Fish guild structure along a longitudinally–determined ecological zonation of Teesta, an eastern Himalayan river in West Bengal, India

Fish guild structure along a longitudinally–determined ecological zonation of Teesta, an eastern Himalayan river in West Bengal, India.— The Eastern Himalaya Biodiversity Hotspot contains exceptional freshwater biodiversity and ecosystems that are of vital importance to local and regional livelihoods, but these are under threat from the developmental and anthropogenic pressures arising from the 62 million people living in the area. Therefore, monitoring the riverine health and considering future conservation approach, the study of fish biodiversity plays a significant role in this region. The River Teesta in the Brahmaputra basin in India forms one of the major rivers in the Eastern Himalayas. In the present investigation, we studied ecological fish guilds as they can enhance the usefulness of fish zonation concepts and serve as tools to assess and manage the ecological integrity of large rivers. We classified fish species according to their water flow preference and spawning substrate preference. Ten spawning habitats were identified, occurring in three water flow guilds. The most widely preferred habitat in upstream zones was lithophils while in lower stretches it was lithopleagophils. On applying predictions of the River Continuum Concept, our results indicated the presence of a zonation pattern based on fish species assemblage and their ecological attributes along the longitudinal stretch of the Teesta River in west Bengal. Along the longitudinal stretch of the river, species richness increased downstream, with maximum richness in the mid–reaches. However, species richness decreased further downstream. The number of ecological guilds also increased downstream, and there were clear shifts in the structure of the guilds.

en la zona. Por consiguiente, el estudio de la biodiversidad piscícola desempeña un papel fundamental en esta región como forma de supervisión de la salud fluvial y de evaluación de políticas de conservación futuras. El río Teesta, en la cuenca india del Brahmaputra, es uno de los mayores ríos del Himalaya Oriental. En el estudio actual hemos analizado los gremios ecológicos de peces ya que ello puede contribuir a mejorar la utilidad de los conceptos de zonación de los mismos y servir como herramienta para evaluar y gestionar la integridad ecológica de los grandes ríos. Hemos clasificado las especies de peces en función de su preferencia por un segmento determinado del curso fluvial y por un substrato de desove. Se han identificado diez hábitats de desove, que se dan en tres gremios de cursos fluviales. El hábitat preferido en mayor medida en el curso superior es el litófilo, mientras que en el curso inferior es el litopleagófilo. Al aplicar predicciones basadas en el concepto de Continuum Fluvial, nuestros resultados indican la presencia de patrones de zonación basados en el ensamblaje de especies de peces y sus atributos ecológicos a lo largo de un tramo longitudinal del río Teesta en Bengala Occidental. A lo largo del tramo longitudinal del río, la riqueza de especies aumenta aguas abajo, con valores máximos en el curso medio, pero la riqueza de especies se reduce en el curso inferior. El número de gremios ecológicos también aumenta aguas abajo, produciéndose claras alteraciones en la estructura de las comunidades.

