Iheringia Seasonal variation of the crustacean fauna in the belowground and aboveground strata in a Halodule wrightii meadow of northeastern Brazil

The diff erent microhabitats provided by seagrasses and the habit of the species determinate the vertical distribution of crustaceans. This study verifi ed the infl uence of the seasonality on the spatial distribution of the crustacean community associated with a Halodule wrightii meadow. From April 2006 to July 2007, we performed fi fteen samplings in Goiabeiras Beach, Ceará State, each with fi ve sectioned replicates in belowground and aboveground. Cumaceans and the amphipod Hyale media (Dana, 1853) had a higher frequency, density, and dominance than the other taxa, in both strata. The community exhibited gradual changes along the study, in both the belowground and aboveground strata, but the seagrass structure was not suffi cient to explain the vertical distribution of the crustacean fauna along the time.

Seagrasses are marine angiosperms that provide a tri-dimensional habitat for benthic organisms, with a canopy along the water column and a root system that penetrates the sediment (Orth et al., 1984;Hemminga & Duarte, 2000;Nakaoka, 2005;Björk et al., 2008).Thus, these habitats are used in several ways, with algae and animals attached to the leaf surface (epifl ora and epifauna), dispersed along the bottom (epibenthic organisms), or in the sediment (infauna) (Nakaoka, 2005).The settlement of these organisms attracts visiting groups, especially fi sh shoals and large herbivores, and promotes increased productivity in the ecosystem (Zieman & Zieman, 1989;Nakaoka, 2005;Björk et al., 2008).
Many species of crustaceans settle on the meadows after the larval stage and remain there until reaching the adult phase (Hemminga & Duarte, 2000;Nakaoka, 2005;Björk et al., 2008).They include a diversity of forms of organisms that can live in different habitats provided by these plants, as the sediment rich in organic matter and the surface of the stable sediment provided by the rhizome/root system, and also the leaves canopy, which can offer food and more stable environmental conditions (Abele, 1974;Zieman & Zieman, 1989;Gambi et al., 1995;Hemminga & Duarte, 2000;Williams & Heck, 2001).
Thus, the literature suggest that the different microhabitats provided by the seagrass added to the species habits are sufficient to determinate the vertical distribution of the crustacean species.The aim of this study was to observe the influence of the seasonality on the crustacean community associated with the belowground and aboveground strata of a Halodule wrightii Ascherson meadow of the semiarid coast of northeastern Brazil.
Beach rocks cover an area of approximately 500 m parallel to the coast, where macroalgae banks cohabit with a small meadow of the seagrass Halodule wrightii, with a maximum width of around 30 m.The meadow is regularly immersed during low spring tides (Barros & Rocha-Barreira, 2014).The local climate is rainy tropical with a long dry season -Aw' (Köppen, 1948).The rainy season occurs between February and May, with weak winds intensifying in the dry season, between August and November (Morais et al., 2006).
Fieldwork and material processing.Fifteen monthly samplings were performed from April 2006 to July 2007, each with five random replicates, collected with the help of a core (0.0078 m 2 ) buried at 10 cm.The number of samples was determined after the elaboration of a performance curve for the macrofauna groups from a previous sampling.The subsequent samplings took into account the size of the small seagrass meadow studied here, as recommended by Duarte & Kirkman (2001), using a sampling design based on Burdick & Kendrick (2001) for patched meadows.Each core was sectioned in belowground and aboveground, the former of which included a thin upper layer (smaller than 0.5 cm) of the sediment.
The samples were transported to the Laboratório de Zoobentos do Instituto de Ciências do Mar, Universidade Federal do Ceará (Brazil) and washed in running water with a 0.5 mm sieve to retain the macrofauna.Specimens were preserved in a 70% alcohol solution, and the crustacean fauna was subsequently identified in the Laboratório de Carcinologia da Universidade Federal de Pernambuco to the lowest possible taxonomic level, excepting cumaceans and mysids, due to the lack of taxonomists.
Statistical analysis.We determined frequency of occurrence (F<10% -rare; 10%< F< 40% -few common; 40%< F< 70% -common; F>70% -more common), density (specimens per grams of dry weight of H. wrightii), and species dominance for the belowground and aboveground strata, in the dry and rainy seasons.We also obtained the Shannon diversity and Pielou evenness indices, which were submitted to a two-way ANOVA in order to observe the isolated and combined effects of the vertical distribution in the meadow (below/aboveground stratum), and seasonality (dry/rainy period).
Bray-Curtis analyses were performed to determine similarities among the samples of crustacean fauna in each stratum of the seagrass meadow, using the abundance of the species, previously transformed in log (x+1).After obtaining the graph of similarity (Cluster), the groups of samples with more than 40% of similarity were compared using the similarity percentage (SIMPER) analysis to determine the taxa that favoured similarity among the samples.An analysis of similarity (ANOSIM) was performed to determine the significance of seasonal variations in the groups of samples in each stratum.
These analyses (diversity, evenness, Bray-Curtis, SIMPER and ANOSIM) were performed with the aid of the Primer ® (Plymouth Routines in Multivariate Ecological Research) programme, 6.1.6version.The two-way ANOVA was obtained with the STATISTICA ® programme (7.0 version).

