First characterization of the spawning habitat and mating behaviour of Twaite shad in the Ebro River (Western Mediterranean)

Abstract

The Ebro River is the largest river of the Iberian Peninsula, and one of the most important western Mediterranean rivers. Mean river flows attain about 400 m3 s−1, with minimum summer flows of around 100 m3 s−1. Peak discharges of one to several thousands of m3 per s are common during winter and autumn, and this occurs despite the significant hydrological disruption of the flow regime imposed by hydroelectric dams constructed during the 1960–70s. Once abundant in the Ebro River, and found upstream for hundreds of kilometers during its spawning migrations, the Allis and Twaite shads became severely restricted to the 63 final km of the river due to river fragmentation caused by the construction of a weir (Xerta-Tivenys weir) as early as the XV–XVII centuries (Fig. 1). The congregation of shads under the weir allowed the existence of local fisheries of these species to persist until the 1970s, when the last shad fishery was abandoned. After 40 years with scarce scientific records regarding the presence of Twaite shad in the lower Ebro River, the occurence of the species seems to be gradually increasing and the species can be considered as a seasonal member of the fish fauna assemblage, appearing to be the only anadromous species that can complete its life cycle under these severe restricted access to upstream habitats (López et al., 2007). Map indicating location of study area upstream from the Ebro River Delta. Extent of marine influence in river channel is shown in dark gray. The freshwater habitat is shown in light gray, and location of the weir indicated by a diagonal black line. Aerial view showing spawning grounds in light gray colored oval areas within river channel. Aerial image obtained from Google Earth The migratory and spawning behaviour of Twaite shad and Allis shad have received considerable attention in different European rivers. Various studies have described that within each species the spawners have similar ecological characteristics in terms of demography, migratory activity and reproductive behaviour (Aprahamian et al., 2003a,b; Acolas et al., 2004). However, in the Ebro River, knowledge on the biology of Twaite shad and its ecology is almost non-existent. Basic information on the characteristics of habitat found useful by this species for successful spawning are mostly missing. Some knowledge on the spawning ground characteristics for Twaite shad in the lower reaches of the Ebro River are urgently needed in order to develop proper conservation measures and locate potential spawning grounds upstream of the river barriers currently in place. In addition, studying fish mating behaviour is a valuable tool for assessing the reproductive potential, and the long-term viability of the population. The objective of this study was, therefore, to identify and describe the spawning grounds recently discovered below the weir at Xerta, while also documenting the mating behaviour and local movements of the Twaite shad spawners in the lower reaches of the Ebro River. The study site is located approximately 63 km from the mouth of the Ebro River next to the town of Xerta and just below a hydroelectric complex next to a weir. The main area of study was just below the weir along a 1 km stretch; however a larger region of 5 km was surveyed for potential spawning behaviour and evaluation of spawning habitat. The study was undertaken between 2007 and 2009. The first localization of spawning grounds was carried out by extensive night surveys following initial data reported by López et al. (2007). Observers were placed according to a predetermined grid of the river where mating episodes from previous years were observed (years 2005 and 2006). The surface delimitation and location of individual mating episodes was made by in-situ measurements with GPS from four independent observers in kayaks that were placed in midstream and 200 m equidistant from one another downstream from the weir where previously, the spawning activity had been noted (total river length analyzed: 750 m). The reproductive activity was determined by visual and auditory recordings of surface mating episodes through a 22:30–05:00 hour time window (GMT + 2), during observation excursions on a weekly basis during three consecutive spawning seasons (May to August: 2007–2009). Spawning habitat characteristics were recorded included depth, water velocity (at 0.5 m depth) and substrate categorization by visual determination. All measurements were performed within transversal transects spaced 150 m apart (5–7 points on each transect) on the main spawning ground. Historical details of the relict fisheries were obtained through questionnaires and interviews with the older local citizens who participated in the fisheries in order to contrast our biological studies with the empirical and traditional knowledge of fishermen. One main reproductive area has been found between 0.5 and 1 km downstream of the river weir. A second area, located at 4 km downstream, represents only 4% of the recorded spawning episodes (Fig. 1). The main area is a continuous stretch of approximately 1.2 km of length located along the right river bank. Total length surveyed comprised a stretch of 5 km of the river downstream of the Xerta weir. River width in this area ranges from 85 to 122 m. Depth and flow velocity profiles of this spawning ground are represented in Fig. 2c,d, respectively. According to data shown in Fig. 2, the spawning grounds for Twaite shad in the Ebro River are characterized by moderate to high water velocities (0.72 m s−1 for a circulating river flow of 220 m3 s−1; CHE, 2009), with variable depths, and a substrate composed of heterogeneous clean gravels without silting. Other areas with silting gravels or dominated by sand and with a broad coverage of macrophytes and moderate water currents (0.33–0.01 m s−1) were not used by the shads. No mating episodes were recorded directly in front of the hydropower station alongside of the weir, probably due to the high water flows and velocities close to the outlet of the power station (approximately 1.5–2 m s−1). The spawning grounds used by Twaite shad during the studied period corresponded to those where old fisheries for this species were located during the 60s. In addition, personal interviews with fishermen aged between 75 and 90 years old revealed that they were able to recognize the Twaite shad mating behavior and profit from recognition of this behavior (without knowing the biological relevance of this) by targeting this area for their fishery. Study area with the location of the episodes within the spawning grounds (a), river transects made for habitat measures (b), profiles of depth in meters (c), and velocity in m per s (d). Horizontal white lines indicate the region within each profile where spawning observations coincide Night observations revealed that mating activity took place between 23:45 and 04:05 hour, with a peak activity being around 02:15 hour. Half of the total events registered took place between 01:45 and 02:45 hour (Fig. 3a). The same pattern of spawning events was observed during three consecutive years (n = 522), but the peak of activity showed some interanual variability (2007 – 02:15 hour; 2008 – 02:05 hour; 2009 – 01:45 hour). Mating activity recorded with eco-cameras (data not shown) was observed only during the night, supporting other visual and auditory observations. During the mating activity window, erratic fast movements of fish appeared, with continuous underwater interactions that we interpreted as an active search by males for females. When a mating event started (perhaps as an interaction with only two or few individuals of opposite sex) it seemed that other individuals quickly reached the area of mating activity. These episodes had a very short duration (approximately 2–5 s) and usually seemed to involve only a few individuals (maybe 2–5 fish estimated from direct visual observation). Some pre-mating surface activity was not uncommon, and when present was located upstream from the spawning episode. This trait was tentatively interpreted as a signal or exposing behaviour conducted by females for attracting males. Other well known mating cues have also been recorded, such as the surface tail splashing behavior which is a common feature during mating activities in species of this group (Beasley and Hightower, 2000; Cassou-Leins et al., 2000; Aprahamian et al., 2003b). Spawning activity expressed as episodes occurring during night hours (time in GMT + 2). Pooled data ranging from May to June from three consecutive years (2007–09) (n = 522 episodes) One of the prominent observations was derived from the knowledge of the older fishermen. These persons exhibited a very high capacity to discriminate between the two species of shads inhabiting the Ebro River during past decades, since both species had different market values. Using the outcome of questionnaires, and using the traditional knowledge to discriminate safely between these species, we concluded that the Twaite shad was relatively common during the second half of the 20th century while Allis shad was present in the Ebro river, but at a low percentage (1–5% of the total landings). In contrast, no recent captures or recent observations of Allis shad have been made during the study period in the Ebro River in which shads extended from 2004 to the present. Present results revealed that Twaite shad in the Ebro River represent an annual and functional migrating population that has existed for at least the last 5 years. Field data showed that despite the insurmountable obstacle imposed by the barrier of the Xerta-Tivenys weir it seems that the species actively selected some characteristics of the habitat for spawning. In particular, fast flowing waters (approximately 1–1.5 m s−1) with a substrate composed by heterogeneous clean gravels without silting were chosen by Twaite shad spawners. Additional research is needed to improve our knowledge that can be used for evaluating the availability of potential upstream habitat which could be increased by approximately 200% if fish were able to overcome the Xerta weir. Future conservation and management plans for Twaite shad in the Ebro River need to consider the removal of barriers through the installation of proper shad-targeted fish-ladders, and surveys of additional potential spawning habitats in portions of the drainage below the Xerta weir not yet investigated. This study was partially funded by the IRTA and the AGAUR (2007ACOM 00063, Generalitat de Catalunya). Funding was provided to E.G. and Fundació del Parc Natural del Delta de l’Ebre (PNDE), respectively. Authors are indebted to V. Castan, T. Cartes, J. Arnal, M. Falcó, A.J. Avinyó and J. Roig for their help and support tasks during field sampling and animal capture. The authors have declared no potential conflict of interests.