Introduction

Intertidal rocky shores are heterogeneous environments (Araújo et al. 2005) and encompass a gradient of environmental conditions from fully marine below low tidal levels to fully terrestrial, where splash and spray reach to the highest levels above high tide (Underwood 2000). Due to this heterogeneity, floral and faunal variation can occur over a short horizontal scale (Masi et al. 2009) and this characteristic of rocky shores has led many studies at various spatial scales (Menge 1991; Underwood and Chapman 1996). Other than spatial variation, a characteristic of rocky shores is vertical zonation with bands of different biodiversity from the upper to the lower intertidal zones (Ellis 2003). These zones are in transition between marine and terrestrial ecosystems; therefore, the transition is not abrupt due to the involvement of the tide and waves. A strong environmental stress gradient, viz. desiccation, temperature and irradiance, occurs perpendicular to the shore, which exhibits increased values towards the upper heights of the littoral zone (Chappuis et al. 2014), and this distribution of organisms along this gradient is called vertical zonation. This pattern of vertical distribution has been studied by many authors (Zaneveld 1937; Ballesteros and Romero 1988; Sibaja-Cordero and Cortés 2008; Chappuis et al. 2014). On rocky shores, substratum heterogeneity can influence environmental variables at various scales, which, in turn, might affect the biological processes (Guichard et al. 2001). If the complexity of substratum is very high, then the assemblages of organisms can change over a very fine spatial scale (Benedetti-Cecchi and Cinelli 1997).

Much attention has been dedicated to the study of organisms on rocky shores, especially their vertical zonation, the upper and lower limits of species, and distribution along gradients of wave exposure. Various studies have been carried out to understand the cause of biotic zonation, including physical factors (Colman 1933) and combinations of physical and biological factors, such as physiological tolerance (Connell 1961a, b) and species interactions (Bruno and Bertness 2001; Menge and Branch 2001), but, arguably, patterns of zonation are very complex to understand because numerous biotic and abiotic factors act at different spatial and temporal scales (Dayton 1971; Menge and Sutherland 1987; Trivedi and Vachhrajani 2015). This variation in the distribution and abundance of organisms in different zones of intertidal areas has provided a basis for ecological experiments and such complex patterns of variation have been studied in depth, especially for the molluscs (Chapman 1994a, b; Chapman and Underwood 2008). To understand the processes of structural assemblages of organisms, quantitative analysis of patterns of distribution and abundance is essential (Moore and Seed 1985; Underwood et al. 2000). In earlier studies, the main attention was dedicated to variation in the assemblages in relation to various environmental gradients like heights of the shoreline (Paine 1974), exposure to waves (Lewis 1964) and biotic factors such as competition (Mangialajo et al. 2012), grazing (Underwood and Jernakoff 1981; Thomas 1994), predation (Underwood and Jernakoff 1981) etc. However, Underwood and Chapman (1996) showed that significant variability in the abundance of organisms on rocky shores is present even at less than 10 m scales. Thus, in recent years, a greater emphasis has been placed on the diversity and abundance of organisms at various spatial and temporal scales (Kelaher et al. 2001). Zonation patterns may vary along the coast due to processes unrelated to vertical gradients (Chappuis et al. 2014). Moreover, different factors act as the main drivers of ecological processes and patterns, and these factors are not scale-independent (Levin 1992; Willig et al. 2003). Therefore, in recent times, many studies have focused on the variability of littoral assemblages at different spatial scales along shores (Cruz-Motta et al. 2010; Valdivia et al. 2011; Veiga et al. 2013). These studies have shown that the distribution of species and assemblages along the horizontal axis (i.e. parallel to the sea surface) is important at different scales.

