Mortality of vertebrates in irrigation canals in an area of west – central Spain

Mortality of vertebrates in irrigation canals in an area of west–central Spain.— Mortality patterns of vertebrates in irrigation canals have been poorly studied despite their potential impact on wildlife. Concrete irrigation canals in a cropland area in west–central Spain were monitored over 13 months to assess their impact on small fauna. A total of 134 vertebrates were found dead. Most were amphibians (86.46%) or mammals (20.90%), though fishes, reptiles and a bird were also recorded. Mortality peaked in autumm months. Corrective measurements are needed to reduce this cause of non–natural mortality.


Introduction
The effects of irrigation canals on vertebrates have been little studied in comparison with other causes of non-natural mortality.However, water conduction systems can affect vertebrates in a similar way to highways (Forman & Alexander, 1998;Rosell et al., 2002), creating a barrier effect or causing mortality.
Several authors (Cushman, 2006) consider that the barrier effect has a more negative impact than mortality in this setting, but certain populations of small and medium-sized vertebrates are at serious risk of drowning in irrigation canals (Ramos, 1992;Arranz, 1994;SCV, 2000;García, 2006a).Such mortality appears to occur in this setting simply because animals may fall into these canals as they move around their home ranges (see however Rosell et al., 2002;Peris & Morales, 2004).Several solutions have been proposed to avoid this cause of non-natural mortality, but little is known about their effectiveness.
In this work, we studied seasonal patterns of vertebrate mortality in irrigation canals in an attempt to further our understanding about the main groups involved and the extent of the involvement.Such knowledge could contribute to effective solutions in future.

Material and methods
The irrigation canals in this study were concrete structures designed for transporting water in the vicinity of Fresno Alhándiga (40º 42' 36.43'' N 5º 36' 16.58'' W;820 m a.s.l.;fig. 1), east of the province of Salamanca, in west-central Spain.This locality has a typical continental Mediterranean climate.Since the construction of the Santa Teresa dam, croplands have become the predominant land use.The area has a complex network of water bodies that includes rivers, streams and gravel pits, and well developed riparian forest.The irrigation canals are small and the water tanks, where the water coming from these ditches is stored, measure on average 81.44 ± 2.03 cm (mean ± SE; n = 29) length, 82.60 ± 2.16 cm wide, 154.97 ± 8.08 cm depth.The canals are used to conduct and distribute water to croplands in the region.Canals transport water from May to August, whereas water tanks remain full almost all the year, although their level diminishes during the winter.
To evaluate the impact of these irrigation canals on vertebrates from October 2004 and December 2005 we carried out monthly monitoring of water tanks along about 2 km of the irrigation canal system as it is in these structures where carcasses accumulated whenever animals died in the ditches.We studied a total of 29 water tanks.Animals found dead in the tanks were removed to avoid duplicate recordings.The average time taken to examine all the water tanks per month was one hour and 30 minutes, totally 24 hours and 30 minutes over the study period.As sampling was performed in the months of October, November and December in 2004 as well as 2005, the number of carcasses found per month was divided by the number of times that surveys were done per month to standardize.The differences in mortality rates between groups of animals and seasons were compared by means of a chi-squared test (χ 2 ).
Months of higher incidence were September, October and November (χ 2 = 194.44,df = 11, P < 0.001; fig.1).Mortality began to diminish as of the second fortnight of November and it remained at minimum levels during the rest of the year, with a small peak between May and July (fig.1).

Discussion
Mortality in ditch-related water tanks may affect animal density and dynamics, usually reducing the population size to some extent (Sccocianti, 2001).The incidence in the study area was relatively high (Ramos, 1992;Arranz, 1994;SCV, 2000;Sccocianti, 2001;García, 2006aGarcía, , 2006b)), with an average of 8.93 ex./month.Most were amphibians (71.64%), in accordance with previous works on canals (Ramos, 1992;Arranz, 1994;SCV, 2000;Sccocianti, 2001;García, 2006aGarcía, , 2006b)).Mortality peaked in September, maybe due to the dispersal of amphibians born in summer and spring (García-París et al., 2004) or as a consequence of changes in the cycles of crops, such as harvesting, causing animals to migrate.Counterintuitively, in spring, when amphibians were supposedly breeding, and therefore more active and carrying out breeding migration (García-París et al., 2004), mortality reached the lowest values, owing those from winter.With presently available data it is difficult to account for factors affecting this temporal pattern.
The construction of infrastructures for water transport for diverse uses with scarcely surface and depth, and also sections that produce less water velocity (Scholz & Trepel, 2004), could be able to avoid the detected small vertebrate mortality.Further research is needed to ascertain the impact of this kind of nonnatural mortality on vertebrate populations.

Fig. 1 .
Fig. 1.Monthly distribution of the animals found dead within the water tanks in the study area: J. January; F. February; M. March; A. April; My.May; Jn.June; Jl.July; A. August; S. September; O. October; N. November; D. December.

Table 1 .
Vertebrate mortality in west-central Spain water deposits from 2004-05: I. Incidence (number of animals found).