A comparative study of biological production in eastern boundary upwelling systems using an artificial neural network

[摘要] Eastern Boundary Upwelling Systems (EBUS) are highly productive oceanregions. Yet, substantial differences in net primary production (NPP) existwithin and between these systems for reasons that are still not fullyunderstood. Here, we explore the leading physical processes and environmentalfactors controlling NPP in EBUS through a comparative study of theCalifornia, Canary, Benguela, and Humboldt Current systems. The NPP driversare identified with the aid of an artificial neural network analysis based onself-organizing-maps (SOM). Our results suggest that in addition to theexpected NPP enhancing effect of stronger equatorward alongshore wind, threefactors have an inhibiting effect: (1) strong eddy activity, (2) narrowcontinental shelf, and (3) deep mixed layer. The co-variability of these 4drivers defines in the context of the SOM a continuum of 100 patterns of NPPregimes in EBUS. These are grouped into 4 distinct classes using aHierarchical Agglomerative Clustering (HAC) method. Our objectiveclassification of EBUS reveals important variations of NPP regimes withineach of the four EBUS, particularly in the Canary and Benguela Currentsystems. Our results show that the Atlantic EBUS are generally moreproductive and more sensitive to upwelling favorable winds because of weakerfactors inhibiting NPP. Perturbations of alongshore winds associated withclimate change may therefore lead to contrasting biological responses in theAtlantic and the Pacific EBUS.

[学科分类] 地球化学与岩石