Pioneering works on zonation by Stephenson and Stephenson (1949, 1972) proposed “universal features of zonation between tide-marks on rocky coasts” after conducting surveys of vertical biotic zonation on littoral shores worldwide and Lewis (1961, 1964) carried out an extensive study on the zonation pattern of the British Isles. Since then, numerous studies have been conducted on rocky shores which showed the habitat selection of organisms (Lubchenco et al. 1984; Olabarria and Chapman 2001), degree of aggregation (Garrity and Levings 1984; Levings and Garrity 1984), assemblages (Menge and Farrell 1989) and some species of molluscs (Bandel and Wedler 1987) have been identified as indicators of specific littoral zones. Simultaneously, experimental approaches to understand the functioning of rocky intertidal zones such as competition (Dayton 1971; Underwood 1978, 1984), predation (Paine 1974; Menge 1976, 1978), disturbance (Sousa 1979), recruitment (Menge 1991), resistance (Bertocci et al. 2010), effect of habitat heterogeneity (Beck 1998, 2000; Olabarria and Chapman 2001) and effects of physical factors (Bird et al. 2013) have been carried out. However, study on the distribution of gastropods off the rocky shores of the Indian coast is very limited, e.g. Misra and Kundu (2005), Gohil and Kundu (2013) and Trivedi and Vachhrajani (2015), whereas no quantitative studies on rocky shores macrofauna in the environment of the Andaman and Nicobar Islands have been attempted. So, the aim of the present study is to document the gastropod zonation patterns and their variation at spatial and temporal scales.

Materials and methods

Study area

The Andaman and Nicobar Islands are a typical tropical ecosystem, comprising 572 islands, with a coastline of about 1962 km. The Andaman and Nicobar Islands are situated on mountain tops which rose from the bottom of the sea from around 2000 m depth. Hence, the continental shelf is steep and very limited. There are numerous creeks and bays bordered with mangroves, coral reefs, rocks and sandy beaches. The influence of oceanic seawater is much more pronounced close to the shore. The climate of these islands is typically tropical, with heavy gales, cyclones, and hot and humid conditions. The annual temperature ranges from 18 °C to 36 °C, with a slight increase from the northern group of Andaman Islands to the southern group of Nicobar Islands. The relative humidity of ambient atmosphere reaches about 80%. The Andaman and Nicobar Islands are covered with clouds during most months and rainfall is heavy, with an average of about 3100 mm/year, which is prolonged from May to December. Major rainfall occurs during late May to early October (southwest monsoon), while a weak spell of northeast monsoon brings rain during November to December. Since most of the islands are covered with dense tropical forests, the coastal waters are greatly influenced by the heavy rainwater draining into the coastal areas. The coastal waters of the Andaman and Nicobar Islands are comparatively pristine, since the seas around the islands are deep, with clear water.

For the present study, six rocky shore locations, viz. Science Centre (St. A), Brookshabad (St. B), Rose Valley (St. C), Burmanallah (St. D), Rangachang (St. E) and Kodiyaghat (St. F), along the east coast of South Andaman were chosen (Fig. 1). The geographical locations of the sampling sites were recorded using a hand-held GPS (Garmin). All locations are directly influenced by the open sea.

Fig. 1
figure 1

Study area showing the sampling locations along the South Andaman coast

Full size image

St. A has exclusively rocky substratum, except for a sandy stretch on the supralittoral zone. This area has sedimentary rocks on the high water line that are covered with algae and support good abundance of gastropods which feed on algae. The substratum of St. B is completely rocky, with no trace of sand or mud, and no freshwater input from any source except rain. This area contains big rocks and that makes this area a very uneven substratum and which, in turn, provides sheltered spaces and crevices for relatively large sized epibenthic organisms. St. C contains relatively big rocks and poorly developed mangroves nearby. St. D comprises a vast stretch of intertidal area which is exposed during low tides. This area receives fresh water input through a narrow creek. This area can be categorised in the mixed substratum category, since the upper shore and part of the mid shore has soft substratum (mud and sand covered with dead, broken coral and stones), whereas the lower shore contains rocky substratum. St. E has relatively big rocks, so the habitat structure is not even. St. F is located just north of Chidiya Tapu. The area has mainly hard rocky substratum with randomly located patches of sands. Poorly developed patches of mangroves are present adjacent to the locations of St. C and St. D.

