Use of small rich patches by Eurasian otter (Lutra lutra L.) females and cubs during the pre-dispersal period

Transcription

1 J. Zool., Lond. (2) 26, 46 C 2 The Zoological Society of London Printed in the United Kingdom DOI:/S Use of small rich patches by Eurasian otter (Lutra lutra L.) females and cubs during the pre-dispersal period Jordi Ruiz-Olmo *, Antoni Margalida 2 and Antoni Batet Direcció general del Medi Natural, Dr. Roux 8, 8, Barcelona, Spain 2 GEPT. Apdo. 4, 22-El Pont de Suert, Spain (Accepted 28 July 24) Abstract Direct observation of Eurasian otters was used to demonstrate the characteristics and interannual use of small patches of habitats by females and litters in stretches of two rivers in north-east Spain. During censuses and other vigils, 2 otter sightings (involving 26 otters) were obtained, 4 8% of which were observations on females and litters. Thirty-five additional sightings were obtained in stretches used by litters of large cubs (4 individuals). Six small cub (2- to 6-month-old) and large cub (> -month-old) different family units (litters) were detected in both rivers as a whole. Average sizes of litters of small and large cubs were significantly similar, being respectively. (SD =.4, range 2, n = 6) and.2 (SD =.46, range 2, n = ), so mortality of cubs was low at the period before dispersal. Litters were detected in significantly different patches from other otters, and litters of small and large cubs were also found in significantly different river sections. Litters of large cubs used small home ranges and core areas, and selected sections at reservoir tails, reservoirs, wider stretches, rough waters and refuges. Small cubs were found in narrow river stretches with calm waters. Selected patches had sheltered zones with abundant fish. This reduced use of space may result in the low rate of mortality. Key words: Lutra lutra, otter, use of patches, cubs, litters, habitat, conservation INTRODUCTION The Eurasian otter Lutra lutra is a semi-aquatic mustelid, mainly solitary, the cubs being exclusively looked after by the females (Mason & Macdonald, 86; Sandell, 8; Ruiz-Olmo, 4; Kruuk, ). There is little information on the pre-dispersal and dispersal periods (Mason & Macdonald, 86; Da Silva, ; Heggberget, ; Kruuk, ; Durbin, 6). These post-reproductive stages are important in the life of carnivores (Ewer, 8; Bekoff, 8; Waser, 6) and particularly in the life of otters (Mason & Macdonald, 86; Kruuk, ) because it is during these periods that: () the highest mortality rate among the population occurs (Gorman et al., 8; Ruiz-Olmo, Delibes & Zapata, 8); (2) much of the dispersal of the species takes place; () prey capture and location training, feeding and choice of resting sites occurs; (4) interaction with other individuals outside the family core starts (Ewer, 8). Several studies have shown that females and their cubs show nomadic behaviour in large home ranges (Durbin, 6). Females and cubs, however, may select small plots (Ruiz-Olmo, Batet & Jiménez, in press), although they may move *All correspondence to: J. Ruiz-Olmo, Direcció general del Medi Natural, Dr. Roux 8, 8, Barcelona, Spain. ajruiol@gencat.net from one plot to another depending on their needs. This plasticity is found among carnivores, and the possibility of finding the most vital needs in small and predictable patches or the lack of a concentrated distribution of resources may affect population density, movements, and the size and distribution of home ranges, in accordance with the resource dispersion hypothesis (RDH) (Macdonald, 8; Bacon, Ball & Blackwell, ; Kruuk, ). Spatial and temporal distribution of food resource patches, overall food availability, the distribution of other resources (resting sites, water, etc.) and spatial and behavioural relationships with other individuals have large influences on differences in home ranges and movement patterns among populations (Kruuk, ; Ruiz-Olmo, Jiménez, Palazón et al., 2; Ruiz- Olmo, Olmo-Vidal et al., 22; Jiménez & Ruiz-Olmo, in press). Much effort is put into establishing, or modelling otter distribution to improve conservation planning (Barbosa et al., 2). Any place with otter signs may be considered otter habitat. But the use of space by otters (living in linear habitats) is different from that of other carnivore species (Kruuk, ), individuals crossing less suitable stretches in patched habitats or other home ranges, but where otter signs (tracks, spraints) can be easily found. Otters have large home ranges, up to several tens of km

