Interactions between small cetaceans and the purse-seine fishery in western Portuguese waters

Many fisheries around the world have significant interactions with marine mammals (Northridge, 1991; STECF, 2002). By-catches and incidental mortality of small cetaceans associated with fishing activities are the best-known and best-documented part of these interactions (e.g. Harwood, 1983; Read, 1996). Gillnets and pelagic trawls are responsible for most of the incidental mortality of cetaceans in European waters, although by-catch events have been reported for the majority of fishing gears (Northridge, 1991; STECF, 2002). Another aspect of cetacean-fisheries interactions is the interSCIENTIA MARINA 71(2) June 2007, 405-412, Barcelona (Spain) ISSN: 0214-8358


INTRODUCTION
Many fisheries around the world have significant interactions with marine mammals (Northridge, 1991;STECF, 2002).By-catches and incidental mortality of small cetaceans associated with fishing activities are the best-known and best-documented part of these interactions (e.g.Harwood, 1983;Read, 1996).Gillnets and pelagic trawls are respon sible for most of the incidental mortality of cetaceans in European waters, although by-catch events have been reported for the majority of fishing gears (Northridge, 1991;STECF, 2002).Another aspect of cetacean-fisheries interactions is the inter ference of cetaceans with fishing activities, which can negatively affect fisheries by resulting in loss of bait, damage to fishing gear, decreased catches and increased time spent in fishing operations (Meyer et al., 1992;Wickens et al., 1992;Nitta and Henderson, 1993;Kobayashi and Kawamoto, 1995;Morizur et al., 1999;Silva et al., 2002;STECF, 2002;López et al., 2003).
In Portuguese continental waters, data on strand ing records and fishermen's reports indicate that incidental capture of small cetaceans and other oper ational interactions take place mainly in gillnet fish eries (Sequeira et al., 1997;STECF, 2002).However, fisheries such as purse-seining for small pelagics (Parente, 2000;Stratoudakis and Marçalo, 2002) provide an opportunity for operational inter actions with small cetaceans for two main reasons: (i) purse-seiners target the main prey species of small cetaceans (Northridge, 1991;Pauly et al., 1998;Silva, 1999;Santos et al., 2004;Santos et al., 2007) and (ii) they operate within their distribution area (Sequeira and Ferreira, 1994;Silva and Sequeira, 2003).Landings from the Portuguese purse-seine fishery (around 70000 t per year from 135 vessels, ICES, 2006) are composed mainly of sardine (Sardina pilchardus), an important prey species of the common dolphin (Delphinus delphis) (Silva, 1999;Santos et al., 2004), which is the small cetacean most frequently observed off the Portuguese continental coast (Sequeira et al., 1997;Silva and Sequeira, 2003).Fishing activity takes place within the coastal waters of the continental shelf, mainly below the 100 m depth contour (Parente, 2000;Stratoudakis and Marçalo, 2002).In addition, purse-seiners operate mainly from sunset to sunrise, coinciding with the feeding period of dol phins (Waring et al., 1990;Aguilar, 1997;Crespo et al., 1997) and thus increasing the potential for cetacean-fisheries interactions.
Portuguese purse-seine fishers have repeatedly expressed concern about the adverse effects (for example reducing catches, increasing fishing time) of dolphins on fishing activities, providing some indication that interactions with small cetaceans may be important in this fishery.To address this hypothesis, data were collected on the occurrence of cetaceans during purse-seine fishing activities by observers onboard commercial fishing vessels and through interviews to skippers.Vessels were regis tered in four ports on the western Portuguese coast (Fig. 1), which contribute 29% of the annual purse-seine sardine landings in the total area.Observations were carried out between July and October 2003, coinciding with the period of most intense fishing activity of the purse-seine fleet.Rates of cetacean attendance and interference with the fishing activi ties are presented and the effects of the presence of cetaceans during fishing activities on the fishing effort and catch rate are described.

