SEARCH

SEARCH BY CITATION

Keywords:

  • Lutra lutra;
  • reservoir;
  • monitoring;
  • impoundment;
  • flooding;
  • ecological requirements;
  • diet;
  • Mediterranean region

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  1. The number of dams is progressively increasing, but nonetheless there are few examples of long-term studies involving otters and these structures.
  2. This study assessed how Eurasian otter Lutra lutra responded over time to environmental changes imposed by the construction of a large reservoir, the Alqueva Dam (250 km2, SE Portugal).
  3. Otter distribution was monitored over four phases (pre-deforestation/flooding, deforestation, flooding, and post-flooding) from 2000 to 2006 in both the flooded and surrounding areas. During each phase, presence was assessed at two resolutions: 25 km2 (76 survey sites in wet and dry seasons) and 1 km2 (39 survey sites, every three months). In addition, otter spraints were collected at eight survey sites in 2000 (pre-dam) and 2006 (post-dam) to assess and compare otter diet with prey availability.
  4. Otters were widespread before dam construction (>82.4% positive cells), decreased during deforestation (68.5%), and particularly during the flooding phase (33.3%), recovered during the post-flooding phase, although not at the level recorded before dam construction (61.5–83.3%). Otter diet was dominated by fish and American crayfish Procambarus clarkii during the pre-deforestation/flooding phase (56.7% fish, 35.3% crayfish) and at the end of the post-flooding phase (60.7% fish, 33.2% crayfish). Species richness of fish prey decreased with flooding (16 to 8), as did the ratio of native/non-native fish species (1.7 to 0.3).
  5. Changes in the availability of the main ecological requirements of otter were similar to the changes in the observed otter distribution in the flooded area throughout the impact phases, and decreased during deforestation and flooding with some recovery in the post-flooding phase.
  6. The results emphasize the importance of long-term monitoring studies that include several post-impact phases, to evaluate species response to impacts. This will allow better planning of mitigation and compensation measures.

Copyright © 2013 John Wiley & Sons, Ltd.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Adaptation to habitat change, loss, and fragmentation is a key aspect of species conservation, especially when resulting from the construction of infrastructure that poses new challenges for their survival. Monitoring species' responses to habitat change is crucial, therefore, to predict species trends in human-altered environments. Climate change and an increase in water demand by humans contribute to conservation pressure for aquatic species. This is especially so in the Mediterranean basin, which is considered to be one of the regions that will face the largest changes in climate worldwide (Giorgi, 2006), and where water management is conducted largely through river regulation (dams) (Collares-Pereira et al., 2000). Dams result in large-scale habitat disturbances with major impacts on fish and other riparian populations and habitats (Collares-Pereira et al., 2000; WCD, 2000). These disturbances can be greatly exacerbated when habitats, such as those in the Mediterranean region, are subject to seasonal variation in water flow. Here, a stress period usually occurs in summer, when flowing waters become ephemeral, especially following long periods of drought. Reservoirs can further affect this situation by influencing water flow regimes and acting as barriers to species movement (Ruiz-Olmo et al., 2001).

The Eurasian otter (Lutra lutra) is a semi-aquatic mammalian carnivore that occupies an apex position in the food web of European fresh waters, feeding on a wide range of prey, especially fishes (Almeida et al., 2012). One of the main threats for otters living in inland waters of the Mediterranean is the exacerbation of a naturally unstable water flow in rivers and streams (Jiménez and Lacomba, 1991). Flow reduction is often a consequence of river damming and of increased water demand, particularly for irrigation (Ruiz-Olmo et al., 2001). It is also very likely that climate change will contribute to the decrease in suitable otter habitat, in particular in the Iberian Peninsula. This is linked to a potential increase in drought frequency, extent, and intensity as the climate warms, which may lead in some cases to the disappearance of shallow water bodies, including small rivers (Cianfrani et al., 2011), thereby causing major declines in prey abundance. Owing to the dependence of otters on freshwater environments, the preservation of riverine habitats is, and always has been, a major priority for otter conservation because it is rich in prey, offers adequate conditions for sheltering, and facilitates animal movement (Foster-Turley et al., 1990). Given these characteristics, the otter is an ideal model species to address animal adaptation to habitat loss and change as caused by dam construction.

Dams, especially large ones (wall height ≥15 m or height 5–15 m and reservoir volume >3×106 m3; WCD, 2000), have been described as having an adverse effect on the distribution of Eurasian otters, and are putatively a contributing factor in their past decline in Europe (Foster-Turley et al., 1990; Macdonald and Mason, 1994). Some of the significant impacts on otters caused by dam construction include direct habitat loss and destruction, fragmentation of populations, and reduction of adequate foraging grounds caused by the creation of large and deep reservoirs (Foster-Turley et al., 1990; Macdonald and Mason, 1994; Kruuk, 2006; Santos et al., 2008). In Mediterranean regions, despite evidence of otter presence and the use of established large reservoirs (Prenda et al., 2001; Pedroso and Santos-Reis, 2006), there is a lack of information on otter response in all dam construction phases (pre-dam, flooding, post-dam). Consequently, evaluation of impacts is currently incomplete. This is especially relevant because for many Environmental Impact Assessments of large dams, evidence of otter presence in the reservoirs is used to justify either that no impacts occurred or that there were no significant consequences for otter populations after the change from a river to a reservoir. Long-term monitoring studies are therefore required, covering all phases of dam construction. This is important because of the increasing number of such structures, especially in Mediterranean countries where water management is largely based on dam building, and under the current scenario of climate change that affects river systems mainly by extending the drought period.

The aims of the present study were to: (1) assess changes in otter distribution during the different impact phases of constructing a large dam; (2) assess changes in otter diet according to changes in prey availability before and after dam construction; (3) assess changes in the availability of the main ecological requirements of otters (prey and feeding areas, resting sites, breeding areas, corridors for movement and dispersal, and fresh water) during and after construction of the dam and relate them to changes in otter distribution.

METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Study area

The Alqueva Dam was built in the Guadiana River valley, in south-eastern Portugal. Construction started in 1998 and its completion, in late 2003, created Europe's largest artificial lake, flooding an area of 25 000 ha (Figure 1). The Alqueva Dam project includes a massive irrigation system affecting 120 600 ha, with profound changes in the surrounding landscape structure, land cover, and use. The Alqueva Dam is in the Mediterranean region, considered to be one of the biodiversity hotspots for conservation priorities (Brooks et al., 2006). South-eastern Portugal is characterized by a multi-patch landscape, dominated by holm oak (Quercus ilex) and cork oak (Quercus suber) woodlands, interspersed with agricultural fields (cereals, vegetables, and olive-yards) and forest plantations (eucalyptus and pine). The climate is characterized by mild winters (minimum temperature average for January is 4 °C) and hot summers (average maximum temperature for July is 34 °C with maximum temperatures that rise to 45 °C). Median annual precipitation varies between 400 and 600 mm. Most rivers and streams in this region are ephemeral.

image

Figure 1. Location of the Alqueva Reservoir in southern Portugal, showing the 25 km2 and the 1 km2 (black squares) otter survey grid cells and the sites (black circles) where otter diet and prey were assessed. Two sites (G1 and G2) were located in the main Guadiana River. Other sites were located in tributaries of the Guadiana River: the Azevel stream (AZ), Álamo stream (AL), Degebe stream (D1 and D2), Alcarrache stream (ALC) and Zebro stream (Z).

Download figure to PowerPoint

The response of otters to the construction of the Alqueva Dam was monitored during four periods (monitoring phases): pre-deforestation/flooding (2000), deforestation (2001), flooding (2002 and 2003) and post-flooding (2004 to 2006). The study area was covered by 11 1:25000 Portuguese military maps (441, 452, 463, 474, 481, 482, 483, 490, 491, 492 and 501) that included the flooded area of the artificial reservoir created by the construction of the Alqueva Dam (the area directly affected) and the surrounding area. An additional 3-year period was monitored to include the post-flooding phase. During dam construction, vast areas of vegetation were removed (deforestation) in order to reduce the biomass and potential eutrophication of the future reservoir waters, resulting in a large area where trees and shrubs had been removed up to the predicted maximum water level (152 m asl). Flooding started in February 2002 and until the end of 2003 the water level and flooded area increased considerably. The post-flooding phase corresponds to the slow final flooding of the area (from 140 m to 152 m asl), and represents, therefore, a relatively more stable water level.

Field methods

Distribution of otters

Otter distribution was studied at two resolutions, 25 km2 and 1 km2. First, the study area was divided into a grid of 25 km2 cells, resulting in a total of 76 survey cells (Figure 1). Otter presence/absence in all 76 cells was assessed following the World Conservation Union (IUCN) Otter Specialist Group (OSG) survey guidelines (Foster-Turley et al., 1990; Reuther et al., 2000) by searching for otter signs (spraints, scent marks, prey remains, and footprints) using a transect length of 600 m per site (Macdonald, 1983). The survey in each cell was interrupted as soon as a sign was detected. Surveys were conducted twice each year (wet and dry seasons). In each year, otter presence in each 25 km2 cell was classified as ‘permanent’, if both seasonal surveys were positive; as ‘occasional’ if at least one seasonal survey was positive; and ‘absent’ if both surveys were negative.

To evaluate in detail the change in otter response in the area directly affected, 39 1 km2 cells were randomly selected within the future flooded area and following the same OSG method but fixing the surveyed length at 600 m, each cell was surveyed every 3 months (trimester) for otter signs. The transect location was adjusted following the rise of the water level, that is, new ‘reservoir transects’ were established as old ‘stream transects’ disappeared under water and were abandoned.

Diet and prey availability

Otter diet and prey availability were assessed in the pre-deforestation/flooding phase (2000, five sampling periods: February, April, June, August and October) and at the end of the post-flooding phase (2006, three sampling periods: April, August and December) for comparison purposes.

Otter diet assessment was based on the identification and counting of prey found in spraints collected at eight fish survey sites (Figure 1) and was expressed as percentage of occurrence (PO (item A) = total number of individuals of prey item A consumed / total number of individuals consumed × 100). Food items were identified to the lowest possible taxonomic level following a standard approach (Sales-Luís et al., 2007) and using a reference collection of fishes and other vertebrates, as well as published literature (Prenda and Granado-Lorencio, 1992; Conroy et al., 1993; Prenda et al., 1997, 2002).

Data on fish availability for the stream survey sites were gathered in 2000 within the framework of another study (M.J. Collares-Pereira et al., unpubl. data). Sampling was carried out using standard electrofishing gear, fitted with a single 30-cm anode (300 V, 4–6 A, DC). In each of the eight sites a 50 m stretch was fished for 30 min. As electrofishing becomes increasingly more selective as waters become more lentic (Godinho et al., 1998; Clavero and Hermoso, 2010), complementary methods were used to assess prey availability in 2006 (after flooding). Because otters prefer shallow waters for foraging (Kruuk, 2006), two fyke-nets were placed near and parallel to the shore for fish capture, combined with three carboy traps, adapted and baited for capturing American crayfish Procambarus clarkii. Also, one trammel net (15 m × 2 m; inner mesh: 25 mm; outer mesh: 100 mm) was set at a depth of 1.5 m at a minimum distance of 150 m from the margin. All sets were left overnight. All captured individuals were identified, counted, weighed, measured, and then released into the water. Data were converted into percentage of occurrence (as defined in Ribeiro et al., 2006). Fish species that appeared rarely both in the otter diet and in the environment were removed from the diet analysis.

Ecological requirements

The availability of otter ecological requirements (IUCN OSG, 2009) was characterized in each surveyed 1 km2 cell, distributed in the future flooded area of the artificial lake. Five main requirement groups were assessed: (1) availability of prey and feeding areas; (2) availability of resting sites; (3) suitability for breeding areas; (4) availability of corridors for movement and dispersal and (5) accessibility to fresh water. Each variable was categorized using a 1 (minimum) to 5 (maximum) scale and ranked from the worst to the best available environment for the otter – availability index (Table 1). The value associated with each requirement in each cell was the mean of the corresponding variables for that requirement (Table 1) in that cell. Otter ecological requirements were assessed in the field by experienced observers and classified using available literature (Beja, 1992; Ruiz-Olmo et al., 2005; Kruuk, 2006; Pedroso and Santos-Reis, 2006; Ribeiro et al., 2006; Pedroso et al., 2007; Sales-Luís et al., 2007; IUCN OSG, 2009; Basto et al., 2011). Results were then related to marking intensity (number of detected otter signs per km). Spraint surveys have been widely used over the years as a surrogate variable to assess otter distribution, identify habitat features considered of importance to otters, and to indicate population status. Sprainting activity varies with sex, season, and with social and reproductive status (Macdonald, 1983; Kruuk, 1992; Kranz, 1996). These observations lead to the conclusion that the number of spraints cannot be used to assess otter densities, although it is a useful tool for comparing habitat, resources, and intensity of use between sites (Kruuk, 2006; Sulkava, 2006; Guter et al., 2008). This is especially relevant when working with data from well-distributed otter populations, when negative sites are usually few, and so the additional information that marking intensity gives is relevant. Therefore, marking intensity was used in this study as a surrogate for the cell's otter use frequency by considering that survey sites with more spraints indicate that the cell is important to the otter in terms of the habitat and/or resources.