2 4 J. RUIZ-OLMO, A.MARGALIDA AND A. BATET River Noguera Ribagorçana River Noguera Pallaresa LCS (.%) Downstream Spain LCS 2 (2.8%) LCS (6.%) LCS (.2%) LCS 4 (28.6%) N. Ribag. France Andorra N. Pallaresa km River Ebro Fig.. Study area showing both stretches studied, and the large cub sectors (LCS) found. Percentage of large cub litters watched during visual censuses at each river is in parentheses: Noguera Ribagorçana (n = 2) and Noguera Pallaresa (n = ). Squares, hydroelectric power stations; dark grey, dams of reservoirs; light grey, reservoirs; dotted line, maximum water level in the Noguera Ribagorçana. (Kruuk, ; Ruiz-Olmo, 2), and habitat features change within this range. So some parts of the otter habitat must be of less concern to them (e.g. where they do not breed or live for long periods; Ruiz-Olmo, 8; Ruiz- Olmo, Jiménez, Palazón et al., 2). Conservation efforts should be focused on important stretches (breeding, cub rearing, feeding, sheltering). The aim of this work is to understand how and where otter litters are living. A knowledge of their ecology and behaviour at this critical and scarcely documented stage of life may help to facilitate and enhance the management and design of conservation areas. METHODS The study area includes 2 river stretches in the Prepyrenean mountains (Lleida and Huesca, north-eastern Spain) (Fig. ). Both were selected because: (a) there was a relatively high otter density (Ruiz-Olmo, López-Martin & Palazón, 2) avoiding stretches less used by otter litters; (b) otter signs were detected all along the river; (c) data for 2 years was available. The main features of the area are described in Ruiz-Olmo, López-Martin et al. (2). The river Noguera Ribagorçana stretch is. km long, of which.6. km are actual river (depending on the water level of the reservoir) and. km are now a reservoir. The reservoir is located, both in the section studied and some km downstream, in a canyon area, with steep banks (frequently a completely vertical slope). The Noguera Pallaresa stretch is 2. km long, of which.8 km belong to the river, and. km to the last part of a smaller reservoir (with a constant water level). The study was carried out during 2 (in Noguera Pallaresa from 2 to 2) and was based on the direct observation of otters by visual censuses (with 8 22 points or vigils of simultaneous observation; Ruiz- Olmo, ; Ruiz-Olmo, López-Martin et al., 2) and on other occasions ( points or vigils of simultaneous observation). Vigils were carried out from April to early August inclusive, when fish migrate upstream (Ruiz- Olmo, Olmo-Vidal et al., 22) and the otters show some diurnal and crepuscular activity (Ruiz-Olmo, ). Observers were placed simultaneously on the banks, at ± m intervals, during twilight periods (22 and 24 sections m long were defined, respectively), which allowed the otters movements to be followed. Used sections were defined previously during census studies (Ruiz-Olmo,, 8), and fixed at the used GIS system. A total of 4 single vigils ( 4 min each) were carried out in the Noguera Ribagorçana, 4 during 2 otter censuses, each census including sunset and sunrise (Ruiz-Olmo, ), the remaining observations () when a female with large cubs (definition as in Ruiz-Olmo, Olmo-Vidal et al., 22) had been previously detected. A large part of the River Noguera Ribagorçana, widely used by these groups of otters, is within a high area of

3 Use of habitat by otter females and cubs during the pre-dispersal period 4 sharp meanders and next to a forest track allowing full observation of a stretch of almost km by simultaneous observations or short walks on foot. Observations were made with binoculars (8 and 4) at a distance of 22 m. For the Noguera Pallaresa, a total of single vigils were carried out during 6 visual otter censuses. Dispersal usually takes place at the age of 8 6 months (Watt, ; Kruuk, ), so females with cubs were specifically sought in the weeks or months before dispersal. In view of the difficulties in estimating the age of cubs, only references to large cubs have been used (Ruiz- Olmo, Olmo-Vidal et al., 22), basically 2 months old, which cannot be completely distinguished from their mothers because their total length (TL) is over % of mother