L.G. Ramensky: estimating the ability of plants using their projective cover

This paper continues the series of publications devoted to the scientific biography by L.G. Ramensky and his contribution to the development of phytocenology. Among the various considerations of methods for studying the vegetation, Ramensky was focused on the assessment of plant abundance using their projective cover (later cover). As a geobotanist, Ramensky began working in the first decade of the last century. In the course of this period, the scale of plant abundance according to Drude [32] became widespread. This scale has no quantitative values: there is a mixture of abundance, occurrence, and quantity values as well as the spatial distribution of plants within communities. As a result, data are difficult to compare, and abundance values cannot be processed using mathematical methods. Drude' scale did not satisfy Ramensky, and in 1915 he proposed assessing the abundance of plants as a percentage of their foliage cover on the sample plot, because he believed that this value « has important biological significance, approximately characterizing the relative area of light use of the plant » [9, p. 106]. Fundamental to the sample-plot description according to Ramensky’ method was the summing the cover of each species. This sum could exceed the total foliage cover by layer closeness value, which according to Ramensky [11, p. 16] was «the covering of the foliage of some plant species by other species that are taller» . Layer closeness sometimes is up to 40-50 % in dense grass meadows. Where the sum of that of individual species is 140 % or more, it can be significantly larger in scrubs and forests. In order to summarize the cover of each species, is necessary to assess and record this value in percent cover for each plant. The geobotanists of Ramensky’ school recorded percent cover in the field [20], and where required, the data in percentages were transcribed into ranks. Ramensky also recommended summing the cover values of each physiognomic groups like narrow-leaved (sedges, cereals) and broad-leaved (forbs) herbs, paying most attention to species that were difficult to distinguish. In the past decade, significant progress has taken place in the remote sensing of plants and the evaluation of their top visible cover. However, in many cases the plant overlap («canopy coverage») makes difficult to identify all species and assess the foliage cover especially understorey species. Phytocenology is benefiting from the proliferation of huge international electronic databases (archives) of geobotanical relevés [30, 31, 47]. These databases include relevés in which the abundance is shown in ordinal ranks of different scales (Drude, Hult-Sernander, Norrlin, Braun-Blanquet, and other authors). For the first three scales, algorithms for their transformation into percentages have recently been proposed [44]. To solve a number of issues in phytocenology and ecology, evaluation of the abundance of various plant groups are often necessary. For example, estimating the total cover of a single species across in different layers, or summing the cover of invasive species or summing cover of plants belonging to various functional, taxonomic or growth form groups. However, the ordinal rank values are not possible to be summarized, therefore these should be converted to a continuous scale, namely, a continuous percentage, which is widely used in plant ecology. This is often the midpoint of the ordinal class range. However, converting data into the midpoint of the class ranges assumes that data are distributed symmetrically within each class, yet individual plant species cover, productivity and biomass typically have a right-skew asymmetric pattern. Converting to the midpoint of the class ranges leads to in gross overestimation of summed cover for aggregated properties of plant communities. Therefore, before summing the midpoints of the classes their adjustment is necessary taking into account the life forms of plants [41]. Thus, when combining relevés by numerous authors into large databases, as well as to simplify using the information about individual plant species cover in ecological studies, it is necessary to assess abundance in field not in ranks, but in percentage (like it was done by the Ramensky’ geobotanical school), entering these observations into electronic databases.