29  SPECIES IDENTIFICATION PRACTICES OF COUNTRIES REPORTED LANDINGS OF CHONDRICHTHYAN¹ FISHES IN THE FAO NOMINAL CATCHES AND LANDINGS DATA BASE
R. Shotton

Marine Resources Service, Fisheries Department, FAO
Viale delle Terme di Caracalla, 00100 Rome, Italy

1. INTRODUCTION

A major purpose of identifying the species of fish taken in commercial fishing is usually to obtain the information required to develop fisheries management plans. If a species, or group of species, is not the subject of a specific management plan, then usually there is little demand for accurate species identification of those particular fish in the catch. Because many elasmobranchs are taken as bycatch, or as incidental species, and are not the object of specific management plans, unsurprisingly, few countries assign high priority to collecting accurate data on the species composition of this group in catches. Because there is little, if any, market differentiation between species of many elasmobranch groups, little incentive exists to undertake the often specialized work necessary to identify the catch by species. Further, because of the similar physical appearance of many elasmobranchs species groups and the removal of body parts (e.g. fins) at sea that are necessary to identify individuals by species, such work, if possible, is difficult. Port sampling technicians usually require special training and species identification guides to do such work accurately.

Details of the composition of elasmobranchs (and most types of fishes) in fish landings in the FAO data base are reported across the full range of possible levels of species aggregation (or disaggregation). In the most aggregated form, FAO records may only show reported fish landings as “nei”, i.e. not elsewhere indicated' - all that can be assumed is that the landings consisted of fish. For elasmobranchs, catches may be reported at the order, family, genus or species level. Note should be taken of the term “reported” landings. It is believed that a large amount of fish catches are discarded at the point of capture and are not included in the reported landings. Alternatively, landings of fish are made that are not reported, either because there is no collection of catch statistics (probably the major cause), or because data on landings are deliberately withheld.

The lack of emphasis placed on accurate reporting of the species composition of elasmobranch catches is a great loss given the characteristics of this fish group. Elasmobranchs seem to occur in all ocean habitats, from ankle deep water (as those who have trod on a stingray will attest), to the deepest parts of the oceans, where their presence has been shown from deepwater photographs of deliberately introduced bait, and increasingly by new deepwater fisheries. And, they are also found through out all ranges of latitude, from polar seas to those of the tropics. Few fisheries do not take at least some elasmobranchs as bycatch and in trawl fisheries there is evidence that this has lead (Brander 1981), or is leading to (Casey and Myers 1998, Walker and Hislop 1998) their local extinction. The sensitivity of elasmobranch species composition and population sizes to fishing should make them an ideal indicator of stresses on the icthyofaunal composition of areas exploited by fisheries. However, because of the lack of demand for information for immediate use in their management,the potential of accurate statistics on elamobranch catches has been, to a great part, forgone. But that is not to say that no further benefits from collection of accurate species composition data would exist.

¹ Chondrichthyan, or cartilaginous fishes consist of two living subclasses, of which by the far the most important is the Elasmobranchii - the sharks and rays. For technical accuracy, and because interesting fisheries exist for the other subclass, the Holocephali, or chimaeras, the name for the aggregate group is used in this chapter.

Formal reporting of catches of some groups of elasmobranchs remains essentially non-existent, as is the case for those living in freshwater habitats. Perversely, these elasmobranchs may also be among those that are the most endangered. However, lamentations will not change national and regional priorities regarding collection of fisheries landings (or capture) statistics. One objective of this chapter is, by providing a comparision of the nature of national reporting practices, to encourage fisheries departments to examine their catch reporting practices and assign greater priority to collection and reporting of accurate data fully disaggregated to the species level.

2. CLASSIFICATION OF CHONDRICHTHYAN FISHES

Several authoritative guides exist that describe the systematic relationship of the fishes comprising the chondrichthyan fishes of which the elasmobranchs are the major group. The following classification combines the taxonomic structure of Nelson (1994) with a working classification of cartilaginous fishes for the revised FAO Shark Catalogue and proposed catalogues of batoids and chimaeroids (L.J.V. Compagno, pers. comm., 30 September 1998):

² The superscripts in this list refer to the chapters of this volume. Look at the start of each chapter to see an example of the respective elasmobranch family. Beautiful aren't they?