Methodology

The rocky shore of South Andaman was surveyed during low tides from August 2013 to July 2015 at 3-month intervals. Due to prolonged monsoon season on these islands, the monsoon season was divided into two phases, viz. monsoon phase I and monsoon phase II. During the study period of 2 years, a total of 576 samples were collected covering all four seasons, viz. monsoon phase I (May to Jul), monsoon phase II (Aug to Oct), post-monsoon (Nov to Jan) and pre-monsoon (Feb to Apr). From each season, 72 samples were collected from six locations (72 × 4 = 288 per year). For the present study, the intertidal area was divided into three traditional zones, upper, middle and lower intertidal shores. At each station, two transects were laid perpendicular to the shore and from each transect, duplicate samples were collected from the high, mid and lower water lines using 1-m2 quadrats (English et al. 1997). From each quadrat, the number of individuals were counted and recorded in the field notebook. Scheduled gastropods species were not collected; instead, photographs were taken in the field itself. In case of juvenile forms which were difficult to identify up to the species level, the numbers of such individuals were enumerated in the field notebook up to the possible genus level. Maximum effort was given to identify the species in the field itself; capturing or collection of organisms were avoided even if the species were abundant in order to avoid stress on biodiversity and only those gastropods which could not be identified in the field were brought to the laboratory (only one representative), where they were washed with tap water in order to clean the algal film and other attached debris which could hinder the identification. Then, they were narcotised using MgCl2 and finally preserved in 5% formalin solution for further identification up to the lowest possible taxonomic level using standard taxonomic identification keys (Cernohorsky 1967, 1972; Abbott and Dance 1982; Subba Rao 2003).

Data analysis

Univariate diversity measures (e.g., species richness, Margalef’s diversity and Shannon–Wiener diversity index) were estimated using PRIMER v6.1 (Clarke and Gorley 2006). Box and whisker plots describing univariate diversity measures and abundance of gastropods were constructed using OriginPro v.8.5 (OriginLab Corporation, Northampton, MA, USA). For the multivariate analysis, species abundance data were transformed to square root values to smooth the extremes in data and allow rare species to exert some influence on the results whilst reducing the potentially overwhelming impact of the most common taxa. From the transformed data, the Bray–Curtis resemblance matrix was prepared, which was used to analyse non-metric multi-dimensional scaling (nMDS) (Clarke and Warwick 1994). SIMPER (similarity percentage) analysis was used to determine the taxa most responsible for the percentage dissimilarities between (and similarities within) each shore level and location. The percentage of variability in sample ordination patterns along nMDS axes were examined using principal coordinate analysis (PCO) and those species that exhibited Pearson correlation value > 0.65 were superimposed. For all of the multivariate analyses, the mean values of four replicate samples of each shore level were used. Three-factor PERMANOVA (location, shore level and season) was tested to examine any differences among gastropod populations. Constrained ordination by canonical analysis of principal coordinates (CAP) was used to characterise and visualise these differences. Both the analyses were done using the Bray–Curtis resemblance matrix and 999 permutations. For biota and environment matching, the BIOENV method was used to observe the correlation of environmental parameters on gastropod assemblages.

Results

A total of 34,004 individuals of gastropods were recorded during the study, contributed by 71 taxa, belonging to 34 genera under 21 families and four orders. During the study, the maximum number of species and abundance were recorded by Mesogastropoda (26 species), constituting 52% of the total gastropods, while Neogastropoda was represented by 23 species (11.6%) and Archaeogastropoda by 21 species (32.2%). Five families, viz. Littorinidae, Neritidae, Patellidae, Planaxidae and Muricidae, showed high abundance and together contributed 89% of the total abundance. Littorinidae was the most dominant family throughout the study and constituted around 42% of the total population, followed by Neritidae (23%) and Planaxidae (significantly, the whole 11% proportion was contributed by a single species). These families showed seasonal variation in their abundance. The PERMANOVA test showed significant differences among all the factors, including shore level, locations and seasons (Table 1). During the study, environmental variables did not show any significant relationship with the gastropod population. The BIOENV results showed that no single or cumulative effects of environmental variables were significant (0.119 Spearman rank correlation, which is too low), the maximum being for air temperature and salinity.

