Optimization of an artificial neural network for identifying fishing set positions from VMS data: An example from the Peruvian anchovy purse seine fishery

The spatial behavior of numerous fishing fleets is nowadays well documented thanks to satellite Vessel Monitoring Systems (VMS). Vessel positions are recorded on a frequent and regular basis which opens promising perspectives for improving fishing effort estimation and management. However, no specific information is provided on whether the vessel is fishing or not. To answer that question, existing works on VMS data usually apply simple criteria (e.g. threshold on speed). Those simple criteria generally focus in detecting true positives (a true fishing set detected as a fishing set); conversely, estimation errors are given no attention. For our case study, the Peruvian anchovy fishery, those criteria overestimate the total number of fishing sets by 182%. To overcome this problem an artificial neural network (ANN) approach is presented here. In order to set both the optimal parameterization and use “rules” for this ANN, we perform an extensive sensitivity analysis on the optimization of (1) the internal structure and training algorithm of the ANN and (2) the “rules” used for choosing both the relative size and the composition of the databases (DBs) used for training and inferring with the ANN. The “optimized” ANN greatly improves the estimates of the number and location of fishing events. For our case study, ANN reduces the total estimation error on the number of fishing sets to 1% (in average) and obtains 76% of true positives. This spatially explicit information on effort, provided with error estimation, should greatly reduce misleading interpretations of catch per unit effort and thus significantly improve the adaptive management of fisheries. While fitted on Peruvian anchovy fishery data, this type of neural network approach has wider potential and could be implemented in any fishery relying on both VMS and at-sea observer data. In order to increase the accuracy of the ANN results, we also suggest some criteria for improving sampling design by at-sea observers and VMS data.     

Comparison of three chemical tests to assess seed viability: the seed dispersal system of the Macaronesian endemic plant <Emphasis Type="Italic">Rubia fruticosa</Emphasis> (Rubiaceae) as an example

Summary. Three chemical viability tests were evaluated in the seed dispersal system of Rubia fruticosa, in which three main groups of dispersers participate: reptiles, birds and mammals. Tetrazolium chloride (TTC) and indigo carmine (IC) indicated a lower viability of seeds from droppings of introduced rabbits (Oryctolagus cuniculus) than of those from control plants and the native dispersers, lizards and gulls. In the rabbit seed treatment, significant differences were observed between results obtained with TTC and IC tests. Interpretation of these data, due to the presence of doubtful embryo staining, was more difficult using the IC test. Furthermore, some seeds that were clearly dead had been underestimated. In contrast with results obtained from the two staining methods, the EC test did not confirm that viability of control seeds and those seeds consumed by native dispersers were clearly higher than in seeds ingested by O. cuniculus. Further, compared to the other two tests, the EC method requires more careful handling of the embryo during the extraction process to avoid errors in viability estimation, since this method measures concentration of electrolytes that are released through cellular membranes. Thus, TTC was the most reliable test to assess seed viability in the seed dispersal system of R. fruticosa, and these results agree with those obtained in previous germination experiments made on the same set of seeds given the same treatments.     

Density surface modelling with line transect sampling as a tool for abundance estimation of marine benthic species: the <Emphasis Type="Italic">Pinna nobilis</Emphasis> example in a marine lake

Density estimation of marine benthic fauna is most often conducted with fishery surveys using dredges or trawls. These estimates are often unreliable due to low and variable efficiency and are inappropriate when dealing with rare or endangered species. In the marine Lake Vouliagmeni, a density surface modelling (DSM) approach using survey data from line transects, integrated with a Geographic Information System (GIS), was used to estimate the population density of the endangered fan mussel Pinna nobilis. This is the first time that such an approach has been applied for a marine benthic species. DSM was beneficial in relation to traditional distance sampling. Apart from providing a more precise total abundance estimate, it related the density of the species to spatial covariates of interest, gave a depiction of the species dispersion in the study area, and provided abundance estimates in any sub-region of the study area. In Lake Vouliagmeni, a marked zonation of P. nobilis distribution was revealed, with the species being restricted in the shallow peripheral zone at depths <22 m. Two density peaks were observed, a major peak at depths between 12 and 13 m and a secondary peak at ∼4 m. A main hotspot of high density was also observed in the northeastern part of the lake. Total abundance of the species was estimated to be 6,770 individuals with a 95% confidence interval of 5,460–8,393 individuals.