TL. For some aspects, however, small cubs (2 to 6 months old) were included, who can be observed outside the dens but are smaller than % of mother s TL. Behaviour patterns enabled us to distinguish cubs from mothers. The main areas used by large cubs and their dens were eventually visited on foot or by boat. Some of the main habitat features were recorded in each m section: presence/absence of hydroelectric power station merging flows (HPSTAT), river-channel junction (> 2 m wide) (> 2 JUNCTION), any kind of junction (JUNCTION), < km away from reservoir tail (RESTAIL), reservoir (RESERV), stream < m wide (STR < ), stream > m wide (STR > ), rough waters (> m/s) (ROUGH), calm waters (<. m/s) (CALM), riparian vegetation > 2 m wide (RIPARIAN), pools > 2 m depth (POOL), small islands (ISLAND), complex section (COMPLEX) defined as sections with islands, several arms or meandriform stream, caves (CAVE) and rocky holt systems (ROCKHOLT). Calculations on distances and areas were carried out using GIS Miramon. 4.yc. Watercourse lengths were calculated along the riverbed at the ortophotomap :2, at screen scales between :4 and :. Surface home ranges were calculated in 2 different ways: (a) as the minimum convex polygon (MCP) of extreme observations; (b) as a polygon (riverbanks polygon, RBP) calculated following riverbanks and the reservoir limits in the areas in which otters were observed, at a band m wide on each bank (in which otters were seen on.6% of occasions). In areas of vertical cliffs, these were the limits of the RBP. Food intake was estimated by considering: (a) the average weight of an adult Iberian otter female is between. and 6. kg (Ruiz-Olmo, Delibes et al., 8); (b) a large cub weighs 4 kg (Ruiz-Olmo, Olmo-Vidal et al., 22); (c) the average size of of these large cub groups. According to Kruuk et al. (), the average daily food requirement of an otter is % of their weight. According to Ruiz-Olmo, Jiménez & Margalida (8), otters do not eat bigger fish completely and cubs may leave part of a fish uneaten, so the actual foraging time spent by families could not be estimated, so the theoretical food intake if all fish had been consumed completely was calculated. Carp Cyprinus carpio spawning areas were located in the reservoir. For fish movements, data from Ruiz-Olmo, Probability (a) (b) (c) m stretch Fig. 2. Results of the visual censuses carried out in the studied stretch of the river Noguera Ribagorçana expressed as the probability of observation of otters Lutra lutra for each m stretch. Litters: (a) small cubs; (b) large cubs; (c) other otters without mothers and cubs. Stretches: (upstream) to, narrower sections with calm waters (4 m wide, m /s); to 22, reservoir (level ranging across the years from 4 to 22) and the wider section with rough waters (2 8 m wide, m /s). LCS, sections 2 2. Olmo-Vidal et al. (22) obtained at the study area were used. Finally, during the studies carried out between 84 and 2, all our sightings and information from tracks (see Ruiz-Olmo, Olmo-Vidal et al., 22) and captured litters on large otters in Catalonia and Eastern Aragón (Cinca, Guadalope and Matarranya basins) were collated to confirm the generality of our results. Statistics A multivariate approach was used to categorize the nature of sections selected by otters, by a factorial analysis (Digby & Kempton, 8). The probability of detection of otter litters within selected sections was tested using the chi-square test (d.f. = ). SPSS version was used for this aim. RESULTS Two hundred and fifty sightings of otters were obtained (corresponding to 2 specimens): 2 sightings during visual censuses and occasional vigils (6 in the Noguera Ribagorçana and 2 in the Noguera Pallaresa) and corresponding to the monitoring of three groups in Noguera Ribagorçana in 2, 22 and 2. Abundance of otters was similar in both rivers (Figs 2 & ), the 2 2 2