Introduction
Analysis of the quality of aquatic environments should ideally incorporate attributes that are able to integrate the behavior of elements and biological processes at various levels of organization expressing multiple scale interferences with aquatic communities. The most recent approaches to assess the integrity of environments are multimetric, aiming to combine attributes that represent the broad existing ecological diversity at different levels of biological organization (Casatti et al., 2009). There is a lack of baseline information on freshwater fish species distributions and their ecological requirements throughout the Eastern Himalayas. It has been found that 31.3% of the 1,073 freshwater species of fishes, molluscs, dragonflies and damselflies currently known in the Eastern Himalaya region, are assessed as Data Deficient, emphasizing the urgent need for new research in the region (Allen et al., 2010). Based on these findings, the study of freshwater fish species holds immense importance. Moreover, analysis of their various ecological aspects can adequately assess the ecological integrity of the aquatic ecosystem. Ecological integrity for streams implies the presence of an adaptive assemblage of organisms having a species composition, species richness, and functional organization comparable to that of natural habitat in the region (Karr, 1995).
The Eastern Himalayan Biodiversity Hotspot region and its foothills are rich in both floral and faunal diversity. Fish diversity, in particular, is very rich because the region is home to many large torrential rivers. Fish populations inhabiting these areas are numerous in variety and taxonomically interesting (Abell et al., 2008). As such, the northern districts of West Bengal, specially the districts of Darjeeling and Jalpaiguri, lying within the Eastern Himalayan biodiversity hotspot range, hold a great faunistic importance. The chief rivers are Mahananda and Teesta, with many tributaries such as Murti, Atrai, Jaldhaka, Karala, and Karotoyar. The Himalayas are the source of all major river systems in India. Like other Himalayan rivers, the Teesta River and its tributaries provide a fair ecological niche for many indigenous, and a few exotic, fish species.
Scientific documentation of the Ichthyofaunal diversity of the River Teesta drainage basin is poor and there is no documentation on its stretch within West Bengal. However, as a whole there are several studies on the fish diversity of all along North Bengal. The most comprehensive account of the fish fauna of North Bengal was published by Shaw & Shebbeare (1937) and Hora & Gupta (1941). Apart, Menon (1962) published a distributional list of the fishes of the Himalayas, followed by Jayaram (1977). Subsequent to these there seems to be no report of any fish biodiversity from North Bengal. Allen et al. (2010) reported work on the IUCN status of the freshwater biodiversity in the Eastern Himalayas but there remains an extensive gap in the study of aquatic ecosystem and fish ecology. Analysis of the integrity of riverine environments using a multimetric approach is therefore needed in this region. This approach should include study of the ecological fish guild because knowledge of fish zonation can be used to assess and manage the ecological integrity of large rivers. Grouping fish species into ecological guilds can be a useful method to assess ecological integrity and functioning of large river systems (Aarts & Nienhuis, 2003). Shifts in the structure of functional groups as a result of environmental degradation can be explained by general theories of river ecology, geomorphology and chemistry that can also set guidelines for ecological restoration of degraded river systems, by elucidating the natural configuration of riverine habitats and processes (Vandewalle et al., 2010). The guild and river continuum concept has been largely applied to European rivers (Noble et al., 2007;Fausch et al., 2002), but such information is lacking in Indian rivers. In the present study, the fish guild approach was incorporated to ascertain fish assemblage patterns along the longitudinal gradient of River Teesta in West Bengal, India.

Study area
The River Teesta, originating from north Sikkim and carving out verdant Himalayan temperate and tropical river valleys, traverses the Indian states of Sikkim and West Bengal and finally descends to Brahmaputra in Bangladesh. The total length of the river is 309 km, draining an area of 12,540 km 2 . The present study area includes the course of the River Teesta in West Bengal divided into ecological zones based on elevation gradient and habitat types (table 1, fig. 1). The river stretch was divided into four zones viz. the upper stretch (Rishi khola and Rungpo) where elevations is higher with low temperatures; the middle stretch (Teesta Bazaar) with low elevation; a lower stretch at Sevoke, where the river hits the plains; and lastly, the river plains (Gojoldoba, Domohoni and Haldibari). Along the longitudinal stretch of the river in West Bengal, covering a distance of 142 km, each site was sampled at regular intervals (bi-annually with pre-monsoon and post-monsoon visits) when flow conditions were the most stable and similar among sites. Local habitat attributes were recorded to find any associations with the variation in fish assemblages. Habitat variables for each sampling sites at each sampling operation were recorded in the field. Stream width was measured along three transects regularly spaced across the stream channel. Water depth, current velocity and temperatures were measured at the mid-point of each transect.