The FAO data base also refers to a grouping, the “Selachimorpha [Pleurotremata]” for a limited number of countries reporting landings from the NE Atlantic. Nelson (1975, 1984) notes that this term refers to “sharks” though in his 1994, 3rd edition (p. 43) he notes that it refers to “the sharks, and Batidoidimorpha, the rays”. For the purposes of the analyses here, the term “Selachimorpha [Pleurotremata]” is taken to mean the sharks, or more specifically, the Selachians or non-Batoid Elasmobranchs, and as such is equivalent to a superorder, referred to as “order” in the following analysis. The other member of this group appearing in the reported landing data was the Rajiformes.

3. THE FAO REPORTED FISH LANDINGS DATA BASE

3.1 Introduction

The Fishery Information, Data and Statistics Unit (FIDI) of FAO's Fisheries Department maintains a data base, Nominal Catches and Landings that contains the amounts of fish catches landed, and aquaculture output by country of production measured in tonnes. Weights are of the whole animal (i.e. live weight). The data base contains annual values for the period 1950 – 1996 and is organized into 250 countries or territories grouped by 27 major fishing areas (See Figure 1). More than 1000 species, or items, are contained in the list. The global fishery production data set is divided in to capture and aquaculture files and the relevant data are maintained by FIDI for the period 1984– 1996 in two separate data bases.

Figure 1

FAO fishing areas for statistical purposes

3.2 Reported or estimated landings data?

The fisheries statistics are usually obtained from national reporting offices and, wherever possible, verified from other sources. In cases where data are not reported or the data that are provided are considered to be unreliable, FIDI uses data provided by other sources (e.g. regional fisheries bodies, scientific projects, etc.). Where FIDI has no confidence in the reported data and no alternative information is available, as a last resort FIDI estimates landings based on the “best available information”. In the worst case, this involves obtaining estimates by extrapolating using data from other years.

Catch data for the period 1950 – 1969 have been extracted from the Fishery Year Books (this series started in 1947) where available. Because of the absence of data for some countries in early years of the period 1950 – 1969, the data for these countries have been estimated by extrapolating backwards. A subjective method was used based on the “best information available” so care should be taken in drawing inferences based on data for the period prior to 1970. It is possible, when querying the Nominal Catches and Landings data base to identify those data that have been estimated as they are stored with a “F” prefix. However, because of the high incidence of estimated data in the data base prior to 1970, none of these data are so marked and thus for this earlier period it is not possible to identify which data have been estimated and which data are those supplied by national authorities. Because part of the FAO data series for the period 1950 – 1969 has been obtained by extrapolation, and to avoid ambiquities arising from possible subjectivity in data estimates, no data prior to 1966 have been considered in this analysis.

As Table 1 shows, the relative and absolute amounts of estimated catch have declined by almost a third for the period 1976–1996. However, for some countries, all, or most, of the elasmobranch catch may be estimated (Table 2). One hundred percent of the elasmobranch landings were estimated by FIDI for fifteen countries; 62.4% of the Spanish elasmobranch landinds were estimated. Of these countries, elasmobranch landings have been estimated for the last 9 years in the case of Brazil and the last 3 years in the case of Spain.

¹ 1996 is the most recent year for which data in the FAO data base are available.

Table 2

Countries whose elasmobranch landings were estimated and were over 1000t in 1996

Country	Estimated catch (t)
Brazil	25 270
Spain	8 310
Oman	6 241
Morocco	3 305
Kiribati	1 840
Sierra Leone	1 402
Mauritania	1 090
Algeria	1 060

4. AGGREGATION/DISAGGREGATION OF REPORTED LANDINGS DATA

4.1 Global practices

A variety of levels of aggregation of elasmobranch landings data occur in the reported information. Table 3 lists summary data concerning reported categories for the years 1966, 1976, 1986 and 1996 which are the decadal years for which data are examined in this report. For the last data year available, only 8.8% of the landings were identified to a species level, a 31.2% decline from a decade earlier. There has also been a considerable drop in the accumulated catch reported to at least the genus level, from 30.9% in 1966 to 18.4% in 1996. The decline in the last decade has been 30.3%. The general trend has been one of declining dissaggregation of the catches on a global basis of the catches dissaggregated at a more detailed level than that of Family.