4 42 J. RUIZ-OLMO, A.MARGALIDA AND A. BATET (a) 2 2. large cub litters/year (SD =., range 2), or.4 large cub litters/year/km. Considering both rivers as a whole, the average size of small cub litters was. (SD =.4, range 2, n = 6) and for large cub litters was.2 (SD =.46, range 2, n = ), not differing statistically (t =.4, d.f =, P = ). Probability (b) (c) m stretch Fig.. Results of the visual censuses carried out in the studied stretch of the river Noguera Pallaresa expressed as the probability of observation of otters Lutra lutra for each m stretch. Litters: (a) small cubs; (b) large cubs; (c) other otters without mothers and cubs. Stretches: (upstream) to 8, narrower sections with calm waters ( m wide, m /s); 8 to 2, wider stretches with rough waters (2 8 m wide, m /s); 22 to 24, reservoir (constant level). For sections, 4 and, less than 2 vigils were carried out. probability of detection during the 4 min vigils normally ranging between.4 and sightings/vigil. Some sections, however, were less used, both in the Noguera Ribagorçana stretch (sections 6 in Fig. 2, with no otters observed in section ) and in the Noguera Pallaresa stretch (sections 2 2 in Fig. ). On the River Noguera Ribagorçana,.8% of sightings corresponded to females with litters during censuses and simultaneous vigils (26 sightings). Among these last groups, five were of small cub (2- to 6-month-old) litters (2, 4, 2, 22 and 2). Some small cub litters may be undetected at this time of the year because they are still in the den. Some 2 sightings of females with large cubs belonged to nine different family units (, two in 2, 4, 6,, 2, 22 and 2). In 6 of the years, no groups of this kind were observed, with an average of. large cub litters/year (SD =.6, range 2), or.6 large cub litters/year/km. Observations of solitary otters included at least one large cub waiting for its mother, while at least observations were of females that had temporarily left their large cubs. On the River Noguera Pallaresa,.% of sightings were of females with their litters. Among these, only one was a small cub litter (). Seven were of females with large cubs, belonging to five different family units (also two in 2,, and 2). During 6 of the years, no large cub litters were found, resulting in Selection and use of the river sections On the River Noguera Ribagorçana, litters (of both small and large cubs) were detected in significantly different m sections than other otters (χ 2 = 6.62, d.f. = 2, P <.), and litters of large and small cubs used also different stretches (χ 2 = 2., d.f. =, P =.) (Fig. 2). Litters of large cubs were found more often downstream (χ 2 =.6, d.f., P <.), at the reservoir or in the 2 8 m wide river (rough waters, m /s), while small cubs were found more often upstream, using the 4 m wide river (calm waters, m /s). Some 88.% of observations of large cub litters were at two stretches totalling. km (.% of the study area length), differing from the remaining river sections (χ 2 = 2.68, d.f. =, P =.4). This result is complemented by the fact that, during simultaneous censuses, only the 2 litter of large cubs was observed by two or more consecutive counters. Therefore large cubs usually stayed in stretches of < m during an activity period. The mid-point of the range of litters of large cubs was an average distance of.4 km (SD =., range., n = 8) with respect to the mouth of the river in the reservoir; although in % of the cases they were < km away (n = litters of large cubs). The altitude of this midpoint of the range was closely linked with the altitude of water level at the reservoir (which means, the situation of the mouth of the river) (r =.8, P <.). Four of the nine litters of large cubs were always observed at a small section in the area of the reservoir tail (LCS ) (Figs & 2, Table 2), one (2) was observed here in April and one (2) was observed mainly on LCS, but also in its vicinity and in LCS 2. Three of the litters were observed only at a section between an island and the merging of two watercourses (LCS 2) placed.6 km upstream of LCS (one litter being detected in April in LCS ), and another. km away from the previous one (LCS ). LCS 2 was only used by females and large cubs while LCS was underwater and the river mouth was > km away. LCS was used only in 2, when two litters of large cubs were detected. In the River Noguera Pallaresa, there were not enough litters of small cubs to compare. However again (Figs & ), litters of large cubs were detected more often downstream (χ 2 =., d.f. =, P =.) at the 2 8 m wide stretch ( m /s), and the only litter of small cubs was detected upstream at the narrower stretch. Two main large cub sections were detected, LCS 4 at the merging of a power station area with another calm stretch (being used by the 2, and 2 litters of large cubs), and LCS, at the reservoir tail (river mouth) (being used by both 2 litters; two and one cubs).