Ichthyologic biodiversity in perspective of longitudinal zonation concepts of River Teesta
Fish sampling was carried out from December 2010 to March 2013 every alternate six months at seven sampling areas (approximately 20-30 km apart) in the four zones covering the longitudinal gradient of the River Teesta at Darjeeling and Jalpaiguri districts in West Bengal. After an initial pilot survey of the entire riverine stretch, these seven areas were chosen based on different habitat patches, high fishing activity, accessibility and availability of local fish markets nearby (for gathering secondary data). Each sampling area was further divided into   Table 1. Longitudinal zonation concepts, the sampling areas and their hydrological and ecological characteristics. Fish zones: HmAz. High-mid altitude zone (moderate to high, 1,093-1,000 ft, elevation watersheds dominated by side slopes with gentle slopes and steep slopes); MAz. Mid-altitude zone (moderate, 628 ft, elevation watersheds dominated by side slopes and gentle slopes); LAz. Low altitudeplain zone (moderate to low elevation, 500 ft, watersheds dominated by gentle slopes with substantial areas of flats and side slopes; the river meets the plain at this site); RPz. River plains (low elevation, 380-187 ft, dominated by flats, pastured land and urban inhabitation). A sample reach of 50 m were fished for 2 hours at every site using the above-mentioned fish nets as well as the electro-fishing method using a single anode electro-fisher (300V, 3-4A, DC) operated by the same person. Captured fish specimens were counted and fixed in 10% formalin solution and, after 48 h, transferred to a 70% ethyl alcohol solution. Fishes were identified to the lowest taxonomic level using Shaw & Shebbeare (1937), Day (1889), Talwar & Jhingran (1991), Jayaram (2006Jayaram ( , 2010, and Menon (1987). All fish specimens were deposited in the fish collection repertoire at the Zoological Survey of India, Kolkata. The status of the species on the IUCN Red List of Threatened Species was incorporated. The divisions (table 1) of the zones is based on Aarts & Nienhuis (2003) and Aquatic Ecological System (AES) classification (Maxwell, 1995) and also adds some later subdivisions based on the present occurrence of the zones.

Ecological fish guilds
Fish data should be interpreted ecologically to yield information about riverine habitats and processes. In ecological studies, fish species sharing more or less the same niche are often grouped into guilds (functional groups) of species that exploit a resource (food or habitat) in a similar fashion (Bain et al., 1988;Bergers, 1991). Distributions of guilds were studied in space to give distinctly different information in prospect of the River Continuum Concept (Vannote et al., 1980). Identified fish species were grouped into guilds on the basis of classification related to river water flow regime, habitat use (Aarts & Nienhuis, 2003) and spawning habitat (Balon, 1975a(Balon, , 1975b(Balon, , 1981 to assess the underlying causes and ecological mechanisms of the present state of ichthyofauna of River Teesta in West Bengal (table 2). Balon (1975aBalon ( , 1975bBalon ( , 1981 classified fishes according to their spawning habitats and habits. His system is now used worldwide, with only minor adjustments, using ethological types (guarders and nonguarders), ecological groups (describing parental investment type), and substrate types as criteria.

Data analyses
Analysis focused on quantifying spatial variation in fish assemblages and identifying habitat variables explaining this variation. Because sampling effort (i.e., sample time, length and procedures) was similar among sites and years of sampling, counts of individual fish species at each sample site were directly used in the analyses. A number of diversity indices of the fish community structure in River Teesta were calculated using PRIMER (Plymouth Routines In Multivariate Ecological Research) v6 software package (Clarke & Gorley, 2006). Diversity indices included species richness (d), Pielou's evenness (J') (Pielou, 1969), and Shannon-Weaver (1949) index.   (Aarts & Nienhuis, 2003) and reproductive guild, based on spawning habitat/substrate (Balon, 1975) applied to the fish species of River Teesta in West Bengal). Tabla 2. Gremios según preferencia del curso fluvial (Aarts & Nienhuis, 2003) y gremios reproductivos basados en hábitat/sustrato de desove (Balon, 1975) (table 3). This is attributed to limitations induced by shifting and homogenous substratum and high turbidity as after this point River Teesta enters the Bramhaputra River drainage. The species richness per zone increases downstream (Gojoldoba and Domohoni) but decreases further downstream (Haldibari). Five freshwater fish orders have been deduced with Cypriniformes being the most dominant, followed by Siluriformes and Perciformes (fig. 2). Widely used in zoology, fish can be grouped into guilds according to their flow regime ecology and spawning habitat selection. The ecological classification applied in this study is the one based on the flow preference of adult fishes. It considers rheophilic (all stages of life confined to lotic waters); eurytopic (all stages can occur both in lotic or lentic waters) and limnophilic (all life stages confined to lentic waters) groups. In the present study, after detailed observation and analysis of the habitat requirements, we classified fish in the River Teesta as rheophilic, limnophilic or eurytopic ( fig. 3). Rheophilic fish species formed the dominant group in the upper reaches of the river where altitude was significantly higher. The proportion of rheophilic fish community more or less decreased sharply downstream and the proportions of limnophilic and eurytopic species increased. The stretches of the river falling in plains viz. Gojoldoba, Domohoni and Haldibari was characterized by stagnant zones, higher temperatures and less water current, as reflected in the increase in limnophilic and eurytopic species in these zones. The fish species were found to use seven spawning habitat types within each site (1 km 2 quadrate area considerations at respective sites) and were accordingly classified into seven spawning preference guilds ( fig. 4). Changes in flow preference and reproductive guilds were closely linked: in the rheophilic zone, lithophilic (50.0-58.0%) and psammophilic spawners (15%) were dominant in upper reaches, whereas limnophilic, phytophilic spawners and eurytopic phytolithophilic or polyphilic spawners predominated in lowland reaches. The regions preferred for spawning for respective fish species are illustrated in table 4.      (Hora, 1921) -  - (Hora, 1921) -

