Afonso-Dias, J. Simões, C. Pinto & P. Sousa, 2002. Use of Satellite GPS data to map effort and landings of the Portuguese crustacean fleet (GeoCrust). EC-DGXIV/STUDY/99/099 - Download (3260 kb)

This project provides the first attempt made in Portugal to use satellite-based vessel monitoring systems (VMS) for scientific research. GPS and vessel speed data provided by MONICAP, the Portuguese VMS, were used, for the first time, to estimate and map fishing effort of the Portuguese Crustacean trawlers operating off the Southwest (Alentejo) and South (Algarve) coasts of Portugal. The estimation and mapping of Landings per Unit Effort (LPUE) was also possible by combining VMS with landing figures.  

The south and south west Crustacean fisheries are multi-specific targeting three commercially important species, the Norway lobster (Nephrops norvegicus-NEP), the Blue and red shrimp (Aristeus antennatus- ARA) and the Deep water rose shrimp (Parapenaeus longirostris- DPS). During 1998 and 1999, only 25 out of 34 licensed trawlers operated monthly. Four of the vessels operating regularly during the two years were freezer trawlers. About 80% of the crustacean trawlers were under VMS control. The GPS data recorded by the 'blue box' installed onboard the crustacean fishing trawlers is transmitted via satellite to Inspecção Geral das Pescas (IGP), the Portuguese monitoring authority responsible for marine surveillance. The VMS dataset used in this study included vessel positions and speed data reported regularly, during 1998 and 1999. As GPS data is geo-referenced with respect to longitude and latitude it is ideal for spatial analyses. 

During this project a dedicated Geographical Information System (GIS) was developed to store, manipulate, analyse and map fishing effort and landings per unit effort (LPUE), in order to improve the data quality available to assess the state of the south Portuguese crustacean stocks. The GIS software created, GeoCrust 1.0 was developed in Visual Basic 6.0. ADODB and MapObjects 2.0 Pro were used to communicate with the DB (in Access 2000) and the maps, respectively and is one of the most important and well-succeeded deliverables of this project. 

The 1998 and 1999 VMS data was coupled with landing records available from Direcção Geral das Pescas e Aquicultura (DGPA), the Portuguese Directorate-General of Fisheries. In the case of non-freezer trawlers, each landing figure was the result of one fishing trip (FT) and corresponded to a sale in the fish market. For the freezer trawlers, since their landings are not sold in the fish market (they sell their 'catches', directly to fishmongers) the landing figures for each fishing trip were pooled catches, registered in logbooks on a daily basis. 

With the data available it was possible to deduce, to some extent, the fishing activity of each vessel, group of vessels and for the whole fleet. This was possible by analysing the variation of the vessels speed reported consecutively and by cross checking that information with landing records. For example, a series of consecutive positions at a speed around 3 knots indicated that the vessel was towing. Conversely, a successive series of positions at higher speeds indicated that the vessel was travelling. Displaying those series of positions in a geo-referenced map, made it possible to visualize the vessel's position. Therefore, it was possible to tell if the vessel had sailed to a port to land, by crossing its position with the landing records. These series of procedures carried out for each vessel, made it possible to define and count the Fishing trips (FT) performed by each vessel. In this study, due to the constrains of the data available it was only possible to define 3044 FT, about 41% of the total fishing trips carried out in 1998 and 1999. The position course of each FT, the duration and corresponding landing were estimated and recorded in the GIS related database (DB) for further analysis. 

Within each FT several trawl hauls (TH) were defined using the same methodology as described to define the FT. A total of 15673 TH were identified. For each TH it was possible to estimate the towing positions, distance, duration and depth range by using the different routines of the software developed, GeoCrust 1.0. This set of information was also stored in the DB. The results showed that, the average TH lasted 4.6 hours, covering 13.8 nautical miles at a speed of 3 knots. 

A grid of 1x1 nautical mile (nm) was used to map different densities corresponding to the number of hours of fishing. The spatial analysis of all TH shaped nine discrete FG: one in Lisbon Region (Lisbon), two in Alentejo (Sines and Arrifana) and six in the Algarve (Sagres, Portimão-Sagres, Olhão-Portimão, Olhão-Tavira, Beirinha and ZEE), covering an area of about 1412 nm2, within the depth range 50 to 1000 m. Although Olhão-Portimão and Portimão-Sagres only represent 29% of the total fishing area, the crustacean fleet spent most time, 65% of the total number of fishing hours, operating in these two FG. Quarterly and annual density maps were presented for the whole fleet for 1998 and 1999. Important differences in the species composition were observed between FG.

The two components of this fleet operated differently. In general, freezer trawlers spent 2.6 days (10 TH per FT) at sea, whereas non-freezer trawlers only spent 1.5 days (4.6 TH per FT) at sea. Different maps were produced to display the different TH performed by vessel or group of vessels for a selected period of time. This way it was possible to visualize the different fishing operations between vessels. 

Mapping the fishing effort and LPUE made it possible to detect areas of overlapping fishing activities between vessels in the same FG and for one vessel between adjacent FG. These facts made it very difficult to separate Fishing effort and LPUE by species. A Hierarchical Cluster Analysis was used to try to determine the FT targeting a particular species, DPS, NEP or ARA. Results showed that the landings could be divided in six groups (FT types), three of them dominated by DPS (~85% of the total weight landed in 1999). Only 1.1 and 8.5% of the FT were classified as targeting NEP and ARA, respectively. The sixth FT type was a mix and less representative group (5%), where the shrimp Palaemon serratus was the most important species. The poor discrimination observed can be largely explained by the higher predominance of DPS in the landings. 

LPUE estimated by species, from exclusive fishing trips showed consistent temporal patterns and differences in magnitude between FG or Regions. LPUE values were, in general higher in 1999 than in 1998, particularly for DPS. Annual mean values of 25kg/h (DPS), 4kg/h (ARA) and 3kg/h (NEP) were observed in 1999, for the total fishing area.