5 Use of habitat by otter females and cubs during the pre-dispersal period 4 Table. Result of the factorial analysis for the first five factors, and percentage of the variance explained by each axis for litters of large Lutra lutra cubs and habitat variables Variable Factor I Factor II Factor III Factor IV Factor V Litters of large cubs HPSTATION > 2JUNCTIO JUNCTION l8 RESTAIL STR < STR > ROUGH CALM RIPARIAN POOL.8. 6 ISLAND COMPLEX RESERV CAVE ROCKHOLT % variance % accumulated variance The factorial analysis has shown a close relationship between the presence of litters of large cubs and some of the habitat variables (Table ), the correlation matrix showing significant differences for RESTAIL (P =.8), RESERV (P =.), STR > (P =.44), ROUGH (P =.2), and CAVE (P <.). Axes I, II and III accounted for a.% of the variance. Axis I accumulated variables according to the habitat size and low stability, the number of litters of large cubs being related to large (wide and depth) and low stable habitats. Axis II aggregated variable according to refuge, the litters of large cubs being also more related to this. Comparison by means of chi-square (d.f. = ) have shown again that litters of large cubs also were found more often at sections in reservoir tails (P <.), reservoirs ( P =.8), complex sections (P =.), rough waters (P =.), presence of caves (P <.) and rocky holt systems (P =.48), than elsewhere. Large cub litters were found less often at narrow sections (P =.4) and calm waters (P =.). Data have been collected on 2 large litters in several basins of north-eastern Spain,.% being found at the reservoir tail, even though this is a rare habitat. Considering only large cub litters detected in the Pyrenean mountains (n = ), 4.% were found at the reservoir tail. All the remaining were observed at rivers. Such results confirm the generality of importance of the situation found for rivers Noguera Ribagorçana and Noguera Pallaresa. Features of the main sections (LCS ) used by the large cubs LCS. This section is at the reservoir ( m stretches 2 and 2, Figs & 2) and is 42 m long and 6.8 ha. It is a half-peninsula, well protected by m rocky cliffs (at both banks). There is an outstanding area of 4 m 2 with big rocky blocks in which the six female otters had all their resting sites detected at a denning area (dens area, DA). The only areas of carp spawning (May) for the whole area were found, 28 and m away from DA. This area contained the main concentration and path of the migratory fish Ebro s barbel Barbus graellsii and Iberian nase Chondrostoma miegii, fish concentration taking place between April and June and migration between March and June. Downstream, these events took place between September and October for both species. LCS 2. This included stretches 6 (Figs & 2) and was.2 km long and. ha. It was at the reservoir s tail when this is at its highest water level and was the main area in which the two main fish species gather and go upriver. LCS. This was of lesser concern for litters of large cubs. LCS 4. This was in the reservoir tail ( m stretch 2, Figs & ) and was.8 long and. ha. As in the LCS, this was the waiting and starting patch for the fish migrations. LCS. This included m stretches 6 8 (Figs & ) and was. km long and. ha. Observations on litters of large cubs at LCS (River Noguera Ribagorçana) The ease with which otters could be observed at LCS and its surroundings allowed a more accurate analysis; 4% of the vigils conducted while litters of large cubs were present (years, 2, 4, 2, 2 and 2) were positive. Without considering the 2 litter, cubs stayed alone 44% of the time, during which at least one individual was tracked down. If the time in which they were tracked down but not seen (on caves or dens, and the mother was far away) is also taken into account, it can be concluded that cubs were with their mothers only % of the time. However, the large cub of the 2 litter stayed with his mother % of the time. Litters of large cubs of 2 and 22 had similar behaviour, with small estimated home ranges and core areas (Table 2). Cubs moved only km away from DA. Visual follow up of other groups of females and cubs (years, 2 and 4) also showed similar behaviour, being always detected inside the home range of these other two litter groups (even though up to 2 sightings were elsewhere in the studied. km stretch). The 2 large cub and his mother had different behaviour, using a longer stretch of up to.6 km, and 4. ha. A core area of ha was found, however, again coinciding with LCS, with % of detected days for this group using and resting there (n = ). Twelve prey items were observed being captured by mothers and litters of large cubs until 22, of which were captured by adult females (.68 prey items/h). Most items were Ebro s barbels between 2 cm, and one other was a viperine snake Natrix maura. This means that between 68 and 62 g of prey was captured per hour (with an average of 48.2 g captured/h). Cubs only captured