Onboard observer trips
Data were collected by observers on board com mercial purse-seiners from four ports off the north ern (Figueira da Foz), central (Sesimbra and Setúbal) and southern (Sines) regions of the west Portuguese coast (Fig. 1).Observations were carried out continuously and the observation effort within each phase of the fishing trip (steaming to or from the fishing areas, searching for fish schools and fish ing) was recorded.Observers noted information on the presence, behaviour and by-catch events of cetaceans (hour, position, species, and number of individuals) and types of interference with the fish ing activities.Cetaceans were considered to be pres ent during a fishing event whenever an individual was sighted in the vicinity of the vessel.The behav iour of fish schools was observed with the help of the boat's equipment, sonar and echosounder, but unfortunately neither printing nor electronic record ing devices were available on the vessels to store the observations.Cetaceans were considered to interfere with the fishing activity when the behaviour of fish schools was different from that usually observed during fishing operations.These unusual reactions were generally abrupt changes in fish schooling behaviour and were classified as scattering (shoal "explodes" and individual fish become scattered), sinking (shoals dive towards the sea bottom) or clus tering (shoals become more compact and cohesive).Observers also recorded whether any damaging of the nets occurred.
Sighting rates for each species of cetacean were calculated as the number of groups sighted per hour of observation for each fishing trip.Mean sighting rates where then calculated for each species for each fishing trip and, in the case of the common dolphin, mean sighting rates were also calculated for each area.Sighting rates of the common dolphin were compared among areas using Kruskal-Wallis tests (Zar, 1999).
The total catch biomass and species composition per fishing set (visually estimated by the skipper) was recorded.Fishing effort was calculated as the sum of the time spent searching and fishing per fish ing trip, as both affect catch rates (Hilborn and Walters, 1992).The catch-per-unit effort (CPUE) was calculated as the catch biomass per unit of fish ing effort.Fishing effort and CPUE were compared among fishing trips carried out in the presence and absence of cetaceans using a Kruskal-Wallis test.Only data from the northern region (Figueira da Foz) were analysed as the majority of fishing events attended by cetaceans occurred in this region.

Interview survey
Interviews were carried out by the observers dur ing fishing trips and focused on the skippers' last trip (usually the day before).They were intended to pro vide information comparable to direct observations, although necessarily less complete.Data on the number of fishing sets, catch biomass and species composition of fishing sets, presence and by-catch events of cetaceans (hour, position and species) and interference with the fishing activity were collected for each fishing trip.As skippers were generally not able to distinguish cetacean species, their sightings were classified by the observers according to the description provided.

RESULTS
Purse-seine fishing practices were similar among ports.Vessels usually left simultaneously from the port and the fishing areas rarely extended beyond one degree of latitude to the north or to the south of their home port, although they could vary consider ably between days.Fishing trips took around 9 hours; steaming comprised 7-15% of the trip time, while the remaining time was split between search ing and fishing operations.The net was generally set once or twice per fishing trip at sunset or at sunrise.The main target species varied according to their local abundance and vessel characteristics.Sardine was the target species of vessels from Figueira da Foz and Sines, while the fleets from the central area directed their effort towards horse-mackerel (Trachurus trachurus) and demersal species.Sardine comprised 72% of the total catch biomass within the study period, followed by Spanish-mack erel (Scomber japonicus) and horse-mackerel (around 5% each).