4.2 Regional practices in aggregation/dissaggregation of reported landings data

Table 4 shows the results from a similar analysis broken down by region. It is apparent that there is wide variation in the levels of disaggregation amongst the regions. No landings are reported at the species level in 1996 from five regions (Atlantic West Central, Eastern Indian Ocean, North Eastern Pacific, Eastern Central Pacific, Western Central Pacific and the Southwestern Pacific). No landings are reported at the species or Genus level from the Eastern Central Pacific. Six of the regions show a decline in the percentage of landings disaggregated at the lowest level (i.e. species of genus), of which the most serious is that which occurred in the NE Atlantic where the amount of catch disaggregated to a species level declined from 53.4% in 1986 to 34.7% in 1986! Best performance in terms of disaggregation to at least the family level occurred in the Northwest Atlantic (93.2%), then Northeast Atlantic (83.5%), Northeast Pacific (69.2%), Mediterranean and Blank Sea (53.0%) and Southeast Pacific (50.3%). Worst performances were given by the Eastern Central Atlantic (1.0%), Eastern Indian Ocean (1.1%), Western Central Pacific (4.8%) and Northwest Pacific (12.8%).

In the region of the greatest reported landings in 1996, the Western Central Pacific, no clear trend is evident at the species, or genus, level but at the level of family, there has been a slight increase in the amount disaggregated to this level in 1996 relative to 1976 and 1986. The decline in the performance in the NE Atlantic is striking and no region is marked by a clear increase in the degree of disaggregation of the reported landings data.

Table 5 gives an indication of regions by rank where disaggregation of the reported landings is better, and worse. At the species level, the Pacific Northeast ranks first, followed by the NE Atlantic. For cumulative disaggregation to the level of genus, the NE Atlantic ranks first, followed by the SE Pacific then SE Atlantic. When reported landings are cumulated to the Family level, the Northwest Atlantic ranks first, then the Northeast Atlantic followed by the Southwest Pacific. Certain areas clearly perform poorly, e.g. the East Central areas of the Atlantic and Pacific followed by the Atlantic Western Central, Eastern Indian Ocean and West Central Pacific.

5. NATIONAL PRACTICES IN DISAGGREGATION OF REPORTED LANDINGS

Table 6 shows the rankings, in terms of their reported chondrichthyan landings, of the top 20 countries. These data show that of the top three countries - Indonesia, India³ and Pakistan, none reports landings in the FAO data base below the taxonomic level of order, and India provides no details at all of the chondrichthyan catch composition, reporting it all as ‘elasmobranchii’. The sixth ranked country, Japan, is the first to report any landings at a species level, 10.7% of their reported total landings.

The results in Table 6 may indicate that the information about species breakdown that is provided may be of little practical use. For example, in the case of Indonesia, the ‘Order’ category landings that are reported are Rajiformes, but taxonomically, Rajiformes are included within the Elasmobranchii. It is possible that what are reported as Elasmobranchii, are non-Rajiforme Elasmobranchii, i.e. selachians, primarily the sharks. The Philippines follows a similar practice. Likewise, Pakistan reports data disaggregated into only two groups, Rajiformes and only one family,

³ Note that the Cochin Marine Fisheries Research Institute report total “shark” landings of 34 750t for 1996 compared with the elasmobranch report of 71 072t in the FAO data base. It is unknown if skates and rays are included in the CMFRI figure.

the requiem sharks! Venezuela follow the same practice. Clearly, this can not represent what is actually being caught. The data reported by the United States is potentially similarly confusing as fish of the genus Raja could also occur in three of the four catch categories that they report - Rajidae, Rajiformes and Elasmobranchii. Spain also reports landings in only two categories - Raja spp. and Elasmobranchs, again the former may also occur in the latter category. Korea, in 1996, reported chondrichthyan landings only of Rajiformes!