6 44 J. RUIZ-OLMO, A.MARGALIDA AND A. BATET Table 2. Litters of large Lutra lutra cubs and home ranges of mothers determined by direct observation at LCS and its surroundings (River Noguera Ribagorçana) during May and early August (, 2, 4, 2, 22 and 2). MCP, minimum convex polygon; RBP, riverbanks polygon; CA, core area of RBP (minimum area % of time) Litter 2 Litter 22 Litter 2 All (6 groups) ( sightings; h 4 min) ( sightings; h 4 min) ( sightings; 2 h 2 min) (4 sightings; 2 h min) River RBP MCP CA River RBP MCP CA River RBP MCP CA River RBP MCP CA length (km) (ha) (ha) (ha) length (km) (ha) (ha) (ha) length (km) (ha) (ha) (ha) length (km) (ha) (ha) (ha) Cubs Groups a a The size of the home range of groups always coincided with the home range of females, because they always included the cubs home range. one fish ( fish captured/h), a 2 cm barbel (8 2 g captured/h). The 2 litter behaved differently, moving more than in previous years. This group captured only.4 prey items/h, estimated as g captured/h. Given an average weight of one of the large cub litters of.. kg, an average need of.. kg of food/day/group (with a maximum of 2. kg/day in a group with a female and two cubs) is estimated. To capture the daily theoretical food intake for most groups (.48 kg captured/h), it would be necessary to spend between 2. and. h foraging (4. h, for litters of large cubs with two cubs). The five groups of females and large cubs using LCS placed their detected resting sites (n = 2) within DA. DISCUSSION Females and litters used specific stretches, which were different to those being used by other otters. Between litters, those with large cubs (pre-dispersal animals) were found selecting reservoirs (mainly tails, and moving the tail following its position), wider sections, rough waters, and refuges, at rich feeding patches. Those with small cubs were upstream, using calm waters and the narrowest stretches. Kruuk () also found that females (often with litters) used different patches from other otters. Females used sheltered coasts and calm waters, whereas the other otters used exposed coasts and rough waters. Most litters of large cubs used space in a similar way over the years, and all used specific and small patches. Females and large cubs concentrated their activity in small home ranges compared with what is known of the species (Kruuk, ; Durbin, 6; Ruiz-Olmo, 2). They established themselves mainly around the area of reservoir tails, in well-protected areas, rich in shelters and resting sites, and with a good availability of fish (the main food in the study area; Ruiz-Olmo & Palazón, ; Ruiz-Olmo, López-Martin et al., 2; Ruiz-Olmo, Olmo-Vidal et al., 22). Although similar to the data obtained by Kruuk () in the Shetlands with otter families using holts for generations and with relatively small home ranges, differences were interesting compared to the largest home ranges exploited by females with cubs at some British rivers (Durbin, 6), probably owing to differences in food abundance and distribution. Over the past 2 years an important number of observations of female otters with large cubs in Spain has been gathered, a major part of them at the reservoirs, specifically at their tails (this study and own pers. obs.). Differences with other studies may be because of the different availability of resources. In these small core areas (..2 km/6 ha), otters had a safe shelter, an extensive and permanent water body (an important factor in Mediterranean environments; Ruiz-Olmo & Delibes, 8; Ruiz-Olmo, Olmo-Vidal et al., 22) and several patches with abundant prey that could be easily captured. These factors are considered to be important or limiting factors for otters (Kruuk & Carss, 6; López-Martín, Jiménez & Ruiz-Olmo, 8; Prenda, López-Nieves & Bravo, 2; Ruiz-Olmo, Batet et al., in press). So the abundance of resources in specific patches over long periods would allow females and cubs to live in small home ranges during a part of the pre-dispersal time (May August), according to the RDH (Macdonald, 8; Kruuk, ). Recently, Dahle & Swenson (2) have shown that Norwegian brown bear Ursus arctos females with cubs and