SciELO - Scientific Electronic Library Online

vol.30 número1  suppl.SympClimate Change and Variability of Pelagic Fish Stocks in the Humboldt Current EcosystemAssessing Vulnerability to Climate Risk: The Case of Small-Scale Fishing in the Gulf of California, Mexico índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Investigaciones marinas

versión On-line ISSN 0717-7178

Investig. mar. v.30 n.1 supl.Symp Valparaíso ago. 2002 

Expert Systems for Fishing Ground
Prediction Models:
A Management Tool in the Humboldt
Ecosystem Affected by ENSO

C. Silva1, K. Nieto1, M.A. Barbieri1,2,
E. Yáñez1

1Escuela de Ciencias del Mar, Universidad Católica de
Valparaíso, Casilla 1020, Valparaíso, Chile;
2Instituto de Fomento Pesquero, Huito 374,
Valparaíso, Chile, E-mail:

El Niño and the Southern Oscillation (ENSO) has severe effects on the fish population of the coastal regions of the Eastern Pacific and specifically in Chile, Perú, Ecuador and Colombia. In 1997-1998, ENSO produced losses to the fisheries sectors that resulted in a worldwide shortage of fishmeal and fish oil (FAO, 1998). Various climatic conditions are involved, including elevated sea surface temperatures (SST), decrease in coastal upwelling due to shifting wind patterns, reduction in primary productivity and a large increase in rainfall. These anomalies result in a decrease of abundance and migration of pelagic fish and sea bird populations during El Niño events.

In the last twenty years, several tools have been developed to study and predict meteorological and climatological changes. Expert systems and early warning systems are used to mitigate the effects of climatic changes. Relations between fishing parameters and environmental variables detected with satellites can be used to develop an expert system to assess probable fishing grounds, which would decrease search times and operational costs of the fishing fleets.

In this study we present a probable fishing grounds (PFG) prediction model for anchovy (Engraulis ringens) in northern Chile (18º21'-24ºS) using historical (1987-96) sea surface temperature (SST), thermal gradients (TGR) and catch per unit effort (CPUE) information. The model also considers the relationships estimated between chlorophyll concentration (Chl) from SeaWiFS and anchovy CPUE during 1999. Later on, we develop an expert system to generate daily PFG charts, taking into account the input environmental data, the relationships established and using the Bayesian theory approach and fuzzy logic. Finally, we validate the application of the PFG model in two scenarios in an ENSO context, one considering the fishing and environmental data of El Niño 1997-98 and the other considering a La Niña and the normal conditions of 2000.

Relationships between SST, TGR, Chl and CPUE

Anchovy is distributed in SST ranging from 16 to 23ºC, the optimum range being between 19 and 20ºC, and in TGR from 0.3 to 3.5ºC/10nm with an optimum between 0.8 and 2.1ºC/10nm. Regarding the Chl data compiled in 1999, these were between 0.2 and 6 mg/m3, and the optimum range varied between 0.3 and 1.3 mg/m3.

The optimum ranges vary monthly, with warmer values during the summer months (January - March). Regarding the TGR, it is possible to observe that in the summer months the most frequent TGR values in fishing zones are higher than in the rest of the year. The high gradient values in fishing zones are related to the upwelling events within the year, with the intensity of the latter being greatest in summer and early autumn. In winter (July - September), the fishing ranges with the lowest optimum SST values are recorded. The TGR begins to decrease towards the end of autumn, reaching its lowest values during the winter months. On the other hand, anchovy shows a coastal distribution related to areas with steep gradients and high chlorophyll levels due to the permanent presence of coastal upwelling.

PFG charts

The application of the PFG model is carried out using daily SST, TGR and Chl images as input variables. Using evidence curves fitted for the corresponding month, the evidence images of each variable are calculated. SST, TGR and Chl evidence images are integrated in one PFG image.

Validation of the PFG model

The validation of the PFG model is made considering two scenarios in an ENSO context, one taking into account the fishing and environmental data of El Niño 1997-98 and the other considering a La Niña and the normal conditions of 2000. During the 1997-2000 period, the anchovy landings (Fig. 1) and the NOAA OI SST anomaly (Fig. 2) in northern Chile had a high variability. The first months (January - July) of the El Niño 1997-98 correspond to high anchovy catches, but this abundance falls abruptly towards the end of 1997, extending this critical situation into 1998. Positive SST anomalies began in February 1997 and grew rapidly until reaching a peak of +2.8ºC during June 1997, this peak in the SST anomalies coincides with the peak in the anchovy landings, which reached 309.254 tons. The anchovy fishery collapsed from August 1997 to December 1998, and successive ban times were imposed. The abundance levels recovered from the beginning of 1999.

Fig. 1 Anchovy landings (t) and ban times during 1997-2000 in northern Chile.

Fig. 2 OI SST anomaly during 1997-2000 in northern Chile (Reynolds et al., 2001).

The PFG model was applied considering the environmental information of El Niño 1997-98. For example, the application of the model taking into account SST, TGR and Chl satellite data from March 17th 1998 indicate that there were only a few fishing grounds for anchovy concentrated in the south area and near the coast between Iquique and Tocopilla (Fig. 3). Anchovy CPUE during warmer months of El Niño 1987 were distributed in the south-centre part of the study area and very close to the coast (Yáñez et al. 1993 and 1995). The satellite images show SST values ranging from 20 to 26ºC, with low thermal gradients and no wind-driven coastal upwelling events that produce phytoplankton blooms. During this date the SST anomaly was +2ºC and the landing levels were very low (Figs. 1 and 2).

Fig. 3 Application of the PFG model using the input SST (ºC), TGR (ºC/nm) and Chl (mg/m3) images from March 17th 1998 and March 15th 2000.

During La Niña and the normal conditions of 2000, the PFG model was successfully implemented and validated, reaching 74% of coincidences among the real fishing grounds and the areas of high probability estimated by the model, for a dataset of 120 observations. The application of the model using the SST, TGR and Chl information from March 15th 2000 generated a PFG image that shows high catch probability in specific areas (Fig. 3). Probable fishing grounds of anchovy are associated with optimal SST, thermal front and high chlorophyll concentrations produced in the coastal upwelling area. The satellite data shows SST varying between 14.5 to 25ºC, and the presence of a thermal gradient with lower temperatures along the coast is visible in the SST and TGR images, associated with coastal upwelling events driven by S and SW winds (Barbieri et al. 1995a and 1997; Yáñez et al. 1994). The Chl image shows an east-west spatial gradient in the area with higher Chl or blooms found in the coastal regions associated with upwelling events.

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons