Tracing the occurrence of the Critically Endangered smalltooth sawfish Pristis pectinata at its southernmost distribution in the Southwest Atlantic Ocean

Sawfishes (family Pristidae) are among the largest marine and euryhaline fishes. Collectively, the 5 extant sawfish species were formerly globally distributed in tropical and subtropical coastal, estuarine, and in some instances, freshwater habitats (Faria et al. 2013, Dulvy et al. 2016). However, with the local extinction and severe depletion of most populations, global sawfish distribution has been vastly reduced; sawfishes are extirpated from 20 countries, and 43 countries have at least one species classified as Possi-

bly Extinct (Dulvy et al. 2016).These population depletions have made sawfishes the most threatened group of chondrichthyans (Dulvy et al. 2014, Fernandez-Carvalho et al. 2014).According to the International Union for the Conservation of Nature's (IUCN) Red List of Threatened Species, 3 sawfish species (Pristis pristis, P. pectinata, and P. zijsron) are currently classified as Critically Endangered (Carlson et al. 2013, Kyne et al. 2013a, Simpfendorfer 2013).The remaining two (Anoxypristis cuspidata and P. clavata) are Endangered (D'Anastasi et al. 2013, Kyne et al. 2013b).All 5 species are also listed in Appendix I of the Convention on International Trade of Endangered Species (CITES) (Carlson et al. 2013).In addition, the smalltooth sawfish P. pectinata has been listed under the US Endangered Species Act since 2003 by the US National Marine Fisheries Service (NMFS 2003).
The historical distribution of P. pectinata includes tropical and subtropical coasts on both sides of the Atlantic Ocean (Faria et al. 2013).However, P. pectinata population reduction is inferred at 95% over a period of 3 generations (Carlson et al. 2013).In addition, this species has suffered an 81% geographical reduction of its original range (Dulvy et al. 2016).As a consequence, the distribution of P. pectinata is now patchy, and it is possible that only remote parts of the Caribbean Sea could host any viable populations outside US waters (NOAA 2013).
Even though this perception of large-scale depletion outside the USA is undisputed, regional and local information about the contemporary occurrence of P. pectinata is mostly scarce outside of Florida.For example, P. pectinata is considered no longer present in the eastern south American countries of Brazil, Uru guay, and Argentina (Charvet & Faria 2014), yet Brazil is the only country in the region for which historical P. pectinata records have recently been reviewed.Historical records are distributed along the Brazilian coast (Faria & Charvet-Almeida 2008), thus suggesting a once large range there.The last known record of P. pectinata for Brazil, however, refers to Thorson's (1974) report for saws traded in northern Brazil in the 1960s and 1970s (Faria & Charvet-Almeida 2008, Charvet & Faria 2014, Manir Feitosa et al. 2017).Unlike P. pectinata records in Brazil, information from the other eastern South American countries, Uruguay and Argentina, is limited.
Besides the aforementioned records for Argentina, no effort has been directed towards investigating other potential sources of sawfish records for Argentina and Uruguay.Such records could clarify its occurrence in these countries.Given this lack of knowledge and the Critically Endangered status of P. pectinata, the aim of this study was to characterize the occurrence of this irregularly occurring species in the Argentinean−Uruguayan Common Fishing Zone (AUCFZ) using quantitative and qualitative data.

Study area
The Southwest Atlantic Ocean (SWA) is a global hotspot of threatened chondrichthyans; 50% of endemic species in this region are threatened with extinction (Dulvy et al. 2014).At the same time, these areas of high chondrichthyan diversity in the SWA are associated with marine front areas of high biological activity (Lucifora et al. 2012).These areas, formed by the confluence of different water masses, have enormous biological productivity as a result of the input of nutrients (Acha et al. 2004).Our study area is located within the SWA and covers the AUCFZ (Fig. 1).It includes the outer region of the La Plata River and its adjacent inner continental shelf, the Argentine and Uruguayan coast (depth < 50 m) and the outer continental shelf (50 to 420 m) in the SWA (34°30' to 39°30' S, 52°00' to 59°00' W), covering 35.848 square nautical miles (nmi 2 ).This area is the meeting point of 3 different water masses: freshwater flow from the La Plata River; the Brazil Current; and the Malvinas/Falkland Current.It is characterized by high spatio-temporal variability in hydrographic parameters and primary productivity (Acha et al. 2004).