Table 1. Otter requirements, and variables related to each requirement, recorded in flooded area of the Alqueva Reservoir. Data for all variables obtained from fieldwork and maps, and for 'resting' variable from fieldwork alone
Requirement / variableVariable descriptionData categories or units (availability index)
FeedingAvailability of prey and feeding areas1 - minimum to 5 - maximum
PreyAbundance of amphibians, crayfish and fish1–absent / 2–scarce / 3–moderate / 4–high / 5–very high
SlopeBank slope1– very steep / 2-steep/ 3– moderately steep / 4–slightly steep / 5–flat
TypologyBank typology1–peninsula / 2–exposed bank / 3-exposed bank with irregular perimeter/4–bay / 5– narrow bay or deep valley
 
RestingAvailability of resting sites1 - minimum to 5 - maximum
Bank refuges (above flooded area)Refuge availability: presence of large rocks, logs and other type of refuge structures1–absent / 2–scarce / 3–moderate / 4–high / 5–very high
Bank refuges (in flooded area) 1–absent / 2–scarce / 3–moderate / 4–high / 5–very high
Bank vegetation (above flooded area)Vegetation availability for otters refuge1–absent / 2–present but offering no suitable cover / 3–present in patches, offering scarce cover / 4–present in large patches offering suitable cover / 5–continuous dense vegetation, providing excellent cover
Bank vegetation (in flooded area)Vegetation availability for otter refuge1–absent / 2–present but offering no suitable cover / 3–present in small patches, offering scarce cover / 4–present in large patches offering suitable cover / 5–continuous dense vegetation, providing excellent cover
 
BreedingSuitability for breeding1 - minimum to 5 - maximum
Breeding watercoursesNumber of watercourses with breeding conditions1–absent / 2–one / 3–two / 4–three /5–more than three
Natal holtsPotential for natal holts (rock formations, deep tree roots with holes, thick vegetation systems)1–absent / 2–scarce / 3–moderate / 4–high / 5–very high
Rearing areasPotential for rearing areas (protected areas with lower current, very dense vegetation and rich food supply, stable availability of water)1–absent / 2–scarce / 3–moderate / 4–high / 5–very high
 
CorridorsAvailability of corridors for movement and dispersal1 - minimum to 5 - maximum
Corridor numbersNumber of potential corridors (watercourses)1–absent / 2–one / 3–two / 4–three /5–more than three
Corridor typeType of corridors1–small stream without water / 2-small stream / 3–stream / 4–river / 5–large river
RefugeRefuges/vegetation availability for otters in potential corridors1–absent / 2- present but offering no suitable cover / 3–present in small patches, offering some cover / 4–present in large patches offering suitable cover / 5-continuous dense vegetation, providing excellent cover
 
WaterAccessibility to fresh water1 - minimum to 5 - maximum
Watercourse numberNumber of watercourses1–absent / 2–one / 3– two / 4–three /5–more than three
Watercourse typeType of watercourses (given resistance to drought and carrying capacity)1–none / 2-small stream / 3–stream / 4–river / 5–large river or reservoir

Statistical analyses

Differences in otter distribution in the 25 km2 grid over the years and between seasons were analysed using a chi-square test of homogeneity. The 1 km2 grid results were used to analyse the impact and/or influence of the different phases on the presence–absence of otter over time. To account for multiple surveys of the same location, a generalized additive mixed model (GAMM) regression framework was used (Wood, 2006). Presence/absence of otter signs was modelled by smoothing of trimester values, with the smoothness chosen by using the default generalized cross-validation procedure in the R mgcv library (Wood, 2006; R Development Core Team, 2011). Grid was included as a random effect and the residuals within sites were assumed to follow a first-order autoregressive model. Spearman correlations were calculated between the five otter requirements rankings and marking intensity in each 1 km2 cell in each sampled trimester. Consumption of prey categories between 2000 and 2006 was also compared using a chi-square test. Otter dietary niche breadth was estimated by a normalized Levins’ index (Bn) (Feinsinger et al., 1981): B = 1/R∑pi 2, where pi is the relative percentage of frequency of each prey category, and R is the food category. The Bn index was used to characterize 2000 and 2006 diet. Shannon–Wiener diversity index (H′) (Krebs, 1989) was also used to characterize the prey diversity in 2000 and 2006. Since fishing methods used in each year are not directly comparable, prey availability results use was limited to comparing species presence and relative abundance within each year. Statistical calculations were performed using R software v. 2.14 (R Development Core Team, 2011).

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Otter distribution

Temporal dynamics of otter distribution over the different phases and years of monitoring at the 25 km2 grid resolution is shown in Figure 2.

image

Figure 2. Dynamics of otter distribution in the Alqueva Reservoir area in a 25km2 grid along the monitoring phases and years.

Download figure to PowerPoint

The percentage of area occupied by the species in these 25 km2 cells was fairly stable across years within seasons and, although in the last two years sampled values were seemingly lower in the dry season, no significant differences were detected (dry seasons: χ2 = 11.8, P = 0.06; wet seasons: χ2 = 7.0, P = 0.32) (Table 2). During all monitoring phases, otter showed a widespread distribution in both seasons, occupying rivers, streams, small reservoirs, and the Alqueva Reservoir itself. Most negative cells found corresponded to areas with few streams or small streams that are dry for long seasonal periods. Also, in cells with occasional presence, otter absence was always recorded during the dry season.

Table 2. Percentage of survey cells with otters present in the Alqueva Reservoir area in a 25 km2 grid during monitoring phases and years
YearDry season (%)Wet season (%)
200084.284.2
200185.585.5
200285.585.5
200384.284.2
200480.380.3
200568.472.4
200676.378.9

The percentage of area occupied was always similar but the location of negative cells shifted between years. During the deforestation phase (2001), three cells that previously had otter signs were abandoned despite having apparently adequate streams and/or small reservoirs. The early flooding phase (2002) showed the most considerable cell shifts; most 2001 cells without otter signs were positive for otter presence in 2002. Since 2003, the otter was no longer found in three grid cells that had become almost totally flooded. During the post-flooding phase (2004–2006) otter distribution did not differ significantly from previous phases, although 2005 had the lowest record of otter presence.