Table 1 PERMANOVA main results based on Bray–Curtis similarities for differences in gastropod assemblages among locations, shore levels and seasons in terms of abundances (Lo = location, Sh = shore level, Se = seasons)
Full size table

Spatial variations

Gastropods showed a significant variation at the species level in the spatial scale, with the highest number of species found at St. D and St. F (54 species each), followed by St. E (50 species) and the lowest number of species were found at St. C (46 species). Gastropods abundance also varied remarkably at different locations (Fig. 2a). The highest abundance of gastropods was found at St. F (88 ind/m2), followed by St. B (66 ind/m2) and the lowest abundance was observed at St. A (40 ind/m2). The highest abundance at St. F was mainly due to the dominance of Littorinidae. This location forms a suitable habitat for littorinids. The supralittoral zone of this location is shadowed by dense vegetation. As Littorinidae occupies the upper zone, this vegetation cover assists to overcome desiccation and exposure to light. Littorinidae constituted 14% of the total gastropods at St. F and the lowest abundance of this family was found at St. D (4.6%), whereas Neritidae was dominant at St. B and the lowest abundance of the family was noted at St. A. Planaxidae showed maximum abundance at St. B and the lowest was found at St. E. In terms of diversity, the most diverse location was St. D (d, 3.290 ± 1.32; H′, 1.690 ± 0.37), followed by St. A (d, 3.203 ± 1; H′, 1.858 ± 0.26) and the lowest diversity was recorded from St. B (d, 2.576 ± 0.88; H′, 1.650 ± 0.35) (Fig. 2b, c). The PERMANOVA (p = 0.001) test also proved significant differences among locations.

Fig. 2
figure 2

Gastropods of the Andaman Islands during August 2013 to July 2015, univariate measures: a abundance, b Margalef’s index, c Shannon’s index H′

Full size image

Temporal variations

Unlike spatial variation, temporal variation was not very prominent. However, a slight variation in abundance and species composition at the seasonal scale was observed. Among the four seasons, the highest number of species was recorded during pre-monsoon (64 species), followed by monsoon phase I (63 species) and the lowest abundance was found during post-monsoon (60 species). The maximum abundance of gastropods was observed during post-monsoon (64 ind/m2), followed by pre-monsoon (62 ind/m2) and the minimum abundance was found during monsoon phase II (52 ind/m2). Some of the dominant families showed variation in their abundance at the seasonal scale. Littorinidae showed maximum abundance during post-monsoon (12%) and the lowest abundance was found during monsoon phase II (9%). Neritidae, which was the second most dominant family, showed maximum abundance during post-monsoon (6%) and the lowest was during pre-monsoon (5%). Planaxidae showed maximum abundance during post-monsoon (3.5%), followed by pre-monsoon (2.8%) and the lowest abundance was observed during monsoon phase II (1.3%). In terms of diversity, the most diverse season was pre-monsoon (d, 3.311 ± 1.22; H′, 1.865 ± 0.32), followed by monsoon phase II (d, 3.021 ± 1.14; H′, 1.703 ± 0.32) and the lowest diversity of gastropods was observed during post-monsoon (d, 2.888 ± 1.05; H′, 1.697 ± 0.40). PERMANOVA (p = 0.042) showed significant variations among seasons but ordination plots did not reveal clear separation.

Vertical zonation

Out of 34,004 gastropod individuals recorded during the study, the upper shore contributed 23,284 (68.47%) individuals, whereas the mid and lower shores contributed 6835 (20.10%) and 3885 (11.42%), respectively. Evidently, the upper shore harboured maximum mean abundance (121.27 ind/m2), followed by the mid shore (35.59 ind/m2) and the lower shore constituted only 22.23 ind/m2 (Fig. 3a). However, diversity (Margalef’s d) recorded a maximum at the mid shore and lowest at the lower shore (Fig. 3b), whereas the maximum Shannon–Wiener (H′) was recorded at the mid shore (Fig. 3c). The results of PERMANOVA showed significant variation of the shore levels (p = 0.001), as well as the interaction of shore levels and locations (Lo × Sh; p = 0.001).