RESULTS
A total of 1,477 specimens of 28 taxa were captured, belonging to the orders Amphipoda, Isopoda, Tanaidacea, Decapoda, Cumacea, and Mysidacea.Cumaceans were the most abundant group, accounting for 50% of the crustacean fauna, followed by amphipods (42%).Isopods, brachyurans, and tanaidaceans each accounted for 1% of the community.
All taxa occurred in both strata, except mysids and brachyurans that occurred only in the aboveground stratum.Cumaceans and amphipods were more abundant in the belowground stratum, whereas isopods and tanaidaceans were more abundant in the aboveground stratum (Fig. 2).Amphipods of the family Megaluropidae and Photis sp. were only associated with the belowground.Brachyuran juveniles and adults were associated exclusively with the aboveground stratum, but the megalops were captured in the belowground stratum during the dry season (Tab.I).
The amphipods Hyale media Dana, 1853 and Ampithoe ramondi Audouin, 1828 were the most frequent taxa in the belowground stratum, whereas Microphoxus uroserratus Bustamante, 2002 was the most frequent taxa in the aboveground stratum.
Hyale media was dominant during the dry season in both the belowground (41.7%) and aboveground (37%) strata, whereas cumaceans dominated during the rainy season in these respective strata (70.9% and 86.3%).In the belowground stratum, cumaceans were the densest taxon during both the dry (0.96 ind/g) and rainy (6.27 ind/g) seasons.In the aboveground stratum, cumaceans (16.9 ind/g) and H. media (10.3 ind/g) were the densest taxa in the dry season and cumaceans were the densest in the rainy season (59.9 ind/g).Among the community descriptors, despite of the variation found the seasonality significantly influenced only on the evenness (Fig. 3).
Although some species occurred in both strata, the similarity analysis indicated gradual changes in the crustacean fauna throughout the study occurring almost simultaneously in both strata, but primarily in the aboveground stratum, which also had a more clearly defined seasonal variation.
The similarity (46%) of the community from the belowground stratum was mainly due to the abundance of H. media (27.7%), cumaceans (20.91%), and A. ramondi (11.2%).Above 35% of similarity, these samples indicated two main groups (groups I and II) following seasonal sequences (Fig. 4).The dissimilarity between these groups and the other samples was of 76.9%, also determined by cumaceans (21.6%) and H. media (21.3%).
Group I consisted of samples of the crustacean fauna captured between November 2006 and March 2007, with a similarity of 51%, mainly due to the abundance of H. media (51%).Group II comprised samples captured from April 2007 to July 2007, with a similarity of 60%, mainly due to the abundance of cumaceans (36%).The dissimilarity between groups I and II was of 50.4%, determined by the abundances of H. media (24.2%), cumaceans (14.8%), and Paracerceis sp.(10.5%).
In the aboveground stratum, we also observed two main groups of samples (groups I and II) (Fig. 5), with a total similarity of 36.5%, mainly due to the abundances of cumaceans (52.7%).The similarity was of 48.4% in Group I, which comprised samples related to the rainy season (April to July 2006 and 2007) and October 2006.The abundance of cumaceans (75.3%) was the main variable explaining the similarity of these samples.Group II comprised the samples from the dry season and the beginning of the rainy season (August 2006to February 2007), with a total similarity of 40.4%, mainly due to the abundance of H. media (46%).The dissimilarity between these two groups of samples (70.2%) was related to the abundance of cumaceans (29%) and H. media (20.7%).
The ANOSIM indicated a similar composition of the samples from the below and aboveground strata of the studied H. wrightii meadow, considering both the seasons and the two main sequences of samples (groups I and II of the two strata), with a similarity greater than 40% in both strata (Tab.II).

