The need to combine with ancillary data to interpret the results
Our approach is not self-sufficient and relies partly on the availability of expert and/or literature knowledge to interpret the spatial patterns identified through EOF as ecological processes. EOF capture the main modes of variation of the spatio-temporal fields. Combining those results with ancillary data and knowledge about the timing of the main phenological events is needed to interpret the results.
Typically, for sole, some studies investigating the timing and the spatial distribution of egg and larvae distribution have been used to corroborate our interpretation of EOF (Arbault et al., 1986; Petitgas, 1997). For hake, survey data have recorded mature individuals during the spawning season (Alvarez et al., 2004) and catch declarations data have been used to evidence reproduction migration at the level of rough statistical rectangles (Poulard, 2001). They were proven to be consistent with our analysis too.
Still, those data are limited. In both cases, the data were restricted to few samples from very old spring surveys and they have not been updated since this period. Furthermore, in many other cases there might not be any reference data to corroborate the EOF results. One should encourage to increase the quantity of available data by conducting new surveys to obtain direct observations of spawning areas (Fox et al., 2008). However, this requires important investments and depends on the ability of institutions and government to raise funds for these surveys.
Alternatively, expert knowledge of fishermen could be a valuable complement to interpret the main modes of variation (Yochum et al., 2011) when other data are missing. Bezerra et al. (2021) and Silvano et al. (2006) proved the usefulness of fishermen knowledge to determine the temporality of fish spawning and to identify some spawning grounds by crossing the information of aggregation areas provided by several fishermen.