Fig. 3
figure 3

Shore level variations of gastropods: a abundance, b Margalef’s index, c Shannon index H′

Full size image

At the upper shore, the most abundant family was Littorinidae (60%), followed by Neritidae (18%) and Planaxidae, which was represented by only one species, Planaxis sulcatus, and contributed 15%. Overall, the upper shore harboured 29 species out of 71 taxa, and the most diverse families at this zone were Neritidae (6 spp.), followed by Muricidae (5 spp., though abundance was much lower, i.e. almost 1%) and Littorinidae (4 spp.). At the horizontal axis of the upper shore, St. F showed the maximum abundance of gastropods (184.28 ind/m2) and St. C showed the minimum (86.31 ind/m2) (Fig. 4a), whereas in terms of species richness, St. A and St. D showed the highest number of species (23 species each) and the lowest number was shown by St. B (14 species). In terms of diversity, the most diverse (Fig. 4b, c) upper shore was recorded at St. A (d = 2.399 ± 0.65; H′ = 1.777 ± 0.35) and the least diverse at St. B (d = 1.716 ± 0.25; H′ = 1.672 ± 0.18).

Fig. 4
figure 4

a Mean abundance of gastropods at each shore level across the locations. Abundance values are natural log transformed. b Variation of Margalef’s d at each shore level across the locations. c Variation of Shannon—Wiener (H′) diversity values at each shore level across the locations

Full size image

At the mid shore, the most abundant family was Neritidae (42%), followed by Muricidae (19%) and Patellidae (12.5%). Engina mendicaria (4.7%) was the most common gastropod species observed at the mid shore. Out of 71 taxa recorded during the study, the mid shore harboured 61 species and the most diverse families at this zone were Muricidae and Cypraeidae (9 spp. each), whereas Neritidae contributed six species. Variation in the abundance of gastropods at the horizontal axis of the mid shore revealed that the maximum abundance of gastropods was harboured by St. F (63.78 ind/m2) and the minimum by St. D (17 ind/m2) (Fig. 4a). The highest species richness was also recorded from St. F (41 species) and the lowest was recorded from St. B and St. C (27 species each). At the mid shore, the highest diversity of gastropods was recorded from St. E (d = 3.893 ± 1.13; H′ = 2.164 ± 0.24) and the lowest from St. B (d = 2.682 ± 0.59; H′ = 1.801 ± 0.24) (Fig. 4b, c).

At the lower shore, the maximum abundance was contributed by Muricidae (26.4%), followed by Patellidae (25.4%) and Neritidae (20%). Latirolagena smaragdulus (Linnaeus, 1758), was the most commonly observed species at the lower shore and contributed 10% of the total abundance. Overall, this zone harboured 53 species and the most diverse families at this zone were Cypraeidae (12 spp.), but the abundance was relatively low (4%), followed by Muricidae (9 spp.)., while Patellidae, Neritidae, Turbinidae and Trochidae contributed three species each at this zone. Variation in the abundance of gastropods at the horizontal axis of the lower shore revealed that the maximum abundance of gastropods was harboured by St. E (33.9 ind/m2) and the lowest from St. D (8.25 ind/m2) (Fig. 4a). However, species richness was highest at St. F (37 species) and lowest at St. D (28 species). Lower shore diversity was maximum at St. D (d = 4.525 ± 0.89; H′ = 1.778 ± 0.27) and lowest at St. C (d = 3.124 ± 0.52; H′ = 1.717 ± 0.22) (Fig. 4b, c).