DISCUSSION
The majority of the taxa identified in the studied Halodule wrightii meadow occurred both in belowground and aboveground strata, which did not allow us to clearly discriminate the groups of species that belonged to the infauna or epifauna, as also observed for the molluscs associated (Barros & Rocha-Barreira, 2013).Moreover, expected significant differences were also not found among different habitats or species of seagrass (Yokel, 1975;Lewis, 1984;Heck Jr et al., 1989;Edgar, 1990;Corbisier, 1994;Garcia et al., 1996;Jernakoff & Nielsen, 1998;Nakaoka et al., 2001;Rosa & Bemvenuti, 2007).The lack of significant differences between infauna and epifauna may be also related to the complexity of these ecosystems (Heck Jr et al., 1989;Jernakoff & Nielsen, 1998) and the interference of other external environmental variables (Edgar, 1990;Jernakoff & Nielsen, 1998;Nakaoka et al., 2001), making the role of seagrasses secondary on the community structure (Harriage et al., 2006).Furthermore, in general the benthic macrofauna exhibit variations in vertical distribution related to seasonal changes in the water and sediment (Jorcin, 1999).
The high frequency of cumaceans and Microphoxus urosserratus in the aboveground stratum is probably associated with their feeding habits, as also pointed out by other researchers for crustaceans and molluscs (Jernakoff & Nielsen, 1998;Nakaoka et al., 2001).Both taxa leave buried or semi-buried and often feed on the sediment surface (MacDonald et al., 2010).The feeding habit could also explain the increased densities of cumaceans in both strata during the rainy season, when occurs an increase of organic matter in the sediment (Barros & Rocha-Barreira, 2013), and probably in the water column.Unfortunatelly, we could not identify them to specific level.It would define key-species of the group in this ecosystem, which could help in future monitoring activities of this meadow.
After cumaceans, the amphipod H. media was an important taxon in terms of density and dominance, as in other seagrass meadows (Jernakoff & Nielsen, 1998).This amphipod lives basically on the surface of the sediment and phytal ecosystems of the intertidal region.Regarding its feeding habits, H. media is essentially a scraper and grazer of microorganisms of macroalgae and microalgae (Tararam et al., 1985;1986;MacDonald et al., 2010).Amphipods may have risen from surrounding macroalgae banks and deposited in the meadow by waves in the dry season or may have been attracted by the availability of epiphytes (Hypnea musciformis Lamourox), as the biomass of which increases in this period (Barros & Rocha-Barreira, 2013, 2014).Species richness may increase when other types of habitats are present (Bologna & Heck Jr, 1999;Gillanders, 2006), although this pattern depends on the type of seagrass substrate (Jernakoff & Nielsen, 1998).
In addition to the probable epiphytic contribution, wave action also favoured the input of specimens surrounding the ecosystem, leading to an increase in the average diversity index during the dry season.This was the case of brachyuran juveniles and adults as well as some species of tanaidaceans, isopods, and amphipods, which only occurred in the aboveground stratum during the dry season, as previously reported for some mollusc species (Barros & Rocha-Barreira, 2013).The patchiness of the meadow (Barros & Rocha-Barreira, 2014) probably contributed to the occurrence of these organisms due to the edge effect, where there is a greater concentration of nutrients and organic matter (Eggleston et al., 1999;Bologna & Heck Jr, 2002;Macreadie et al., 2010).This factor may have contributed to the significant increase in evenness indices in the dry season.
The influence of seasonality on crustacean abundance demonstrated in the similarity analysis, with some interannual differences between corresponding samples from the different studied years, especially considering the rainy seasons, was also observed by Harriage et al. (2006) for the macrofauna associated with Posidonia oceanica (Lipkin) in Prelo Bay, Italy.However, these authors considered that physical factors play a secondary role in the structure of the benthic macrofauna, contrary to what was observed in the present study and in a study of Nakaoka et al. (2001), who investigated benthic macrofauna associated with Zostera marina Linnaeus and Zostera caulescens Miki in Otsuchi Bay, Japan.However, this factor is directly associated with the characteristics of the studied site.
The similarity dendrogram revealed slow changes in the community along the study in response to environmental changes.The community from belowground was more stable, as suggested by other authors (Gambi et al., 1995;Jernakoff & Nielsen, 1998), probably because the leaves changed their arrangement primarily in relation to the belowground stratum (Barros & Rocha-Barreira, 2014).In addition to the seagrass structure, the temporal dynamics of the leaves may exert a significant influence on the epifauna (Jernakoff & Nielsen, 1998).Furthermore, the aboveground fauna seems to be more strongly affected by environmental actions (Barros & Rocha-Barreira, 2013).Edgar (1990) also found a significant seasonal difference in the captured fauna associated with seagrass leaves.The lowest seasonal fluctuations are found for the rhizomes and roots (Barros & Rocha-Barreira, 2014) and for the fauna associated with the belowground stratum (Gambi et al., 1995), since this stratum is a less stressful and more stable environment than the aboveground stratum.
Thus, the crustacean community exhibited gradual changes along the study, in both the belowground and aboveground strata, but the seagrass structure was not sufficient to explain the vertical distribution of the crustacean fauna along the time.The taxa were also influenced by external processes associated with the effects promoted by the seasonality on the ecosystem as a whole (i.e. the patched meadow and the edge effect, as well as the action of waves and the input of organic matter in the meadow), which have an important role in the productivity of this meadow.