yearlings also used small home ranges during some periods. This restriction in range size was linked to a reduction in cub mortality (in this case to avoid infanticide by males). Some specific areas (LCS, 2, and 4) attracted females with large cubs, when the level of the water was close to/or at the edge of possible dens (although sometimes they lodged in the highest rocky area and had to cross over m of rocks before reaching water). Our data on food needs and capture success again corroborate the high availability of fish in the patch in normal years, compared to capture rates in other studies (Watt, ; Kruuk, ; Ruiz-Olmo, Jiménez & Margalida, 8). The fish captured were mainly big barbel, a large species and important for cubs in this area of the Iberian Peninsula (Ruiz-Olmo, López-Martin et al., 2; Ruiz-Olmo, Olmo-Vidal et al., 22). The only litter of large cubs having a relatively large home range (in 2) had a lower fishing success and food intake rate. The average size of the litters of small and large cubs was similar, and close to values found by Ruiz-Olmo, Olmo-Vidal et al. (22) (.), indicating a low rate of mortality between the exit from dens and dispersal, supporting the evidence of high-quality habitat. This is the time in which litters of large cubs were in such rich

7 Use of habitat by otter females and cubs during the pre-dispersal period 4 sections. Occupying safe patches with plenty of food, thus reducing the need to move, should minimize mortality (as for Norwegian brown bear litters; Dahle & Swenson, 2). The use of only one or few resting sites in an area over many generations has been described by Kruuk (), mainly because the areas in which good dens are located may be a limiting factor (Beja, 6; Kruuk, ; Yoxon, 2). In the study area, DA was the most diverse for potential dens and also the most used plot for this. Knowledge of these quality areas for raising cubs before dispersal should allow greater success in the management and preservation of otter populations. Rivers are heterogeneous, and not all sections have the same significance and quality for supporting otters. The detection of areas important for litters of small cubs (Ruiz-Olmo, Batet et al., in press) and for large cubs (this study) must be a main focus for otter conservation, and should be included in any conservation action plan. This will direct resources to areas of main concern for otters. Although reservoirs are interesting and selected habitats for the rearing of large cubs, it must be kept in mind that Mediterranean reservoirs are new artificial habitats and that many litters of large cubs must use stretches of rivers and streams. Acknowledgements We are grateful to those who helped us in the field, especially: Victoria Asensio, Josep Bardina, José Manuel Bolado, Leocadio Cruz, Josep Jordana, Sisco Mañas, José Antonio Muñoz, José María López-Martín, Josep María Olmo-Vidal, Miquel Palacín, Santiago Palazón, Maribel Pérez and Just Sorando. Paul Chanin helped us by commenting on the typescript. REFERENCES Bacon, P. J., Ball, F. G. & Blackwell, P. G. (). A model for territory and group formation in a heterogeneus habitat. J. theor. Biol. 48: Barbosa, A. M., Real, R., Oliveras, J. & Vargas, J. M. (2). Otter (Lutra lutra) distribution modeling at two resolution scales suited to conservation planning in the Iberian Peninsula. Biol. Conserv. 4(): 2 8. Beja, P. R. (6). Temporal and spatial patterns of rest-site use by four female otters Lutra lutra along the south-west coast of Portugal. J. Zool. (Lond.) 2: 4. Bekoff, M. (8). Behavioural development of terrestrial carnivores. In Carnivore behaviour, ecology, and evolution: Gittleman, J. L. (Ed.). New York: Comstock Publishing Associates (Cornell University Press). Da Silva, P. K. (). Aspects of behaviour and ecology of Old World otters. A review. Habitat 6: 6. Dahle, B. & Swenson, J. E. (2). Seasonal range size in relation to reproductive strategies in brown bears Ursus arctos. J. Anim. Ecol. 2: Digby, P. G. N. & Kempton, R. A. (8). Multivariate analysis of ecological communities. Population and Community Biology Series. London: Chapman & Hall. Durbin, L. S. (6). Some changes in the habitat use of a freeranging female otter Lutra lutra during breeding. J. Zool. (Lond.) 24: 6 8. Ewer, R. (8). The carnivores. London: Weidenfeld & Nicolson. Gorman, M. L., Kruuk, H., Jones, C., McLaren, G. & Conroy, J. W. H. (8). The demography of European otters Lutra lutra. In Behaviour and ecology of riparian mammals: 4. Dunstone, N. & Gorman, M. (Eds). Cambridge: Cambridge