La population d’Avicennia germinans du delta du Saloum est-elle relictuelle depuis la dernière période humide ?

Abstract

Across the globe, mangroves are being deforested at a higher rate than tropical forests. The Saloum Delta mangroves in Senegal declined significantly in the 1980s and 1990s but have been regenerating spontaneously since then. This paper tests the hypothesis that the population of Avicennia germinans in the Saloum delta is a relict population that became established during the exceptionally humid years of 1950 to 1967. This article provides a synthesis of two different botanical campaigns conducted at an interval of 15 years. We also conducted an experiment on the flotation of propagules according to water salinity. By 2003, Avicennia germinans reproduction had not been occurring in the mid-tidal zone for several years. Young shoots were only appearing, in the upper tidal zone, when trapped by Sesuvium portulacastrum, so the problem is one of propagule deposition and rooting. In 2018, only 2 individuals, both on the same site, seem to have followed a normal regeneration process, in the mud in the mid-tidal zone, without a propagule trapping process. The recent decrease in salinity seems to be again, and exceptionally, allowing propagules to sink, and thus to germinate as they did during the humid period. A clear linear relationship has been observed between flotation and salinity. In the Gambia and Casamance, where rainfall is higher and salinity lower, reproduction of this species has been normal without being affected by the period of drought. This has allowed us to estimate that annual rainfall or minimum salinity in excess of the 1,000\xa0mm and 20‰ thresholds respectively, after the exceptionally wet period (1950-1967), have since prevented reproduction of the species. If this population, which has been scientifically studied since the 1970s, is indeed a relict population from an abnormally wet period, a whole body of literature describing and explaining the degradation of the Saloum mangroves would be based on erroneous baseline conditions.