6. NATIONAL PERFORMANCE IN DISAGGREGATING SPECIES LANDINGS

6.1 In terms of total landings

There is no apparent pattern in the quality of the data reported by countries. Table 7 shows those countries that reported that greatest absolute amounts of chondrichthyans disaggregated by (1) species, (2) species or genus, and (3) species or genus or family. In terms of the greatest amount of landings reported as a single species in 1996, ranked first is Sri Lanka with 75.1%. But, FAO data bases reports landings of only species, Carcharhinus falciformis⁴. However, Amarasooriya and Dayaratne (1993), cited in Joseph (1998 - this volume)) report that 46 species of sharks occur in the landings of sharks in Sri Lanka, though some are rather rare. Joseph reports that for the years 1994/1995, C. falciformis represented 60.9% of the catch, the next most common species, Prionace glauca, comprised 12.3% of the sharks landed. Joseph cites Sri Lankan shark landings in 1995 of 26 134t compared to 21 400t of Carcharhinus falciformis reported in the FAO data base, plus 7077t of fishes classified as “Elasmobranchii”. Similar differences occurred in the statistics reported for other years. Argentina ranked second in terms of total chondrichthyans reported disaggregated to a species level - 33.9% of their landings, followed by France (43.1%), UK (44.4%) and Canada (46.0%). If landing reports are considered disaggregated to a species and/or genera level, the ‘results’ provided by Sri Lanka rank it still first, but the second (i.e. first) best performance is given by the UK (87.3% of chondrichthyan landings) then France (58.3%), Argentina (37.3%) and Malaysia (33.7%). When this comparison is extended to species and/or genera and/or family, the greatest amount so classified is by the United States (83.6%), then Pakistan (69.7%), Sri Lanka (75.1%), France (94.1%), Argentina (37.3%) and the U.K. (89.8%). However, the data provided by Pakistan also deserve comment. Only one family of sharks is reported as being landed - the Carcharinidae. Two families of batoids are reported as being landed, Pristidae and Rhinobatidae together with a higher group labelled Rajiformes. It is relevant that Bianchi (1985) reports the presence of 12 families of sharks and 13 families of batoids in Pakistan waters.

⁴ These figures are officially reported by Sri Lanka as “Sharks & Skates/Rays”.

Results are shown disaggregated by species; species or genus; species, genus or family.

Table 8 shows those countries with the worst record in disaggregation of reported landings of chondrichthyans in to groups. Worst performance in terms of amount of landings disaggregated is that given by India, followed by Indonesia, Taiwan Province of China, Japan, Mexico and Brazil. If disaggregation to the level of Subclass or Order is considered, Indonesia moves into top (i.e. worst) place followed by India, again Taiwan Province of China, Mexico, Japan and Brazil. If the consideration is of Subclass, Order or Family, third worst performance is now given by Mexico, then Taiwan Province of China, the United States and Japan.

6.2 Relative performance in disaggregated landings

Of the countries landing significant quantities of chondrichthyans, 100% of the reported landings by India, The Yemen, Uruguay and Costa Rica are aggregated into the subclass level (Elasmobranchii). In the case of countries reporting landings wholly aggregated into the Subclass and Order categories, Indonesia, Taiwan Province of China, The Maldives, Australia, Oman, The Philippines, Senegal and Nigeria join the list. When reported landings are aggregated to the Family level, Mexico, Pakistan, Korea, Venezuela and Morocco become included. At this level, the best performance amoung significant chondrichthyan fishing countries is the U.K. (12.7%) followed by Sri Lanka with 24.9% though it should be noted that the classification to species level is undertaken by FAO staff (M. Perotti, FIDI, pers, comm.)

6.3 Single group only reporting practices

As is apparent, many countries report chondrichthyan landings in only one taxonomic level. But this practice, as Table 9 shows, is not restricted to the higher taxons. Table 9 can be summarized as follows:

Number of countries reporting only at the level of:

Again, it is difficult to conceive, particularly as the higher taxon levels, that the reporting practices represent reality.

7. POSSIBLE CONFUSION REGARDING THE STATUS OF RAJIFORMES

The countries listed in Table 10 reported chondrichthyan landings of only two types, (i) “Elasmobranchii” and (ii) “Rajiformes”, with the exception of Angola which reported Raja spp. and Eritrea that reported Rhinobatidae instead of Rajifomes. Many other countries also reported Elasmobranchii and Rajiformes, but other categories of Elasmobranchii were also reported as landed. Because, taxonomically, the Rajiformes form part of the Subclass Elasmobranchii, it is possible that it was intended that what were reported as Elasmobranchii were in fact the ‘sharks’, or Selachians. However, efforts should be made to ensure that there has not been double counting of Rajiformes, first at the Order level and then at the Subclass level.