At the 1 km2 scale otters were widespread before dam construction (>82.0% positive cells), then decreased during deforestation and the flooding phase, and recovered in the post-flooding period (Figure 3). The trimester surveys showed major impacts during early deforestation (first trimester of deforestation – 68.5%), with some signs of recovery immediately afterwards, and another major decrease in otter presence during early flooding (first two trimesters after the beginning of flooding – 33.3%). Cell occupancy showed a recovery during the post-flooding phase, although not to the level before dam construction (61.5–83.3%). Modelling presence/absence showed that the smooth term is highly significant (smooth term P-value < 10-6) emphasizing that there was a non-linear relationship between otter presence and time.

image

Figure 3. Probability of otter presence as a function of trimester in the flooded area of the Alqueva Reservoir. Data are represented by points and fitted model by a black line.

Download figure to PowerPoint

Otter diet and prey availability

In total, 858 otter spraints were collected and analysed (pre-deforestation/flooding phase, year 2000, n = 675; end of post-flooding phase, year 2006, n = 183) resulting in 2583 identified prey occurrences. Otter diet was dominated by fish and crustaceans (P. clarkii) both in the pre-deforestation/flooding phase (56.7% and 35.3% of occurrences) and at the end of the post-flooding phase (60.7% and 33.2% of occurrences; Figure 4). Insects and mammals were only consumed in the streams and/or rivers. No significant difference was found in the overall consumption of fish, crustaceans, amphibians, reptiles and birds between both phases.

image

Figure 4. Percentage of occurrence (PO) of prey classes in otter diet in the flooded area of the Alqueva Reservoir (year 2000 = pre-deforestation/flooding phase; year 2006 = end of post-flooding phase).

Download figure to PowerPoint

The number of fish species preyed on by otter decreased after dam construction (16 to 8) as well as the ratio of native/non-native fish species (1.7 to 0.3). In 2000, diet was dominated by pumpkinseed Lepomis gibbosus, barbel Barbus spp. and chub Squalius spp. By 2006 diet was dominated by L. gibbosus, mosquito fish Gambusia holbrooki and largemouth bass Micropterus salmoides, all non-native species. Calandino Squalius alburnoides, southern Iberian chub Squalius pyrenaicus, Iberian long-snout barbel Barbus comizo, Iberian small-head barbel Barbus microcephalus, Iberian arched-mouth nase Iberochondrostoma lemmingii, Iberian straight-mouth nase Pseudochondrostoma willkommii, Southern Iberian spined-loach Cobitis paludica, chameleon cichlid Australoheros facetus and freshwater blenny Salaria fluviatilis were not recorded in the diet after dam construction. The black bullhead Ameiurus melas, only captured in 2006, also became important in the otter diet in the post-flooding assessment (Figure 5). Significant differences were found in consumption of Barbus sclateri/steindachneri2 = 5.9, P < 0.05), which was higher in 2000, and G. holbrooki2 = 9.3, P < 0.05), higher in 2006.

image

Figure 5. (A) Fish species captured in 2000 by electro-fishing in streams and rivers of the future flooded area of the Alqueva Reservoir; in 2006 by fyke-nets and carboy traps in the Alqueva Reservoir margins, by trammel nets in the Alqueva Reservoir (relative abundance). (B) Percentage of occurrence (PO) of fish species in otter diet in the flooded area of the Alqueva Reservoir (year 2000 = pre-deforestation/flooding phase; year 2006 = end of post-flooding phase); non-native (nn) species indicated; Amel – Ameiuras melas; Afac – Australoheros facetus; Cpal – Cobitis paludica; Ghol – Gambusia holbrooki; Ccar – Cyprinus carpio; Psp – Pseudochondrostoma sp.; Pwil – Pseudochondrostoma willkommii; Bsp – Barbus sp.; Bscl – Barbus sclateri/steindachneri; Bmic – Barbus microcephalus; Bcom – Barbus comizo; Spyr – Squalius pyrenaicus; Salb – Squalius alburnoides; Msal – Micropterus salmoides; Lgib – Lepomis gibbosus; Unid – Unidentified (species with PO < 1% both in the otter diet and in the environment not included: Carassius auratus, Iberochondrostoma lemmingii and Salaria fluviatilis).

Download figure to PowerPoint

Seventeen fish taxa (15 species and two genera) were captured during 2000, five of them non-native. Lepomis gibbosus, G. holbrooki and S. alburnoides dominated (77.0% of relative abundance; Figure 5). In 2006, the prey captured using fyke-nets and carboy traps in the Alqueva Reservoir were fish (89.4%), reptiles (Mediterranean turtle Mauremys leprosa – 7.5%), and crustaceans (P. clarkii – 3.1%). Of the fish species, L. gibbosus dominated captures in the reservoir margins (fish traps) and in the reservoir itself (trammel nets). Seven fish species, five of them non-native, were captured in the margins. Seven fish species were captured in the reservoir, but five of them were native (Figure 5).

Diet analysis reported similar and low Levins’ index values for both years (2000: B = 0.239; 2006: B = 0.238) suggesting that otters preyed consistently on a small number of abundant species (fish and crayfish; Figure 5). The Shannon index also showed a low diversity in the diet along with a small decrease in that diversity from 2000 (H′ = 2.260) to 2006 (H′ = 2.029).

Otter ecological requirements

Most suitability indices for otter ecological requirements declined steeply during the flooding and deforestation phases, particularly in the early stages (Figure 6). During the late- to post-flooding period (2004–2006), feeding, corridors and resting showed slight increases. The only otter requirement that increased markedly was the availability to fresh water, hardly surprising given the presence of the new reservoir. Marking intensity was consistently correlated with several otter requirements through time, although there were changes in the degree of correlation and the type of requirement (Table 3). Significant negative correlations were only found between water availability and marking intensity.

image

Figure 6. Yearly average of availability index of otter ecological requirements from 2000 (pre-deforestation/flooding) to 2006 (post-flooding) in the flooded area of the Alqueva Reservoir.

Download figure to PowerPoint

Table 3. Correlations between otter requirements and marking intensity in the flooded area of the Alqueva Reservoir during monitoring phases and years (Cc – Spearman correlation coefficient; P – significance)
Monitoring phase / yearOtter requirements
 feedingrestingbreedingcorridorswater
  • Requirements: feeding – availability of prey and feeding areas; resting – availability of resting sites; breeding – suitability for breeding areas; corridors – availability of corridors for movement and dispersal; water – access to fresh water.

  • **

    Highly significant (P < 0.001);

  • *

    significant (P < 0.05).

Pre-deforestation/flooding – 2000Cc0.708**0.753**0.736**0.874**0.483**
Marking intensityP0.0000.0000.0000.0000.009
Deforestation – 2001      
Marking intensityCc0.373*0.488**0.351*0.562**0.190
 P0.0130.0010.0190.0000.216
Early flooding – 2002      
Marking intensityCc0.673**0.451**0.535**0.570**-0.153
 P0.0000.0000.0000.0000.094
Late flooding – 2003      
Marking intensityCc0.387**0.2010.262*0.297**-0.220**
 P0.0000.0700.0170.0070.047
Post-flooding – 2004      
Marking intensityCc0.2060.407**0.364**0.458**-0.348**
 P0.0610.0000.0010.0000.001
Post-flooding – 2005      
Marking intensityCc0.243*0.234*0.345**0.372**-0.373**
 P0.0340.0410.0020.0010.001
Post-flooding – 2006      
Marking intensityCc-0.0850.365**0.354**0.376**-0.141
 P0.5050.0030.0040.0020.265

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Results showed that during all dam construction phases, otters were widespread. This was consistent at both scales of analysis.

At the 25 km2 scale, the percentage of positive cells for otter presence surrounding the flooded area was always similar but the location of negative cells shifted between years. Most absences reflected areas of less suitable conditions for otters. In the post-flooding phase, 2005 had the highest absence of otters, reflecting the extreme drought of that year. Otter distribution was also related to the disturbance caused in the deforestation phase, since cells that included large areas of deforestation became negative. During deforestation, and especially during flooding, otters were forced to look for territories away from the main rivers and streams. However, otters did not settle permanently in these secondary locations, with some of these cells recording otters as absent in 2003.

Results at the 1 km2 scale showed that the first months of deforestation caused a major decrease in otter presence in the flooded area. This corresponds to the period of high disturbance caused by the machinery and manpower used to cut the vegetation from the watercourses. Although otter presence partly recovered immediately after this, the complete lack of vegetation in the watercourses may have hindered otter resettlement. The effect of flooding and the abrupt loss of rivers and streams caused significant impacts in the first 6 months. Otter presence was fairly constant after the water level stabilized, with some signs of recovery in the sites surveyed, although not reaching the level before dam construction.

Dam construction led to changes in otter diet, reflecting major shifts both in abundance and composition of prey communities. The analysis of dietary niche breadth and prey diversity index suggests that otter is an opportunist predator feeding consistently on crayfish and a small number of fish species (Copp and Roche, 2003; Sales-Luís et al., 2007). There was a significant prey switch from a diet dominated by native fish species to one of non-native fish and crustaceans (P. clarkii). Most native species were caught more, or in many cases only, in streams and rivers in 2000. These are species mainly adapted to a lotic environment and when found in lentic systems tend to be at a distance (20–50 m) from the reservoir margins. Otters usually forage in shallower waters near the reservoir margins (Sales-Luís et al., 2007). Reservoirs provide stable lentic habitats in which non-native species often develop thriving populations (Clavero et al., 2004), for they are largely lentic in their native range (Filipe et al., 2004; Ribeiro et al., 2008). A previous study by Ribeiro et al. (2006) in the Alqueva Reservoir found that 95% of the captured fish were non-native. These authors also found a large increase in abundance and distribution of A. melas after flooding. Translocations or the natural spread of individuals from Spanish populations may explain the presence of A. melas in Portugal (Gante and Santos, 2002), which was first confirmed in the Guadiana basin with the capture of a single specimen in 2001. No specimens of A. melas were detected before 2003 in the lower reaches of the Guadiana River basin (Collares-Pereira et al., unpubl. data). As with other introduced species, flooding provided A. melas with highly productive and suitable habitats (Ribeiro et al., 2006). The data also show an increase in the proportion of A. melas after flooding, both in the fish community and as otter prey.

The change in the composition of the fish community is perhaps the most substantial impact of dam construction on otter ecology. In the reservoir, otter became more dependent on fewer species. When these species are non-native, which is the most common situation in a reservoir, this creates several constraints for otters. The non-native species that dominated otter diet in 2006 generally had lower biomass per individual than native species that formed most of their diet in 2000. The mean biomass of fish caught in 2006 was generally lower than in 2000, as the Barbus spp. (508.3 g) were mostly replaced in otter diet by G. holbrooki (0.3 g) and M. salmoides (33.2 g). This means that otters must capture a larger number of individuals to meet the same biomass requirements. Sales-Luís et al. (2007), in a study in a large reservoir in central Portugal, found a preference for greater length classes of L. gibbosus, providing greater energetic profits. In addition, anti-predator body structures (i.e. tough skin and sharp spines) of both L. gibbosus and M. salmoides may impede easy handling and capture by otter (Blanco-Garrido et al., 2008).

Procambarus clarkii was the single most highly consumed species. Basto et al. (2011) showed that the abundance of this crayfish was one of the most important variables positively associated with otter use of small and medium-sized reservoirs. Other studies in Mediterranean areas have also demonstrated the importance of P. clarkii as prey for otters (Magalhães et al., 2002; Clavero et al., 2004; Pedroso and Santos-Reis, 2006). Procambarus clarkii represents an important food resource, especially in stressful periods of severe drought, and is likely to have increased the carrying capacity of the environment for aquatic predators such as otter.

The results also illustrated a change in the availability of otter requirements over time, consistently correlated with otter marking behaviour. Otters usually forage in shallow waters (Kruuk, 2006; Nawab and Hussain, 2012), and in the Alqueva Reservoir prey capture efficiency can be expected to decrease. Although in some cases the flooding of natural deep valleys may increase the amount of shallow foraging areas, this is not the case in the Alqueva Reservoir. During most of the year the rivers and streams flooded by the reservoir were characterized by shallow waters that offered better otter foraging opportunities. Now they are open areas with steep margins where it is more difficult for otters to pursue and trap prey, and where fishing at greater depths is energetically more demanding with longer subsequent recovery on land from the dives (Nolet et al., 1993; Kruuk, 2006). This can affect not only the size of home ranges but also lead to the absence of otters in less adequate foraging areas. This appears to be confirmed by the highly significant correlation between availability of prey and feeding areas and otter marking behaviour in the years of flooding. In the post-flooding phase, this relationship does not exist or is less important, owing to the increase in prey availability.