Gastropod assemblages and multivariate analyses

nMDS yielded three distinct groups based on Bray–Curtis similarity. Among the 72 samples used in this analysis, group I comprised samples collected from the upper shore (n = 24), group II consisted of samples from the mid shore and group III consisted of samples from the lower shore (Fig. 5). The nMDS results indicated that the upper shore formed a separate cluster but there was some overlapping between mid- and lower-shore gastropods, indicating significantly different gastropods on the upper shore, whereas there was common gastropods occurrence at the mid and lower shores. SIMPER analysis supported the outcome that showed 31% of gastropods overlapping (similarity) between the mid and lower shores. The most common species responsible for zone overlapping were Nerita costata Gmelin, 1791, Cellana radiata (Born, 1778), Patelloida saccharina (Linnaeus, 1758), Neothais marginatra (Blainville, 1832), Thalessa virgata (Dillwyn, 1817) and Acmaea achates (Reeve 1855). SIMPER analysis was carried out to evaluate the species that defined similarity within and dissimilarity between shore levels. The analysis revealed that the average similarity of samples within each shore level was 66.5%, 46.4% and 41.4% for the upper, mid and lower shores, respectively. The cumulative contribution of the top seven listed taxa to within-shore level similarity decreased with shore height. 84.09% of the upper shore sample similarity was because of its seven most important taxa, the highest seven mid-shore taxa accounted for 77.37% of the similarity, while the same is true of only 72.02% of the seven key taxa for lower-shore similarity. The average dissimilarity between each pairwise combination of shore height showed maximum dissimilarity between the upper and lower shores, i.e. 92.8%, followed by the upper and mid shores (83.4%), whereas the mid and lower shores showed the least dissimilarity (71.1%). It is evident from the analysis that the maximum dissimilarity of the upper shore to the mid and lower shores was caused by the higher abundance of Littorinidae at the upper shore. The species which are responsible for causing dissimilarity is given in Fig. 6. The SIMPER analysis (Table 2) was also used to investigate the most important discriminating taxa between the shore heights, using transformed species abundance data. The PCO ordination based on species association patterns revealed the importance of the first two principal coordinates, accounting for 59.4% of the total variation in species distribution patterns (Fig. 7). The first PCO axis showed a clear separation of upper shore samples from the other two shore heights, whereas the second axis separated the mid and lower shores. PCO confirmed the result of cluster analysis that the variation between the mid and lower shores was lower (15.4%) compared to the upper shore (44%). PERMANOVA showed significant variation in the gastropod populations at different locations, shore levels and seasons. CAP revealed these differences which were not clearly visible by other ordination plots. CAP however, could not reveal a clear picture of gastropod population at the seasonal level but, location wise, it separated St. A and St. D from other locations (Fig. 8).

Fig. 5
figure 5

Non-metric multi-dimensional scaling (nMDS) plot based on gastropod abundance data showing three separate groupings of samples, each corresponding to different shore levels

Full size image
Fig. 6
figure 6

Results of SIMPER analysis (modified from Davidson et al. 2004) showing the most important taxa responsible for similarity within and dissimilarity between the shore levels

Full size image
Table 2 Results of SIMPER analysis: characterising species (mean ind/m2) for each shore level
Full size table
Fig. 7
figure 7

Two-dimensional principal coordinate analysis (PCO) showing superimposed vectors for important gastropod species (which contributed to r > 0.65) that shaped gastropod assemblage patterns for different shore levels

Full size image
Fig. 8
figure 8

Two-dimensional canonical analysis of principal coordinates (CAP) showing the assemblage pattern of gastropods for different locations. CAP ordination plot separated St. A and St. D from other locations

Full size image

Discussion

Most of the works on zonation have been done considering all flora and fauna of the intertidal zone (e.g. Sibaja-Cordero and Cortés 2008; Chappuis et al. 2014) but in the present study, gastropod zonation was studied to investigate the gastropod community changes with respect to different intertidal zones. A similar study on the vertical zonation of a single family of gastropoda (Littorinidae) was carried out by Lee and Lim (2009) along the coast of Singapore and showed that littorinids preferred specific zones. Even after the individuals were manually translocated to different zones, they returned to their preferred zones in the field. Underwood (1975) also studied the intertidal zonation of four coexisting species of gastropods.

The observed vertical differences are possibly a result of marked variation in environmental stress which occur between low and high elevations (Menge and Branch 2001) and organisms are scattered according to their preferences of specific zones best suitable for their survival, as the distribution pattern of organisms in intertidal habitats is attributable to physiological needs in relation to physical stress and biological interactions (Schonbeck and Norton 1978, 1980; Kastendiek 1982; Benedetti-Cecchi et al. 2000a, b; Araújo et al. 2005). During the analysis of individual distributions of organisms along transects, it was clear that the majority of the species appeared only at some levels on the shore, e.g. Latirolagena smaragdulus at the lower shore, Engina mendicaria at the mid shore and Littorinidae at the upper shore. Underwood (1981) has documented the occurrence of any given species at a particular height on the shore. However, within each height level on the shore, different assemblages of gastropods existed and the dominance in each of these assemblages can be attributed to a particular species or family with high cover percentage. Therefore, it was considered to classify those assemblages (or zones, as, in this study, it was found that there were different assemblages of gastropods across the shore heights) based on dominant taxa. The present study revealed that the rocky intertidal zones of the Andaman Islands can be divided into three zones based on gastropods abundance at different heights of intertidal zones, such as littorinid zone, neritid zone and muricid zone. The upper shore was excessively dominated by periwinkles (littorinids) and the mid shore, despite being the most diverse zone, was dominated by Neritidae, whereas the lower shore harboured many gastropod families, but the total abundance of Muricidae outnumbered other gastropod families.