Observer trips
On-board observers monitored 48 fishing trips during which 72 fishing sets yielding 132 t of fish were recorded (Table 1).In a total of 356 h 32 min of observation (53.4% between sunset and sunrise) there were 31 sightings of cetaceans, corresponding to 201 individuals and a mean sighting rate of 0.89 groups/hour (Table 2).42% of the sightings took place during navigation while the remaining ones occurred almost equally during fishing operations and while vessels were searching for schools of fish.
Three cetacean species were recorded: the com mon dolphin, the bottlenose dolphin (Tursiops trun catus) and the harbour porpoise (Phocoena pho coena).The common dolphin was sighted across the study area and showed the highest frequency of occurrence (80.7%) and relative abundance (mean sighting rate = 0.72 groups/hour) of the three species (Table 2; Fig. 1).Sighting rates of the common dol phin were higher in the north, intermediate in the south and lower in the central regions but significant differences were observed only between the north ern and the central region (Q = 1.75, p<0.015) (Table 3).The size of common dolphin groups ranged from 1 to 60 individuals (median = 7.6 indi viduals).Bottlenose dolphins and harbour porpoises were sighted twice (0.09 groups/hour) and once (0.03 groups/hour), respectively (Fig. 1).
During fishing operations, most of the sightings occurred after the net started to be set and all indi viduals identified to the species level were common dolphins.Overall, common dolphins attended 12.5% of the fishing sets, being considerably more frequent during fishing operations off the northern (Figueira da Foz; 31.6%) and southern (Sines; 22.2%) regions than in the central region (Setúbal/Sesimbra; 2.3%) (Table 3).Interactions were observed in 4% (n = 3) of the fishing events: fish schools were observed to sink and scatter (2 cases) and to cluster (1 case).Except for the assem bling effect on fish, which facilitated encircling of the school, interactions were considered to negative ly affect the fishing activity.Observers also regis-  tered some opportunities for fishing events aborted due to the presence of dolphins.In fact, although skippers know about the association between sar dine and dolphins, they believe that dolphins are one of the major predators of sardine and responsible for reducing catches, so they tend to avoid them.The mean CPUE was substantially higher in the southern (Sines, 1362 kg/hour) and northern regions (Figueira da Foz, 1110 kg/hour) than in the ports of the centre (Sesimbra/Setúbal, 154 kg/hour).In the northern region, fishing effort and CPUE were not significantly different between fishing trips attended by small cetaceans (n = 12, 5.4 ± 2.6 h, 1148.5 ± 1145.9 kg/hour) and fishing trips not attended by small cetaceans (n = 4, 5.6 ± 1.6 h, 992.8 ± 480.6 kg/hour) (H = 0.06 and H = 0, both p>0.05).However, catch rates were more variable in the pres ence of small cetaceans (Fig. 2), although no signif icant differences were found (F = 2.62 and F = 5.68, both p>0.05).Onboard observers registered no acci dental captures of cetaceans.

Interview survey
A total of 36 fishing trips were monitored, during which 65 fishing sets yielding 190 t of fish were recorded (Table 1).Skippers reported 18 sightings of small cetaceans (all classified in the category Delphinidae).Half of the sightings occurred during fishing operations, 39% during navigation and 11% during searching for fish schools.Cetaceans attend ed 9 (13.8%) of the total number of the fishing sets, affecting fishing operations on 8 occasions (12.3% of the fishing sets).Skippers reported that cetaceans appeared after the net started to be set and disturbed fishing operations by causing fish schools to sink (4 cases) or scatter (1 case) and by damaging the fish ing gear (1 case).In two cases, in which dolphins clustered the fish school, cetaceans were considered not to affect the fishing activity, whereas the other six interactions were considered to negatively affect the fishing activity.Both the number of sightings per trip and the rates of attendance and interaction with fishing operations were considerably higher off the northern than off the central area (Sesimbra/ Setúbal), supporting the data from onboard observers (Table 4).The limited number of trips sampled off Sines precludes a meaningful compari son among the three regions.Skippers reported three events of accidental catches of cetaceans (eight individuals being caught) off the northern area, corresponding to 4.6% of the fishing events.All individuals were released by low ering the head rope of the net.One of the animals died in situ but it was impossible to ascertain whether the others survived the interaction or whether any injuries occurred that might have caused death after release.