8. LANDINGS OF FISHES NOT IDENTIFIED BY SPECIES OR OTHER GROUP

This report has used data specifically identified as pertaining to elasmobranchs. However much data are reported as fishes “not elsewhere indicated” (nei). Because of the low importance often given to elasmobranchs, it seems reasonable to expect that a significant amount of elasmobranchs, in terms of elasmobranch landings that are counted would be included in the nei category, the most aggregated level of catch reporting. Table 11 lists the top twenty countries in terms of landings of fish in the “nei” category taken from 119 countries in the sample. For example, Thailand reports 42.6% of its catch landings as undifferentiated in any way - 1 251 50lt. In terms of the percentage of catch reported as elasmobranchs, it ranks 81^st. In terms of the relative amount of reported landings categorized as nei, it is 9^th; the others fall off the table in terms of absolute amounts of catch reported in this category. Thus, it is reasonable to expect that much of the elasmobranchs taken by Thai fishing vessels is grouped in the nei category. The highest proportion of catch reported as nei, is returned by Hong Kong. This territory (as it was at the time) ranked 88^th in terms of the percentage of the reported catch being elasmobranchs — so again, it would seem safe to conclude that much of the elasmobranchs were included in the nei category.

9. DISCUSSION

Reporting of aggregate levels of fish landings is obviously considered important, as is indicated to the headline treatment afforded to the announcements of successive annual total global landings. However, such information is of little practical use in providing information for preparation of fisheries management plans or the evaluation of fisheries on a fishery or stock basis — what should be the major purpose in collection of catch data. For this, detailed and accurate data are required of the species, genus, or even family level of the different major fishery groups on an appropriate geographical basis.

NEI = reported landings in the ‘not elsewhere indicated’ category.
Elasmo = Reported landings of elasmobranchs.
NEI % = Percentage nei is of total reported landings.
Elasmobranch % = Landings of elasmobranchs as a % of total landings. NEI covers reported landings that are not attributed to any group, family, genus or species.

While the performance of no country in supplying catch data on elasmobranchs approaches what should be possible, some countries are clearly committed to provision of detailed information to FAO regarding chondrichthyan landings. France, the U.K., Argentina and Canada are some examples of creditable reporting practices. However, a satisfactory standard of catch data reporting is clearly not happening in the majority of countries reporting catches of chondrichthyan fishes. Thus, while there are now major concerns about the conservation, species diversity and even potential local extinction of members of this class of fish, the quality of the reported landings int he best global data base is insufficient to confidently monitor or measure changes in taxonomic composition of the catch at an appropriate level.

It is clear that many countries collect and record chondrichthyan data at a disaggregated level from their national fisheries, but for some reason choose to aggregate the data before forwarding the data to the Fisheries Department of FAO. Australia and New Zealand, among others, appear to be examples of such a practice.

A tactical issue for some member countries is what is the ‘best’ practice regarding the level of disaggregation of reported landings that should be recorded. If data managers insist on a level of disaggregated reporting that requires identification of the catch to a level that is beyond the technical ability of the data recorders, or the fishermen completing log forms, e.g. in terms of their ability to correctly identify the species, genus or family, then the data may be incorrectly identified. This will causing subsequent errors by those who use the data on the basis that it is accurate. Clearly, accurate aggregated data are preferably to disaggregated data of unknown reliability. It is clear form the results described in this paper that many countries should audit the performance of those responsible for identification of the landings. And, it goes without saying, countries with non-functioning fisheries statistics systems who estimate the current landings based on the previous years data (which were in turn, derived from those of the previous year ad infinitum) should address the seriousness of reporting data that, at a minimum are useless and, possibly worse, are misleading to those unfamiliar with the actual national reporting practices.

10. ACKNOWLEDGEMENT

I am grateful to Mr. Maurizio Perotti, FAO Fisheries Information, Data and Statistics Unit, for revising and correcting an earlier version of this report and Ms Sara Montanaro, of the same service for accessing the data bases for me.

11. LITERATURE CITED

Bianchi, G. 1995. Field Guide: Commercial Marine and Brackish Water Species of Pakistan. FAO Species Identification Sheets for Fishery Purposes. Proj. UNDP/FAO/PAK/77/033. 20 pp. + 143 plates.

Brander, K. 1981. Disappearance of common skate Raia batis from the Irish Sea. Nature 290(5801):48–49.

Casey, J.M. and R.A. Myers 1998. Near Extinction of Large Widely Distributed Fish. Sci 281:690– 692.

Nelson, J.S. 1994. Fishes of the World. 3rd Ed. John Wiley & sons, Inc. New York, N.Y. 600pp.

Walker, P.A. and J.R.G. Hislop 1998. Sensitive Skates or Resilient Skates? Spatial and Temporal Shifts in Ray Species Composition in the Central and North-western North Sea between 1930 and the Present Day. ICES J. Mar. Sci. 55:392–402.