Fig. 3 .
Fig. 3. Mean Shannon diversity and Pielou evenness indices of crustacean fauna associated with belowground and aboveground strata of Halodule wrightii meadow in dry and rainy seasons on Goiabeiras Beach, city of Fortaleza, state of Ceará, northeastern Brazil (a, indices for community associated with belowground stratum; b, indices for community associated with aboveground stratum).

Fig. 4 .
Fig. 4. Similarity among samples from belowground stratum of Halodule wrightii meadow on Goiabeiras Beach, city of Fortaleza, state of Ceará, northeastern Brazil; Legend: I) samples collected in predominantly dry months; II) samples collected in predominantly rainy months; III) samples collected in both climatic periods.

Fig. 5 .
Fig. 5. Similarity among samples from aboveground stratum of Halodule wrightii meadow on Goiabeiras Beach, city of Fortaleza, state of Ceará, northeastern Brazil; Legend: I) samples collected in predominantly dry months; II) samples collected in predominantly rainy months.
Tab. I. Frequency of occurrence and spatiotemporal distribution of crustacean fauna in Halodule wrightii meadow on Goiabeiras Beach, city of Fortaleza, state of Ceará, northeastern Brazil.
Few Common X X Fig. 2. Abundance of crustacean fauna captured in belowground and aboveground strata of Halodule wrightii meadow on Goiabeiras Beach, city of Fortaleza, state of Ceará, northeastern Brazil.