DISCUSSION
Small cetaceans were frequently sighted within the operation area of the Portuguese purse-seine fishery, as shown by data collected by on-board observers.Three cetacean species were recorded: the common dolphin, the bottlenose dolphin and the harbour porpoise.Bottlenose dolphins were sighted in the Sado estuary near Setúbal, where there is a resident population (dos Santos et al., 1998), and harbour porpoises were sighted in the Figueira da Foz area, where there is also evidence of the exis tence of a resident population (Ferreira, 2000).Cetaceans were recorded in 31 out of 48 fishing trips (65%) accompanied by observers.The common dol phin showed the highest sighting rate, 0.72 groups/hour, and accounted for 80% of the sight ings.Common dolphins attended 12.5% of the fish ing sets and, in all cases, were detected after the net started to be set.The presence of common dolphins was observed to negatively affect fishing operations in 3% (n = 3) of the net sets by causing the fish schools to sink and scatter.Although most interac tions were judged to negatively affect the outcome of the fishing activity, neither the CPUE nor the fish ing effort changed significantly in the presence of cetaceans during fishing trips.However, catch rates were more variable in the presence of cetaceans.No by-catches of small cetaceans took place when the observers were on board.The highest interaction rates were observed in the northern and southern regions, where the highest catch rates were also reg istered, suggesting that both cetaceans and fishers are attracted to fishing grounds with high abundance of the prey/target species.
Data collected during interviews with skippers of purse-seine vessels showed that small cetaceans (Delphinidae) were sighted on 50% of fishing trips and attended 13.8% of the net sets, mostly in the northern region, generally supporting observer data.However, skippers reported higher interference rates (12.3%)than on-board observers and, in addition to sinking or dispersing fish schools, small cetaceans were reported to damage the fishing gear.Skippers reported small cetacean by-catches (8 individuals) in 3 of the 65 fishing events.Although seven individu als were released alive it is impossible to know whether they survived the interaction.The reason for the different interference rates is unclear but it is possible that the occurrence of by-catch events is responsible for this difference.The fact that skippers reported by-catch events makes us believe that they did not fear sanctions and that probably no occur rence of cetacean by-catches was concealed.Interview surveys, although not necessarily a reli able source of quantitative data, provide a means of obtaining minimum estimates of fishing sets attend ed by cetaceans, by-catches of cetaceans and cetacean interference with fishing activities.
Purse-seining is mostly dependent on detection and location of fish schools by hydroacoustic instru ments (Misund, 1997 in Fréon andMisund, 1999) and is conducted on fish aggregated in dense shoals normally during night-time (Pitcher, 1983 in Fréon andMisund, 1999).Field studies of fish behaviour are performed under uncontrolled conditions and therefore provide contrasting results.Also, fish schools' behaviour is affected by several external factors, such as sound stimuli from vessels, visual stimuli of the net and the presence of predators (Fréon and Misund, 1999), so it is difficult to identi fy a single cause of changes in their behaviour.However, there is evidence indicating that the swim ming speed of schooling herring and mackerel increases in the presence of an active vessel (Misund, 1994), and that they tend to avoid them by moving horizontally but remain at about the same swimming depth during circling and pursing (Misund, 1993).More abrupt changes in schooling behaviour like splitting, escaping and panic swim ming of the schools may indicate that visual stimuli from the gear or the presence of predators have a stronger effect on their behaviour than the sound stimuli from the vessel (Misund, 1993;Schwarz, 1985in Misund, 1994).On the other hand, the majority of net sets registered by onboard observers occurred between sunset and sunrise, when fish schools seem to show little avoidance of the net.Misund (1993) reported that mature, shoaling her ring in the spawning grounds off western Norway show little avoidance of the net in the winter dark ness but, in contrast, in the North Sea summer day light, the herring schools often avoid the net and escape in about 35% of the sets.In fact, in purseseine fisheries along the Norwegian coast, shoaling herring seldom escape capture at night (Misund, 1990in Misund, 1994).We therefore consider that, most probably, changes in fish schooling behaviour during net sets attended by cetaceans, as reported by observers, were due to fish schools' awareness of the presence of a predator.Predators are known to not only change the microdistribution of their prey, which usually react by increasing their packing den sity and school splitting and avoidance, but can also cause a large decrease in abundance in the area due to long distance flight by the prey (Fréon and Misund, 1999).In acoustic surveys of Sardinella aurita in Venezuela, Fréon and Misund (1999) observed that fish disappeared when dolphins arrived in the area.