Otters are less selective with respect to resting sites than for breeding areas, and they may use several in their home range (Jiménez et al., 1998). However, the availability of both became restricted in the Alqueva Reservoir. There are refuges such as large rocks, logs, and other structures and substantial vegetation cover is present mainly in small bays of the reservoir, while most of the banks are dominated by aquatic vegetation that does not offer refuge for otters (e.g. Juncus spp). Marking behaviour in the post-flooding phase indicates that post-impact colonization was significantly related to refuge and cover requirements. Melquist and Hornocker (1983) suggested that the availability of adequate escape cover and shelter were reasons that Canadian otters preferred streams over lakes, reservoirs, and ponds.

Before dam construction, evidence of otter breeding was found in the Guadiana River and adjacent streams (Saavedra, 2002). After dam construction no indication of cub or juvenile presence was found in the reservoir margins. Otter dens, particularly natal ones, are difficult to find without radio-tracking and often there is no evidence of their presence, such as spraints (Moorhouse, 1988). Some studies in the Iberian Peninsula showed that den locations were restricted to rock formations, deep tree roots with holes, bankside vegetation, tangled vegetation carried downstream by floods, and helophytic vegetation systems (Ruiz-Olmo, 1995; Jiménez and Palomo, 1998; Ruiz-Olmo et al., 2005). Flooding caused the disappearance of previously existent valleys of rivers and streams, along with potential natal holts. In the reservoir margins, refuges or potential dens are now restricted to small rocky agglomerates, all quite exposed. Rearing areas where cubs stay after having moved from the breeding areas are also important. Permanent access to stable water levels and undisturbed areas play an important role in determining good rearing areas where small cubs can safely learn to swim (Kruuk, 2006). Ruiz-Olmo et al. (2005) also found female otters rearing small cubs in selected stretches with high food availability, with suitable dens and deeper and calm water. Although there are several undisturbed areas in the Alqueva Reservoir, these do not sustain high concentrations of prey populations, or suitable refuges. Conditions for breeding were positively correlated with marking behaviour in all of the years. The few highly marked cells recorded after dam construction reflects a lack of appropriate areas around the reservoir. Breeding may, therefore, be the otter requirement most affected by dam construction.

Foster-Turley et al. (1990) and Ruiz-Olmo (2001) have argued that the physical characteristics of dams have adversely influenced the distribution of otters in Europe. Alqueva's dam wall is 96 m high and its insertion in the Guadiana river valley effectively cuts off the movements of otters, isolating upstream and downstream populations. After flooding, otter movements were limited within the reservoir and the adjacent watercourses. Smaller streams constitute movement corridors between suitable habitats such as larger rivers, or are themselves suitable habitats. Pedroso (2012) found that streams flowing in and out of reservoirs were one of the main factors influencing otter presence and use of several large reservoirs. A similar conjoint system of reservoir–streams was found in the centre of Portugal (Aguieira Reservoir), with streams identified as important otter refuge areas and even possible (and main) breeding areas (Pedroso et al., 2007; Sales-Luís et al., 2007). Significant correlations were found between the presence of corridor structures and otter marking behaviour in all of the years, including post-flooding. This suggests that corridors are necessary for otter presence and colonization of the reservoir.

Permanent accessibility to fresh water was the only otter requirement that increased after dam construction. This is one of the most important factors influencing the occurrence of otters in dry areas in summer (Beja, 1992; Prenda et al., 2001). With the construction of the dam, a permanent water source is available for the otter all year. However, the disproportionate abundance of this requirement had a negative impact as, after early flooding, marking behaviour was negatively correlated with water accessibility, and, in the post-flooding phase, both marking behaviour and otter presence were negatively correlated with water accessibility.

IMPLICATIONS FOR OTTER CONSERVATION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Suitable habitats for otters are connected aquatic environments with high prey abundance, dense bankside vegetation cover, shelters and breeding dens, and foraging grounds that are easy to find and use (Macdonald and Mason, 1982; Ruiz-Olmo et al., 2005). With the exception of the availability of fresh water, all of these ecological requirements were less available after construction of the Alqueva Dam. Breeding, feeding, and resting areas are critical for a species (Fernández and Palomares, 2000; Kruuk, 2006). Specifically in recently constructed reservoirs, where the vegetation is non-existent and fish populations still have not had the time to colonize, reproduction and breeding may be limited or non-existent.

Nevertheless, these impacts may not be a serious setback for otter conservation in areas where otter populations thrive. The favourable otter population situation found in the Guadiana river basin in southern Portugal (Saavedra, 2002) may have been the key to the otter colonization of less adequate habitat areas like the large Alqueva Reservoir, since the streams surrounding it are well occupied. Yet in other areas where otter populations are present at low density, this scenario may be quite different, with a lower colonization effect and consequently higher impacts of dam construction (loss of occupied area/otter distribution).

Moreover, dams do not offer equal opportunities for otter populations compared with a network of rivers and streams. The Alqueva Dam was built on a stretch of a large river (Guadiana River) at a confluence with two other large streams (Degebe and Alcarrache). These, before impoundment, were able to sustain high-density otter populations. In large reservoirs, the areas of highest use are those near inflowing streams, suggesting a complementary use (Sales-Luís et al., 2007). As a result it is expected that the carrying capacity for otters in the Alqueva Reservoir is now lower than that in the area before its construction.

Dam construction and water development projects create wide-ranging social and environmental consequences, with impacts extending well beyond the initial planning area. Ecologists and environmentalists are often challenged by the complex interaction of forces at work in these environments, making the prediction of overall effects difficult (Maingi and Marsh, 2002). The results emphasize the importance of long-term monitoring studies that include several phases of construction, post-construction and flooding to evaluate species response to impacts, sometimes beyond the mandatory framework of Environmental Impact Assessment. Such studies will enable better planning of mitigation and compensation measures, such as protection of riparian vegetation in streams near to the reservoir, promotion of vegetation in the margins of the reservoir, retention of islands in the reservoir as refuges for otter, construction of inlets and bays to improve foraging efficiency, and the protection of rocky formations and possible rearing areas for otters.

ACKNOWLEDGEMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES

Nuno M. Pedroso was supported by a PhD grant (SFRH/BD/17495/2004) from the Fundação para a Ciência e a Tecnologia. Field data on otter from 2000 to 2003 were collected under the Monitoring of Carnivores in the Alqueva Dam Project, and data for fish species in 2000 were collected under the Monitoring of Fish in the Alqueva Dam Project, both executed by the Centro de Biologia Ambiental – Faculdade de Ciências da Universidade de Lisboa, and funded by EDIA, SA – Enterprise for the Development of the Infrastructures of Alqueva. Special thanks are due to Maria João Collares-Pereira for fish data, and to Maria João Santos, Hugo Matos and Teresa Sales-Luís for assistance in field work on otter in the first years of the monitoring. The authors thank Mafalda P. Basto, Carla Barrinha, Carla Marques and Ana Rita Martins for their assistance in diet analysis. Special thanks are due to Hans Kruuk, Maria João Santos and Filipe Ribeiro for contributions to an earlier draft and to John Bissonette for manuscript revision. Fishing permits for the use of all methods described were issued by the Direcção Geral das Florestas (Ministry of Agriculture, Rural Development and Fisheries) and all animals were handled to ensure their welfare. The authors would also like to thank the editor, Professor Philip J. Boon and two anonymous reviewers for their constructive comments that helped improve the manuscript considerably.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. IMPLICATIONS FOR OTTER CONSERVATION
  8. ACKNOWLEDGEMENTS
  9. REFERENCES
  • Almeida D, Copp GH, Masson L, Miranda R, Murai M, Sayer CD. 2012. Changes in the diet of a recovering Eurasian otter population between the 1970s and 2010. Aquatic Conservation: Marine and Freshwater Ecosystems 22: 2635.
  • Basto MP, Pedroso NM, Mira A, Santos-Reis M. 2011. Use of small and medium-sized water reservoirs by otters in a Mediterranean ecosystem. Animal Biology 61: 7594.
  • Beja PR. 1992. Effects of freshwater availability on the summer distribution of otters Lutra lutra in the southwest coast of Portugal. Ecography 15: 273278.
  • Blanco-Garrido F, Prenda J, Narvaez M. 2008. Eurasian otter (Lutra lutra) diet and prey selection in Mediterranean streams invaded by centrarchid fishes. Biological Invasions 10: 641648.
  • Brooks TM, Mittermeier R, da Fonseca GB, Gerlach J, Hoffmann M, Lamoreux JF, Mittermeier CG, Pilgrim JD, Rodrigues ASL. 2006. Global biodiversity conservation priorities. Science 313: 5861.
  • Cianfrani C, Lay GL, Maiorano L, Satizábal HF, Loy A, Guisan A. 2011. Adapting global conservation strategies to climate change at the European scale: the otter as a flagship species. Biological Conservation 144: 20682080.
  • Clavero M, Hermoso V. 2010. Reservoirs promote the taxonomic homogenization of fish communities within river basins. Biodiversity and Conservation 20: 4157.
  • Clavero M, Blanco-Garrido F, Prenda J. 2004. Fish fauna in Iberian Mediterranean river basins: biodiversity, introduced species and damming impacts. Aquatic Conservation: Marine and Freshwater Ecosystems 14: 575585.
  • Collares-Pereira MJ, Cowx IG, Ribeiro F, Rodrigues JA, Rogado L. 2000. Threats imposed by water resource development schemes on the conservation of endangered fish species in the Guadiana River Basin in Portugal. Fisheries Management and Ecology 7: 167178.
  • Conroy JWH, Watt J, Webb JB. 1993. A guide to the identification of prey remains in otter spraint. Occasional Publication of the Mammal Society No. 16. The Mammal Society: London.
  • Copp GH, Roche K. 2003. Range and diet of Eurasian otters Lutra lutra (L.) in the catchment of the River Lee (south-east England) since re-introduction. Aquatic Conservation: Marine and Freshwater Ecosystems 13: 6576.
  • Feinsinger P, Spers EE, Poole RW. 1981. A simple measure of niche breadth. Ecology 62: 2732.
  • Fernández N, Palomares F. 2000. The selection of breeding dens by the endangered Iberian lynx (Lynx pardinus): implications for its conservation. Biological Conservation 94: 5161.
  • Filipe AF, Marques TA, Tiago P, Ribeiro F, Da Costa LM, Cowx IG, Collares-Pereira MJ. 2004. Selection of priority areas for fish conservation in Guadiana River Basin, Iberian Peninsula. Conservation Biology 18: 189200.
  • Foster-Turley P, Macdonald S, Mason C. 1990. Otters, an action plan for conservation. International Union for the Conservation of Nature: Gland.
  • Gante HF, Santos CD. 2002. First records of the North American catfish Ameiurus melas in Portugal. Journal of Fish Biology 61: 16431646.
  • Giorgi F. 2006. Climate change hot-spots. Geophysical Research Letters 33: 14.
  • Godinho FN, Ferreira MT, Portugal e Castro M. 1998. Fish assemblage composition in relation to environmental gradients in Portuguese reservoirs. Aquatic Living Resources 11: 325334.
  • Guter A, Dolev A, Saltz D, Kronfeld-Schor N. 2008. Using videotaping to validate the use of spraints as an index of Eurasian otter (Lutra lutra) activity. Ecological Indicators 8: 462465.
  • IUCN OSG. 2009. Otters in Environmental Impact Assessments. Recommendations for Environmental Impact Assessments.
  • Jiménez J, Lacomba J. 1991. The influence of water demands on the otter Lutra lutra distribution in Mediterranean Spain. Habitat 6: 249254.
  • Jiménez J, Palomo J. 1998. Utilización de refugios por la nutria en el río Bergantes (Cuenca del Ebro). Galemys 10: 167173.
  • Jiménez J, Ruiz-Olmo J, Pascual A. 1998. Uso del espacio en una población de nutrias en el río Bergantes (C. H. Ebro). Galemys 10: 201208.
  • Kranz A. 1996. Variability and seasonality in sprainting behaviour of otters Lutra lutra on a highland river in Central Europe. Lutra 39: 3344.
  • Krebs CJ. 1989. Ecological Methodology. Harper Collins: London.
