Persistence of benthic communities: a case study from the Valli di Comacchio, a Northern Adriatic lagoonal ecosystem (Italy)

Michele Mistri Data collected from a three-year continuing sampling program were used to study the persistence of the macrofaunal communities living in the Valli di Comacchio (44 41 N, 12 10 E), a complex of shallow-water brackish lagoons located in the southernmost part of the Po River deltaic area (Italy). Univariate and multivariate analyses were used to assess the variability and persistence of the community at each of the four sites investigated. A total of 46 benthic taxa were identified from bottom grab samples: Annelida was the most abundant group, followed by Crustacea and Mollusca. The abundance of some species exhibited marked temporal fluctuations. At three out of four sites, abundance fluctuation were unpredictable in terms of both timing and magnitude. There was little evidence that factors operating at lagoon-wide scale contributed to variability in community structure. The balance of taxa within each community was probably most affected by forcing factors acting on a local scale. The perception of macrofaunal communities as being persistent through time decreased as the analysis moved from coarser (presence/absence) to finer (absolute abundance) levels of numerical resolution. Analyses showed that macrobenthic assemblages at some sites were both persistent and variable: this paradox reflected viewpoints of different analytical scales. Because of this, a multi-level approach is probably more useful in gaining insight into community persistence and variability. 1054-3139/02/040314+09 $35.00/0 - Introduction In ecological theory, the concept of disturbance and its role in structuring populations and communities has been given considerable attention (Pickett and White, 1985). Physical disturbance (flood and droughts, unusually high or low temperatures, salinity changes) that result in changes to population structure or resources availability are common in shallow-water coastal and estuarine habitats (Nichols and Thompson, 1985; Beukema, 1992; Giangrande and Fraschetti, 1996; Koutsoubas et al., 2000). Disturbance may influence the structure and functioning of benthic animal assemblages by directly or indirectly killing, displacing or damaging components of the community. The effects depend on the frequency and intensity of the disturbance (Connell, 1978; but see Collie et al., 2000, and literature cited herein, for disturbance on aquatic sedimentary communities), whereas the responses depend on the inherent resilience and stability of the community. Unstable communities are often the most resilient (Holling, 1973) because they contain species adapted to variable environmental conditions; a number of studies have demonstrated that unstable communities are more likely than stable communities to return to their previous composition and structure following some form of disturbance (Pickett and White, 1985). Stability has been shown to exhibit some distinct attributes (Begon et al., 1986), such as persistence and variability. Persistence is the constancy in some parameter (number of species, taxonomic composition, population size) of the system over time, while variability is the degree to which a parameter fluctuates over time. The ability to measure such attributes is crucial to our understanding of patterns or processes in benthic faunal assemblages because persistence in species abundances can be viewed as evidence for a deterministic organization of communities, whereas Valli di Comacchio Porto Navigabile Magnavacca Campo variability often implies a greater role of stochastic processes in structuring communities (Grossman et al., 1982). Data collected from the first three years of a continuing sampling program undertaken to monitor the macrofaunal communities living in the Valli di Comacchio, northern Italy (44 41 N, 12 10 E), are presented here. In the last decades, the Valli underwent important anthropogenic impacts, from land reclamation to the effects of contamination on the remaining areas (Sorokin et al., 1996). In this paper, macrobenthic community persistence over the duration of the sampling period as well as patterns of temporal variability in community structure are discussed. Materials and methods Study site The Valli di Comacchio (Figure 1) are a large (100 km2) complex of shallow-water (with depth ranging from 0.51.5 m) brackish lagoons located in the southernmost part of the Po River deltaic area (northern Italy). This lagoonal system is nowadays connected with the Adriatic Sea by two marine channels (Portocanale and Logonovo), which enter the lagoonal complex through the small Valle Fattibello (which also receives a small amount of continental waters) and then flow into the wider basin of Valle Magnavacca. A third channel, the Gobbino which formerly connected the easternmost part of Valle Magnavacca with the sea, is impounded since the beginning of the 1990s. The volume of the Valli is roughly 108 m3, and the water exchange rate with the sea has been estimated in 10 m3 sec 1; theoretical water residential time can roughly be estimated in 115 d, quite a long time when compared, for example, with the neighboring Sacca di Goro, whose water residential time resulted in 5 d (Brath et al., 2000). This makes the Valli di Comacchio an environment characterized by limited water renewal. The bottoms of the Valli are typically muddy. Four permanent sampling sites (A, B, C and D; Figure 1) were chosen as being characteristic and representative of the shallow-water habitats in the Valli di Comacchio. Observations at all the sites indicated that sediment composition was consistent over the duration of the sampling program; sites A, B and C were characterized by clay-silt bottoms, while site D had clay sediments; moreover, sites A, B and D had nonvegetated sediments, while at site C a sparsely vegetated meadow of the seagrass Ruppia cirrhosa (Petagna) Grande occurred. All sites were sampled at roughly 3-month intervals between the end of December 1996 (winter 1997) and the end of January 2000 (winter 2000). No sample was collected from any site in winter 1999. Three replicate bottom grab samples were collected at each site at each sampling date using a 0.06 m2 Van Veen grab (depth of sampling: about 8 and 10 cm in clay and clay-silt sediment, respectively). The contents of the grab were gently washed on a 0.5 mm sieve. Material retained on the sieve was fixed in 8% buffered formalin, and stained with rose bengal to facilitate sorting and identification. The abundance of individuals, identified at the species level when possible, was measured for each sample. Water parameters (temperature, salinity, and dissolved oxygen) were measured at each site at approximately monthly intervals by means of a Idronaut Ocean Seven 316 multiparameter probe at about 20 cm above the bottom. Values of water parameters were seasonally averaged. Measures of community constancy over the three years of the monitoring program were calculated for each site. These inc (...truncated)