By-catches of cetaceans provide ample evidence that interactions with small pelagic fisheries have a high probability.Our results showed that Delphinidae are accidentally caught in the Portuguese purse-seine fishery in accordance with data collected from similar fisheries in neighbouring areas off northwest (Galician waters, López et al., 2003) and southern Spain (southeastern Mediterranean waters, Aguilar, 1991) and across Mediterranean waters (Tudela, 2004).Although cetacean-fishery interactions are known to affect fisheries, few studies have been specifically dedicat ed to the evaluation of cetacean interference with fishing operations.The types of interference observed in this study are comparable to those reported by purse-seine fishermen in Galician waters (northwest Spain), where dolphins were observed to scare and eat fish and to damage the fishing gear (López et al., 2003).The observed rate of cetacean attendance (12.5%) to nets sets is con siderably higher than that reported for pelagic trawl ing on the coast of Scotland (4%), where, addition ally, no interactions were observed (Pierce et al., 2002).Interestingly, both attendance and interaction (3%) rates have comparable magnitude to those reported for the tuna fishery in the Azores (northeast Atlantic, Silva et al., 2002).In this fishery, cetaceans were present in <10% of the fishing events and inter fered with fishing in <5% of the cases by causing the fish schools to sink and feed on live bait.Small pelagic fish may mediate interactions in tuna fish eries since these are the prey species of both tuna and cetaceans.
Most of the observed interactions took place in the northern region of the Portuguese coast (Figueira da Foz) and involved the common dolphin.This pat tern is consistent with the distribution of cetaceans in Portuguese waters, since the common dolphin is the predominant species and shows a decreasing abundance from north to south along the coast (Silva and Sequeira, 2003).It also agrees with the overall distribution of sardine, the most abundant small pelagic fish in Portuguese waters and the target of the purse-seine fishery.Data collected during acoustic surveys between 1997 and 2002 (Marques et al., 2003) show that 34-88% of estimated sardine (in numbers) was distributed in northern Portuguese waters (between 39º and 42ºN).Interactions in this fishery may be partly explained by prey-predator relationship between sardine and the common dol phin, since sardine is an important food resource for the common dolphin across Iberian waters (Silva, 1999;Santos et al., 2004).Silva (1999) reports that 63% of stranded and 96% of accidentally caught common dolphins off the Portuguese coast con tained mainly juvenile sardine in their stomachs, corresponding to 24% and 56% of stomach content biomass.The fact that the northern region is the main recruitment area of sardine, and that young-of the-year (<16 cm) are usually observed in the autumn, when most observations were carried out, may have contributed to higher interaction rates in the area.
A total of 84 fishing trips (on-board observers and interviews) and 137 fishing sets were monitored within this study, as a first attempt to describe cetacean interference with the Portuguese purseseine fishery.Observations covered 4% of the fish ing trips and 5% of the landings of the fleet operat ing at the observed ports during the study period.In spite of the substantial observation effort, this study reports information from a short period that may not be representative of the average situation, so gener alisation of interference or by-catch rates was not attempted.The low frequency of interactions and apparently stable catch rates do not support the hypothesis that small cetaceans are harmful to the Portuguese purse-seine fishery.However, additional data is needed to account for temporal and spatial variability of interactions, and estimates of fishing effort and cetacean abundance are required to assess the importance of by-catches in this fishery.
FIG. 2. -Boxplot of (A) fishing effort and of (B) CPUE of fishing trips without (NC) and with (WC) the presence of small cetaceans.Boxes are drawn with width proportional to the square-roots of the number of observations in the groups.The box stretches from the 25th percentile to the 75th percentile.The line across the box rep resents the median.The ends of the vertical line indicate the minimum and maximum data values.

TABLE 1 .
-Summary information of the monitored fishing trips off the Portuguese west coast, June-October 2003.