A one-century suite of seagrass bed maps: can we trust ancient maps?

Abstract

To assess the dynamics of seagrass beds, it is necessary to be able to compare their present day and former distribution patterns. In most cases, ancient data are available in the form of maps whose reliability must be estimated before comparison. In the study area (near Marseilles, France), a series of maps of the seagrass Posidonia oceanica beds spanning more than one century (dating from 1883, 1897, 1958, 1975, 1979, 1991 and 2000, respectively) are available. In addition, a new map was established, using aerial photography (shallow areas), side scan sonar (deeper areas) and data validation by ground truth. In order to help assess the value of the ancient maps, a Reliability Index (RI) is proposed. The RI, rated from 0 to 50, weighs three parameters: (i) the initial scale of the map (source map) and the working scale (target map); (ii) the method of data acquisition (e.g. dredges, grabs, aerial photography, side scan sonar, scuba diving); and (iii) the method of data georeferencing. Dramatic differences in the distribution of seagrass beds are apparent between the eight maps. In view of the biological features of P. oceanica (e.g. slow rate of change of the meadow limits under natural conditions and the persistence of rhizomes after its death), it is possible that some of these differences may not reflect change over time but may be due to errors in the ancient maps. In contrast, other differences are plausible and validated by field data, e.g. the regression of the lower limit of the meadow. It was thus possible to compare the calculated RI for each map and the plausibility of observed differences as a basis for calibrating the RI.