The most diverse gastropod zone was the mid shore. This may be due to the optimum conditions for survival of the majority of gastropods, as this zone is subjected to less duration of exposure to air compared to the upper shore and reduced wave action compared to the lower shore. Apart from environmental factors, increased availability of food and shelter also makes this zone suitable for gastropods. Families such as Neritidae graze upon the algae and Muricidae, which are carnivore gastropods, feed upon other gastropods, barnacles (Subba Rao 2003) and rock oyster Saccostrea cucullata (Taylor 1990), and these are available aplenty at the mid shore. The upper shore was mainly dominated by Littorinidae, because of their high tolerance to desiccation and adaptation for survival under prolonged air exposure. Lam (1980) found that Echinolittorina malaccana (family Littorinidae) was most tolerant to desiccation at 38 ± 1 °C, an adaptation which shows its successful survival at the upper shore. Another factor that may play an important role for the survival of gastropods at the upper shore is colour, which affects the absorption and reflection of heat. Vermeij (1973) stated that a light-coloured surface absorbs less radiation from visible light. Therefore, the majority of littorinids during this study were light coloured. In addition, Neritidae, which also contributed notable abundance at the upper shore, were light coloured (e.g. Nerita plicata Linnaeus, 1758, Nerita polita Linnaeus, 1758 etc). Mostly all of the Littorinidae and Neritidae collected from the upper shore were pale yellow or light brown in colour. Contrary to the upper shore, mid and lower shore Neritidae were dark in colour. The mid shore was dominated mainly by Neritidae, Muricidae and Patellidae. The majority of Neritidae observed from this zone were found to be dark in colour, such as Nerita albicilla Linnaeus, 1758 and N. chameleon Linnaeus, 1758. Together, these two species contributed 37% of the 42% total Neritidae at the mid shore. On the other hand, Muricidae and Cypraeidae, even though their abundance was less compared to Neritidae, contributed the maximum number of species at this zone. The lower shore was dominated by Muricidae. This carnivorous family has a worldwide distribution on rocky coasts and reef ecosystems. These active predators feed upon other gastropods, bivalves and barnacles by means of softening secretion and scraping action of radula. Muricidae was found on the mid shore as well as the lower shore, but higher abundance at the lower shore showed the preference to the lower shore. However, the considerable amount of abundance at the mid shore makes it unambiguous that they move upshore for food and optimum surviving conditions (such as to avoid harsh wave action).