University Press. Heggberget, T. M. (). Reproductive strategy and feeding ecology of the Eurasian otter Lutra lutra. PhD dissertation, University of Trondheim. Jiménez, J. & Ruiz-Olmo, J. (In press). Home ranges and movements of Euroasiatic otter (Lutra lutra), influence of sex and age: a case study in Mediterranean rivers. J. Wildl. Res. Kruuk, H. (). Wild otters. Predation and populations. Oxford: Oxford University Press. Kruuk, H. & Carss, D. N. (6). Costs and benefits of fishing by semi-aquatic carnivore, the otter Lutra lutra. InAquatic predators and their prey: 6. Greenstreet, S. P. R. & Tasker, M. L. (Eds). Oxford: Fishing News Books. Kruuk, H., Carss, D. N., Conroy, J. W. H. & Durbin, L. (). Otter (Lutra lutra L.) numbers and fish productivity in rivers in north-east Scotland. Symp. zool. Soc. Lond. No. 6:. López-Martín, J. M., Jiménez, J. & Ruiz-Olmo, J. (8). Caracterización y uso del hábitat de la Nutria Lutra lutra (Linné, 8) en un río de carácter mediterráneo. Galemys (Spec. No.):. Macdonald, D. W. (8). The ecology of carnivore social behaviour. Nature (Lond.) : 84. Mason, C. F. & Macdonald, S. M. (86). Otters. Ecology and conservation. Cambridge: Cambridge University Press. Prenda, J., López-Nieves, P. & Bravo, R. (2). Conservation of otter Lutra lutra in a Mediterranean area: the importance of habitat quality and temporal variation in water availability. Aquat. Conserv. mar. fresh. Ecosyst. : 4. Ruiz-Olmo, J. (4). Reproducción y observación de grupos de nutria (Lutra lutra L.) en el Norte de España. Misc. Zool. : Ruiz-Olmo, J. (). Estudio bionómico sobre la nutria (Lutra lutra L., 8) en aguas continentales de la Península Ibérica. PhD dissertation, University of Barcelona. Ruiz-Olmo, J. (8). Influence of altitude on the distribution, abundance and ecology of the otter (Lutra lutra). In Behaviour and ecology of riparian mammals: 6. Dunstone, N. & Gorman, M. (Eds). Cambridge: Cambridge University Press. Ruiz-Olmo, J. (2). Pla de Conservació de la Llúdriga a Catalunya: Biologia i conservació (Generalitat de Catalunya, Departament de Medi Ambient). Doc. Q. Medi Ambient 6: 8. Ruiz-Olmo, J., Batet, A. & Jiménez, J. (In press). The habitat selection of females and small cubs of otter (Lutra lutra) in Southwest European Mediterranean freshwater habitats. Lutra. Ruiz-Olmo, J. & Delibes, M. (8). La nutria en España ante el horizonte del año 2. Málaga: Sociedad Española para la Conservación y el Estudio de los Mamíferos. Ruiz-Olmo, J., Delibes, M. & Zapata, S. C. (8). External morphometry, demography and mortality of the otter Lutra lutra (Linneo, 8) in the Iberian Peninsula. Galemys (Spec. No.): 2 2. Ruiz-Olmo, J. & Jiménez, J. (In press). Importance of ponds for the otter Lutra lutra during droughts in the Mediterranean ecosystems: a case study in Bergantes river. Mammalia. Ruiz-Olmo, J., Jiménez, J. & Margalida, A. (8). Capture and consumption of prey of the otter (Lutra lutra) in Mediterranean freshwater habitats of the Iberian Peninsula. Galemys (Spec. No.):

8 46 J. RUIZ-OLMO, A.MARGALIDA AND A. BATET Ruiz-Olmo, J., Jiménez, J., Palazón, S. & López-Martín, J. M. (2). Ecologie et conservation de la loutre (Lutra lutra) et du visond Europe (Mustela lutreola) aux milieux Mediterranèes. In L Etude et la conservation des carnivores: 4 2. Chapron, G. & Moutou, F. (Eds). Paris: SociétéFrançaise pour létude et la Protection des Mammifères. Ruiz-Olmo, J., López-Martín, J. M. & Palazón, S. (2). The influence of fish abundance on the otter (Lutra lutra) populations in Iberian Mediterranean habitats. J. Zool. (Lond.) 24: 2 6. Ruiz-Olmo, J., Olmo-Vidal, J. M., Mañas, F. & Batet, A. (22). Influence of seasonality of resources on the Eurasian otter (Lutra lutra L.) breeding patterns in Mediterranean habitats. Can. J. Zool. 8: Ruiz-Olmo, J. & Palazón, S. (). The diet of the European otter (Lutra lutra L.) in Mediterranean freshwaters habitats. J. Wildl. Res. 2(2): 8. Sandell, M. (8). The mating tactics and spacing patterns of solitary carnivores. In Carnivore behaviour, ecology, and evolution: Gittleman, J. L. (Ed.). New York: Comstock Publishing Associates (Cornell University Press). Waser, P. M. (6). Patterns and consequences of dispersal in gregarious carnivores. In Carnivore behaviour, ecology, and evolution 2: Gittleman, J. L. (Ed.). New York: Comstock Publishing Associates (Cornell University Press). Watt, J. P. (). Ontogeny of hunting behaviour of otters (Lutra lutra L.) in a marine environment. Symp. zool. Soc. Lond. No. 6: 8 4. Yoxon, P. (2). Geology and otters. IUCN Otter Spec. Group. Bull. (2):