Ernesto Regueira Linares¹,
Jorge Calderón V²,
Bonny Bayot A.²

1 Grupo Empresarial del Cultivo de
Camarón, Ministerio de la Industria
Pesquera, 5^ta avenida y 248,
Barlovento, Santa Fé, Playa,
Ciudad de la Habana, Cuba, E-mail:
gccendes@fishnavy.inf.cu
2 Centro Nacional de Acuicultura e
Investigaciones Marinas, CENAIM,
Campus Politécnico
P.O. Box 09-01-4519,
Guayaquil, Ecuador.

A study of spatial and temporal patterns of shrimp production in the Gulf of Guayaquil was carried out. Data of 3,952 production cycles were assembled in a database. The database contains stocking parameters, nutritional inputs and production variables of the shrimp species Litopenaeus vannamei stocked in ponds. To conduct a spatial study, a total of 1,063 production cycles in fifteen different shrimp farms was selected. Principal component analysis (PCA) was used to examine correlation patterns between the above mentioned variables. A component score was calculated for each shrimp farm. With the calculated values, isolines, which represented production trends, could be drawn. For the temporal study, 2,990 production cycles implemented between 1986 and 2000 in four of the fifteen shrimp farms were chosen. The farms are located in Puná island, Taura region, La Seca island and the Sabana Grande region. The production variables total yield, expressed as kg/ha/cycle and survival, expressed as a percentage of all four shrimp farms, registered over a period of 15 years, were used for time series analysis. By means of PCA, the common variability between total yield and survival series at the four shrimp farms were calculated and evaluated. Spectral analysis (SA) was used in order to determine cyclical patterns in the more significant components of the time series. Cross spectrum analysis between significant components of total yield and survival series were compared with sea surface temperature time series in El Niño 1+2 region in order to find relationships between climate (temperature) and shrimp production variability. The results of PCA in the spatial analysis showed that the majority of the variability between the data in each year (1998, 1999 and 2000) included in the study was due to the first three principal components, which respectively accounted for 87.4%, 87% and 82.2% of the overall variability. Through means of spatial analysis, it was shown that the farms located in the southern regions of the Estero Salado and northeastern Guayas River region reached higher production levels. On the contrary, farms located in the central and northern regions of the inner estuary of the Gulf of Guayaquil produced less. Time series analysis showed high variability in the monthly data. On the other hand, cyclical components were observed at the interannual scale. Highest values of final yields and survival coincide with the warm periods during the El Niño phenomenon of ENSO events, whereas the lowest values were registered during the cold periods, coinciding with the phenomenon of La Niña of 1999 and 2000. Through spectral analysis, it was shown that peaks in production occurred at periods of between 4 and 7 years. This demonstrates the high contribution and influence of ENSO events on the production variables analyzed. Through cross spectrum analysis, significant correlations between time series of total yield and temperature of the seawater at the surface in the El Niño 1+2 region were found at 34, 12 and 6 months, which indicates that production variability is influenced by environmental factors such as climatic changes on seasonal and annual scales and ENSO events, independently of the geographic location of the shrimp farms.