Historical records and interviews
First, an exhaustive literature review was carried out to search for old records in local fish catalogues, published journals, theses, and technical reports as well as local newspapers and fishing magazines.Then, the ichthyological collections of museums (Table 1), research institutions, and universities in Argentina and Uru guay were examined in 2015 for historical records of Pristis pectinata, and each individual sawfish was identified according to Lahille (1906), Bigelow &Schroeder (1953), andFaria et al. (2013).
Survey interviews (N = 275) were then conducted with ichthyologists (Argentina, n = 42; Uruguay, n = 9), recreational anglers (Argentina, n = 138; Uruguay, n = 5), artisanal fishermen (Argentina, n = 28; Uru guay, n = 51), and the 2 oldest and largest frozen seafood companies in the AUCFZ.These interviews were conducted from September 2015 to December 2016.A simple questionnaire was used for interviews (see Text S1 in the Supplement at www. int-res.com/ articles/ suppl/ n038 p001 _ supp.pdf).Each interview was conducted in person, by telephone, or through social media (Facebook).In Uruguay, interviews were conducted with artisanal fishermen at La Pa loma.In Argentina, interviews with recreational anglers and artisanal fishermen were focused on the Buenos Aires coast to cover areas close to the AUCFZ (communities 2-6 in Fig. 1).Interviews for the 2 frozen seafood companies were conducted at their headquarters in Mar del Plata (Fig. 1).Currently, 14 frozen seafood companies operate in Uru guay and are responsible for fisheries landing re cords for the ports of Montevideo and La Paloma.
Finally, all fishery landing records belonging to the Dirección Nacional de Recursos Acuáticos (DINARA; The National Directory for Aquatic Resources) from 1977 to 2017 were checked for any sawfish landings.Additional consultations were conducted with regional experts (n = 54), including authors of published chondrichthyan species catalogues for Argentina and Uruguay that include P. pectinata.We assume here that the trawling speed and duration (15 and 30 min) used during the hauls carried out in each country could potentially capture sawfishes.This is based on the fact that these trawls captured other benthic batoids with similar size ranges as P. pectinata.On the other hand, we considered that the likelihood of these surveys capturing sawfish was low.This is because sawfish global declines were al ready in place by the time these surveys started in the 1980s.

Interviews
Interviews conducted with fishermen did reveal new evidence of the presence of sawfish in the region.
However, these new instances lacked associated reference material such as photos or dried rostra and are therefore currently considered anecdotal.
Interviews with Uruguayan artisanal fishermen resulted in 3 oral reports of Pristis pectinata (Table 2); their fishing experience ranged from 25 to 45 yr (mean ± SD = 25.6 ± 3.1 yr).The 3 records were as follows: (1) 1 specimen of about 150 cm TL landed at La Paloma port in 1954; (2) 1 specimen (unknown TL) caught by commercial fishing vessels near Pozo del Fango between 1997 and1998;and (3)  The presence of sawfish in the area was also reported by anglers, who described reported 2 sawfish encounters in Argentina: (1) 1 specimen of less than 100 cm TL from the Mar del Plata port in 1958 and (2) 1 specimen (unknown TL) ~2 km east of the mouth of Mar Chiquita Lagoon in 2012.Angler fishing experience ranged from 2 to 57 yr in Argen tina (23.4 ± 12.6 yr) and from 13 to 30 yr in Uruguay (16.4 ± 7.6 yr) (Table 2).
The remaining sources of information did not produce any new sawfish records.These were: (1) interviews with Uruguayan and Argentinian ichthyologists -their research experience ranged from 15 to 50 yr in both countries (Argentina, 24.1 ± 12.3 yr; Uruguay, 27.2 ± 11.5 yr); (2) historical records from seafood companies -the 2 oldest fro zen seafood companies in Argentina: Moscuzza (96 yr) and Frigorífico del Sud Este (73 yr); and (3) Uruguayan fishery statistics -based on all declared captures from Uruguay landed in the ports of Montevideo and La Paloma since 1977.
No sawfish specimens with locality data associated with Argentina or Uruguay were found in the regional ichthyological collections visited.Therefore, the only P. pectinata dried rostrum found (10914; Table 1) without a locality or collection date cannot be assumed to have been removed from a specimen collected in the region.Sawfish rostra deposited in ichthyological collections of Argentina were mostly acquired from European private collectors in the early 1900s.

DISCUSSION
The present study suggests that the view of sawfishes as extirpated from Uruguayan and Argentinean waters should be revisited.This is based on anecdotal evidence of the relatively recent presence of sawfish in the AUCFZ.
Pristis pectinata is currently still considered possibly extinct along the coast of 15 countries in the east Atlantic and 16 countries in the west Atlantic (Carlson et al. 2013).The absence of any sawfish records in ichthyological collections, bibliographies, fisheries records, and the last 35 yr of research cruise data from the AUCFZ may match an extinction scenario.However, interviews conducted in the present study with fishermen have now challenged this regionally extinct status.
The new reports obtained from fishermen interviews rely on the testimony of 2 professional fishermen and 2 recreational fishermen.Due to the absence of any specimens, photos, or other documented evidence, these records can not be regarded as conclusive evidence of the present-day occurrence of P. pectinata in the AUCFZ.Nevertheless, it is valuable that these fishermen knew what a sawfish looks like.In addition, no other similar species (i.e.sawsharks or swordfish) occur in the study area.Therefore, a misidentification of a sawfish in the region is highly unlikely.
At least 2 of the newly obtained oral records for relatively recent occurrences of sawfish in the AUCFZ appear to be reliable.These were obtained from 2 experienced fishermen from Uruguay.Colleagues from DINARA conducted the interviews (DINARA is an institution that has historically collaborated close ly with local fishermen).In the 1997− 1998 Pozo del Fango case, the fisherman was fishing and witnessed the capture of the specimen.In the 2010 record, the fisherman was fishing in the sea between the Uruguayan− Argentinean border.The specimen was killed and brought on board, but it was lost (either thrown away or stolen).The experience of the fishermen and the detailed accounts provided for the fishing capture event, such as site locality, bottom and vegetation type, and gill net mesh size of 18 cm (Table 2), make this report credible.
It is likely that P. pectinata has always been a vagrant in the region.Vagrancy is defined as those individuals that are/were 'recorded once or sporadically, but […] known not to be native to the area' (IUCN 2013, p. 64).In this sense, the largetooth P. pristis and smalltooth sawfishes have been recorded as vagrant in the Mediterranean Sea (Ferretti 2014) and in southwest Australia (Last & Stevens 2009).Since the first sighting of P. pectinata in 1906 in Argentina, only a few historical confirmed and sporadic records have been documented in the area (Tables 3 & 4).In addition, fishermen are mostly unaware of the existence and/ or decline of this species in the region.Overall, this suggests that a population was never established in Argentina and Uruguay.Thus, old records could refer to vagrant individuals from northern habitats (possibly the red mangroves from southern Brazil; Fig. 1), juveniles searching for new suitable estuarine habitats, or adults following prey as other demersal elasmobranchs do during the summer southward arrival of the Brazilian current into the AUCFZ (several elasmobranch species in the region display this seasonal behavior; Vooren 1997).
The fact that this species has not been recorded in Brazil for so long weakens a hypothesis of presentday migrants coming from Brazil.Nevertheless, the northwestern Atlantic population of P. pectinata in Florida and the Bahamas has been a successful example of sawfish conservation and recovery (Wiley & Simpfendorfer 2010, Guttridge et al. 2015).Therefore, it is possible that this effort will gradually result in the presence of sawfish in areas where they have long been gone.This may help to enable the return of P. pectinata to the eastern coast of South America.
In conclusion, P. pectinata was always a vagrant to the region, and we suggest that the view of this species as extirpated from Uruguay and northern Argentina should be revisited.The present study is the first to focus on reviewing the occurrence of sawfishes within their distribution in the SWA.Despite the limitations of the evidence obtained, these records may serve as a call for further research in the region, leading towards collection of more rigorous evidence.For instance, one potentially useful technique is environmental DNA (eDNA).This technique investigates the presence of a species using samples collected in the environment.It has already been successfully applied to sawfishes (Simpfendorfer et al. 2016), as well as other elasmobranches (Boussarie et al. 2018).
In Argentina, data collected from 73 bottom trawl research cruises in the AUCFZ between 1981 and 2015 by research vessels (RV) operated by Argentina's Instituto Nacional de Investigación y Desarrollo Pesquero (INIDEP; National Institute for Fisheries Research and Development) were analyzed for records of sawfishes.Of these cruises, 33 were conducted in coastal waters (depths < 50 m), and the remaining 40 were in the outer continental shelf (depths between 50 and 250 m).Each coastal trawl station (haul) consisted of a 15 min tow during daylight at a speed of 3−4 knots.The outer stations used 30 min tows at the same speed.During both hauls, a high-opening 'Engel' type bottom trawl was used with a stretched mesh of 100 mm, cod-end, and horizontal mouth opening of 19−22 m.In Uruguay, data from 79 bottom trawl research cruises collected in the AUCFZ between 1984 and 2015 were examined and analyzed.Fifteen of these re search cruises were conducted by the RV 'Cruz del Sur' (1984−1989), while the remaining 64 were conducted by the RV 'Aldebarán' (1991−2015) operated by DINARA.Of these 79 cruises, 32 were coastal and 47 were conducted on the outer continental shelf.Hauls were selected using a stratified random sample design, and each trawl consisted of a 30 min tow during daylight at a speed of 3−4 knots.A high-opening 'Engel' type bottom trawl with a stretched mesh of 80−100 mm, cod-end, and 55−60 m horizontal mouth was used.

Fig. 2 .
Fig. 2. Spatial distribution of the 8490 hauls from 152 research cruises by season conducted by Argentina and Uruguay in the study area between 1981 and 2015 during (a) autumn, (b) winter, (c) spring, and (d) summer

Table 1 .
1 specimen (~80 cm TL) caught by commercial fishing vessels in the Papamoscas fishing zone between Uruguay and Argentina in 2010.Interviews with Argentinian artisanal fishermen yielded no records of P. pectinata; these fishermen's fishing experience ranged from 7 to 73 yr (23.8 ± 15.7 yr).Rostral records of sawfishes from ichthyological collections in Argentina and Uruguay.MLP: Museo de la Plata.(−) No data

Table 2 .
spring: Interviews with fishermen with sightings of sawfishes in the Argentinean−Uruguayan Common Fishing Zone.Oral reports lacked reference material (i.e.no rostra, photo or video).(-) No data available a Sampling date could be earlier; only date of citation was available.b Catalogued as Pristis spp.; probably P. pectinata

Table 3 .
Confirmed records of sawfishes in the Argentinean−Uruguayan Common Fishing Zone 3274; summer: 1797; Tables 4 & S1−S4) with a total covered area of 185.78 nmi 2 yielded no sawfish records in Argentina or Uruguay between 1981 and 2015 (Fig.