Discussion
Already in the 19th century, Eastern European ichthyologists had drawn up a rough classification system for the longitudinal succession of characteristic or dominant fish species that occur in rivers (Holcík, 1989). Huet (1949Huet ( , 1959Huet ( , 1962 improved this classic scheme by determining the characteristic physical and chemical parameters of each zone: the slope, the width, the depth, the current velocity and the water temperature. Downstream changes in fish assemblage structure along river courses have been a dominant theme in running water ecology (Hawkes, 1975). Aarts & Nienhuis (2003) divided the entire course of a river, from the spring to the sea, into five basic zones: trout (Salmo trutta), grayling (Thymallu thymallus), barbel (Barbus barbus), bream (Abramis brama) and smelt (Osmeruse perlanus) zone in near-natural and in regulated large rivers in Europe (the River Doubs in France and the Rivers Rhine and Meuse in the Netherlands). However, no such zonation concept has been applied to Indian rivers as such. In the present study, the stretch of River Teesta in the state of West Bengal, India was classified into four zones based on the physiological attributes of the riverine habitat and accordingly fish assemblage patterns were analyzed. The biodiversity pattern was delineated according to the river continuum model. What sets this model apart from others is the importance of the spatial arrangement of habitats for spawning and regugia (Schlosser, 1991). Incorporation of the River Continuum Concept in view of multimetric approach considers a river system as a longitudinal gradient of environmental and ecological conditions. This gradient ranges from a heterotrophic headwater regime (allochthonous nutrient sources) to a regime of autotrophy in midreaches (autochthonous production), followed by a gradual return to heterotrophic processes in the semi-lentic downstream waters. In view to the River Continuum Concept, fish diversity and composition of ecological guilds (functional groups) change longitudinally along this river continuum (Van der Velde & Van den Brink, 1994), and the highest biodiversity normally occurs in the midreaches, which are more productive due to warmer temperatures and high nutrient load creating maximum habitat diversity and environmental variations. Fish populations show a shift from cool water species low in diversity to more diverse warm water communities (Huet, 1949;Vannote et al., 1980). In the present study, highest species diversity was recorded at the midreaches (Gojoldoba) and the fish assemblage shifts from tolerant groups to thermally inclined specific groups both upstream and further downstream reaches. The study therefore elaborates the patterns in ecological guild structure that can be inferred from the predictions of the river continuum concept (Ward, 1998;Bhat et al., 2012). Besides, the elevation gradient (driven by water temperature and river substratum variations) seems to act as the main influencing factor for the observed fish assemblage patterns.
Moreover, in our study, the highest species evenness at Gojoldoba and Domohoni stretches was attributed to optimal habitat and environment conditions as seen by high water clarity and substratum with mosaic habitat patches. The Teesta River bed can be seen as a mosaic of different substratum patches, viz. mainly stones, gravels, sand and silt,that may affect differential nutrient uptake. The size, distribution and density of the patches enable the catchments as a whole to retain nutrients. Variations in patch characteristics that occur over extremes in spatial scales can influence stream structure and function (Angermeier & Karr, 1984;Hunt, 1971;Sheldon, 1968). The dynamics of fish populations are influenced by spatial variations in habitat patch mosaics, ranging in size from localized substratum patches to entire catchments. In stream segments, the presence of instream cover or habitat patches such as undercut banks, logs, etc., are important determinants of fish biomass, species diversity, and community composition. High water clarity ensures an increase in primary production of the aquatic body causing uniform nutrient cycling in the water column. This enables availability of a wide variety of food resources for fishes, so that all feeding groups can be sustained.