Some gastropods move up and down the shore with the tide. Species such as Nerita yoldii Recluz, 1841 can modulate its behaviour in response to local changes in environmental conditions. Such plasticity by these gastropods is important in ensuring that individuals maximise opportunities to gain energy or minimise risk to extreme physical conditions (Yeung and Williams 2012). Therefore, such foraging behaviour can pose a problem while investigating zonation as they can overlap with different species at different zones. These species are zone shuttlers and found in abundance at the mid shore as well as the lower shore. Nerita costata (Neritidae), which was responsible for overlapping between the shores occurring throughout all shore heights during the study, are known to move downshore for foraging. Yeung and Williams (2012) also revealed in their study that Nerita species can forage at long distances during summer and neap tide. All Nerita species are zonal shuttlers (Chelazzi et al. 1988) and, therefore, other neritids (in addition to Nerita costata) were found throughout the intertidal zone, but the abundance of other neritids were much lower compared to Nerita costata. Different species have different foraging behaviours, e.g. N. scabricosta Lamarck, 1822, which forages whilst awash, was able to avoid risk of predation by fish whilst immersed by aggregating above the high tide mark (Yeung and Williams 2012). These differences in foraging behaviour in species co-occurring on the same shore confirm the importance of shore level in determining species’ potential foraging strategies (Williams and Little 2007) and their preference of different zones. Therefore, other neritids were not as widely distributed as Nerita costata. On the other hand, muricid gastropods, which feed upon neritids (Taylor and Morton 1996), also forage during high tide when immersed (Chow 2004), which can be the reason for the overlapping of muricids in different zones. Moreover, the lower shore is a very harsh condition in which to survive due to high wave action; thus, the degree of wave exposure may be the most important factor affecting communities, disabling the identification of clear distribution patterns in relation to shore height. The grade of exposure is an important factor regulating the faunal composition and has been mentioned by many authors (Underwood 1981; Hawkins 1983; Menge et al. 1993). The exposure grade can also alter the spatial distribution patterns of organisms along the shore. Underwood (1981) documented that there is a tendency of wider organisms distribution with increasing wave exposure. This tendency was also observed in the present study, as the gastropods occurring at lower levels of shore moved upshore and overlap their distribution to avoid wave action.

During the study, the species composition was more or less same at different spatial scales (i.e. across sites) but abundance and diversity varied remarkably across the shore height. Various abiotic and biotic factors act in synchrony for these variations. Recent findings have shown that small-scale substratum heterogeneity is important in determining the assemblage structure within any particular shore height (Beck 1998; Araújo et al. 2005). Archambault and Bourget (1996) also found that variation in abundance is attributed to the small scale. The colonisation history, small-scale substratum heterogeneity, biological interactions and organisms’ physiological needs can, altogether, contribute to the heterogeneity in assemblage structure found within each site. Lapointe and Bourget (1999) found that the substratum heterogeneity was the most important factor determining community structure. Bourget et al. (1994) demonstrated that substratum heterogeneity and complexity strongly structured the small-scale distribution of species during the early development phases of a marine epibenthic community (Araújo et al. 2005). Thompson et al. (1996) stated that structural complexity may reduce environmental stress experienced by intertidal communities during low tide and which, in turn, allow the development of a higher diversity of assemblages. Pandey et al. (2016) found that the heterogeneous substratum support a higher diversity of gastropods. Furthermore, biological interactions are also responsible for the development of mosaics of communities at small spatial scales (Kaehler and Williams 1997; Johnson et al. 1998; Leonard 1999; Benedetti-Cecchi et al. 2000b; Williams et al. 2000; Araújo et al. 2005).

In this study, more pronounced and a major source of variability was the gastropods species composition across shore height (i.e. vertical axis) than across sites for the same shore height. The relative importance of physical factors such as tidal height and of physical attributes of the habitat, such as small-scale spatial variability, could depend on the magnitude of each of these factors (Benedetti-Cecchi 2001). However, in the present study, physical factors such as salinity, temperature and pH did not show any pronounced effect or significant correlation with the results; thus, the physical attributes of habitat may be important factors determining the gastropods community and assemblages structure as observed in the Mediterranean Sea (Benedetti-Cecchi 2001).

Conclusion

In the Andaman Islands, quantitative studies on the gastropods assemblages and population ecology are scarce. Hence, in the present study, an attempt is made to enrich our knowledge on the gastropods community of these islands. The present study is based on gastropod samples collected from six locations along the eastern coastline of South Andaman Island. This study highlights some important aspects of gastropods zonation, diversity, abundance and assemblages. Based on multivariate analysis (using Bray–Curtis similarity), the rocky shore of Andaman can be divided into three distinct zones, viz. the littorinid, neritid and muricid zones. Gastropods showed a clear demarcation of zones, especially for the upper shore. The most diverse gastropod zone is the mid shore, whereas the upper shore is gastropod abundant but the least diverse. The gastropod population is scale dependant, as it varies with location for each shore; however, there is no significant variation at the temporal scale, i.e. the gastropod population varies across location but not significantly with seasons. The environmental variables like air temperature, water temperature, salinity, dissolved oxygen and pH did not show any significant effect on the gastropod distributional patterns.