young status of the Persian Gulf and its ecosystems (Sheppard et al. 2010), we can account the lack of proper time for the formation of separate niches (Chesson 2003;Munoz et al. 2023) for the low functional diversity of fish in mangrove associated estuaries.As mentioned earlier, creek and mudflat environments in mangrove forests have more diverse and separate habitats, therefore, there will be more completely separate niches in these regions.Hence, we expect the functional diversity of their fish inhabitants reach its optimum sooner.
Considering MPD and MNTD indices, as measures of functional dispersion and evenness, the results suggest that the dispersion of functional traits in the estuary community is significantly low (observed MPD is lower than the expected value).We can attribute this result to the harsh environment of the mangrove associated regions in the Persian Gulf (Bayani 2016).A recent study suggests the larger contribution of environmental filtering than the competition for the assembling process of mangrove crabs in this region (Raeisbahrami and Naderloo 2023), therefore, we can expect the species in the Persian Gulf occupy restricted preliminary niches that results in a cluster of trait values (Peres-Neto 2004).In homogenous environments, such as mangrove estuaries habitats, this process can produce communities with a limited functional richness and dispersion (Biswas and Mallik 2011).The lack of significant difference in the aforementioned functional diversity indices in the creek/mud community shows a rather neutral assembly in these regions.Also, we should bear in mind the more heterogeneous and separated niches in this area which can produce a community with varied and distinct functional traits.
The Beta diversity D pw index shows a relatively high difference between the two communities (D pw < 0.3).Also, D nn indicates the existence of species with a rather distinct functional role in both communities (this can be shown by the D nn < 0.3).This result is caused due to the species with completely distinct functions in creek/mud species.Mudskippers are the most obvious examples, for they have a complete set of functional traits that cannot be found in estuary species.

Functional Value
In this study, the contribution of each species to the functional richness of the communities was calculated.The five most important species for the functional diversity of the creek/mud community are representatives from the four distinct functional guilds.Surprisingly, from three species of mudskippers, none of them were among the most important species.This we can relate to the substitution of each species with the functional role of another mudskipper.For instance, the elimination of Periophthalmus waltoni can be compensated by the existence of Scarlateus tenuis or Boleophthalmus dussumieri.However, the elimination of Acentrogobius dayi causes a significant decrease in the functional richness of the creek/mud community.This species is among the burrow dwelling species of tree covered zone, yet is not a mudskipper, which are mainly in non-vegetated nearby mud flats.Therefore, its deletion cannot be completely compensated by the mudskippers.Hemiramphus archipelagicus and Tylosurus crocodilus are both carnivores, with large or very large body sizes.Hence, they are key species in the food web of these region.Arius arius and Chanos chanos are both migratory species, therefore, have a substantial influence on 79 Page 12 of 15 the transmission of resources for mangroves and adjacent ecosystems.
In the estuary community, the elimination of Maculabatis gerrardi has the most negative impact on the functional richness of the community.This is due to its peculiar set of functional traits, because the species is a large stingray that swims over sandy and bottoms.Again, Chanos chanos is among the most functionally important species of the community.It should be noted that Chanos chanos is the only living representative species of the family Chanidae, so it has a special phylogenetic status in this community as well as its functional situation.This concurrency can be due to the much important role of environmental filtering in the assembling process of the community, and a sign of a strong phylogenetic signal in fish communities (Swenson 2014a), a result which is reported from the studies on the brachyuran crabs of this region (Raeisbahrami and Naderloo 2023).Future comparative phylogenetic studies can reveal more about the ecological and evolutionary structure of the mangrove fish communities in the Persian Gulf.

Conclusion
There are two different fish communities in the Persian Gulf mangrove forest in estuaries and creek/mud habitats.Functional diversity analysis of these communities reveals two distinct functional compositions in estuaries and creek/ mud communities.Functional diversity of estuaries species is relatively low, due to its homogenous environment and its preliminary stage of formation.However, the functional diversity of creek/mud community is in the range of the expected values.Functional structure of the both communities reveal four functional guilds that are in creek/mud species are based on more specified traits (feeding habit and habitat), despite the more general based guilds in estuaries species.Therefore, we can infer that fish communities in shallow creeks-muddy substrate have reached their optimum functional diversity.Also, important species were specified based on their contribution on the functional diversity of each community.Considering these results, conservation policies can be made to protect the mangrove ecosystem of the Persian Gulf which have a harsh environment and are the subject of many biodiversity threats in this region.

Fig. 1
Fig. 1 Mangrove forests in the Persian Gulf.Adjoining stations which were considered as a single site are represented by the same color.Example of sampling positions for estuaries community (represented by the dotted line) and creek/mud community (solid line) is shown

Fig. 2
Fig. 2 Dendrograms of a) estuaries, and b) creek/mud communities' functional traits.Four functional guilds are specified by different colors in both trees

Fig. 3 Fig. 4 Fig. 5
Fig. 3 Histograms of random FD values for each community.a) estuaries species and b) creek/mud community.Observed FD values are represented by a black line

Table 2
Functional traits and modalities used for the calculation of functional diversity metrics.A brief description for each trait and how codes are assigned to each modality is given Reported total size of the species.When maximum size belongs to another modality, we assigned 1 for the maximum reported modality, and 2 for the average reported modality BS Medium (30-50) BM Large

Table 4
The importance of each species for the functional richness of estuaries and creek/mud communities.Numbers represent the FD index value of the community after the elimination of the corresponding species