  • Kruuk H. 1992. Scent marking by otters (Lutra lutra): signalling the use of resources. Behavioural Ecology 3: 133-140.
  • Kruuk H. 2006. Otters: Ecology, Behaviour, and Conservation. Oxford University Press: Oxford.
  • Macdonald SM. 1983. The status of the otter (Lutra lutra) in the British Isles. Mammal Review 13: 1123.
  • Macdonald SM, Mason CF. 1982. The otter Lutra lutra in central Portugal. Biological Conservation 22: 207215.
  • Macdonald SM, Mason CF. 1994. Status and conservation needs of the otter (Lutra lutra) in the Western Palaeartic. Nature and Environment 67: 1–54. Council of Europe Press: Strasbourg.
  • Magalhães MF, Beja P, Canas C, Collares-Pereira MJ. 2002. Functional heterogeneity of dry-season fish refugia across a Mediterranean catchment: the role of habitat and predation. Freshwater Biology 47: 19191934.
  • Maingi JK, Marsh SE. 2002. Quantifying hydrologic impacts following dam construction along the Tana River, Kenya. Journal of Arid Environments 50: 5379.
  • Melquist WE, Hornocker MG. 1983. Ecology of river otters in west central Idaho. Wildlife Monographs 83: 360.
  • Moorhouse A. 1988. Distribution of holts and their utilisation by the European otter (Lutra lutra L.) in a marine environment. MSc thesis. University of Aberdeen, Aberdeen.
  • Nawab A, Hussain SA. 2012. Factors affecting the occurrence of smooth-coated otter in aquatic systems of the Upper Gangetic Plains, India. Aquatic Conservation: Marine and Freshwater Ecosystems 22: 616625.
  • Nolet BA, Wansink, DH, Kruuk H. 1993. Diving of otters (Lutra lutra) in a marine habitat: use of depths by a single-prey loader. Journal of Animal Ecology 62: 2232.
  • Pedroso NM. 2012. Otters and Dams in Mediterranean Habitats: A Conservation Ecology Approach. PhD thesis, University of Lisbon, Lisbon.
  • Pedroso NM, Santos-Reis M. 2006. Summer diet of Eurasian otters in large dams of South Portugal. Hystrix 17: 117128.
  • Pedroso NM, Sales-Luís T, Santos-Reis M. 2007. Use of Aguieira Dam by Eurasian otters in Central Portugal. Folia Zoologica 56: 365377.
  • Prenda J, Granado-Lorencio C. 1992. Claves de identificacion de Barbus bocagei, Chondrostoma polylepis, Leuciscus pyrenaicus y Cyprinus carpio mediante algunas de sus estruturas oseas. Doñana Acta Vertebrata 19: 2536.
  • Prenda J, Freitas D, Santos-Reis M, Collares-Pereira MJ. 1997. Guía para la identificación de restos óseos pertenecientes a algunos peces comunes en las aguas continentales de la Península Ibérica para el estudio de la dieta de depredadores ictiófagos. Doñana Acta Vertebrata. Doñana Acta Vertebrata 24: 155180.
  • Prenda J, López-Nieves P, Bravo R. 2001. Conservation of otter (Lutra lutra) in a Mediterranean area: the importance of habitat quality and temporal variation in water availability. Aquatic Conservation: Marine and Freshwater Ecosystems 11: 343355.
  • Prenda J, Arenas MP, Freitas D, Santos-Reis M, Collares-Pereira MJ. 2002. Bone length of lberian freshwater fish, as a predictor of length and biomass of prey consumed by piscivores. Limnetica 21: 1524.
  • R Development Core Team. 2011. R: A Language and Environment for Statistical Computing. Vienna, Austria: the R Foundation for Statistical Computing. Available online at http://www.R-project.org.
  • Reuther C, Dolch D, Green R, Jahrl J, Jeffereies DJ, Krekemeyer A, Kucerova M, Madsen A, Romanowski J, Roche K, et al. 2000. Surveying and monitoring distribution and population trends of the Eurasian otter (Lutra lutra). Habitat 12: 1148.
  • Ribeiro F, Chaves ML, Marques TA, Costa LM. 2006. First record of Ameiurus melas (Siluriformes, Ictaluridae) in the Alqueva reservoir, Guadiana basin (Portugal). Cybium 30: 283284.
  • Ribeiro F, Elvira B, Collares-Pereira MJ, Moyle PB. 2008. Life-history traits of non-native fishes in Iberian watersheds across several invasion stages: a first approach. Biological Invasions 10: 89102.
  • Ruiz-Olmo J. 1995. Estudio bionómico sobre la nutria (Lutra lutra L., 1758) en aguas continentales de la Península Ibérica. PhD Thesis. University of Barcelona, Barcelona.
  • Ruiz-Olmo J. 2001. Pla de Conservació de la Llúdriga a Catalunya: Biologia i Conservació. Generalitat de Catalunya, Departament de Medi Ambient, Catalunya. Documents dels Quaderns del Medi Ambient 6: 187.
  • Ruiz-Olmo J, López-Martin JM, Palazón S. 2001. The influence of fish abundance on the otter (Lutra lutra) populations in Iberian Mediterranean habitats. Journal of Zoology 254: 325336.
  • Ruiz-Olmo J, Batet A, Jiménez J, Martínez D. 2005. Habitat selection by female otters with small cubs in freshwater habitats in northeast Spain. Lutra 48: 4556.
  • Saavedra D. 2002. Reintroduction of the Eurasian otter (Lutra lutra) in Muga and Fluvià basins (north-eastern Spain): viability, development, monitoring and trends of the new population. University of Girona, Girona.
  • Sales-Luís T, Pedroso NM, Santos-Reis M. 2007. Prey availability and diet of the Eurasian otter (Lutra lutra) on a large reservoir and associated tributaries. Canadian Journal of Zoology 85: 11251135.
  • Santos MJ, Pedroso NM, Ferreira JP, Matos HM, Salea-Luis T, Pereira I, Baltazar C, Grilo C, Cândido AT, Sousa I, Santos-Reis M. 2008. Assessing dam implementation impact on threatened carnivores: the case of Alqueva in SE Portugal. Environmental Monitoring and Assessment 142: 4764.
  • Sulkava R. 2006. Ecology of otter (Lutra lutra) in central Finland and methods for estimating the densities of populations. University of Joensuu, Joensuu.
  • WCD. 2000. Dams and development: a new framework for decision-making. The report of the World Commission on Dams. Earthscan Publications, London.
  • Wood SN. 2006. Generalized Additive Models: An Introduction with R. Chapman & Hall/CRC: Florida.