ABSTRACT

Benthic diatoms' species richness was analyzed based on materials collected from 1996-2009 at 8 near-shore locations of SW Crimea. A total of 93 stations were sampled on sandy-muddy substrate within 6-48 m deep. In total, 433 diatom species and infraspecific taxa were found by the results of cell calculations in a Goryaev chamber and observations of permanent slides under a microscope. Prognostic estimation of the expected diatom species richness (S_exp) was performed by the application of 4 estimators (Jack-knife-1 & 2, Chao-2 and Karakassis-S_∞). Statistical assessment of the accuracy of the estimators' results and evaluation of optimal ratio between minimal sampling efforts and maximal information about diatom species richness at specific sampling locations were conducted. The estimation accuracy of the S_exp is increased proportionally to sampling efforts. Magnitude of S_exp, resulted by estimator S_∞, displayed the most similar values to the actually observed species number (S_obs). Overestimation of S_obs values (not more than 10-13%) was found under consideration of 12-15 samples or less, and slight underestimation (3-5%) when the number of samples exceeds 40-43. The other estimators gave large overestimated results of the expected species richness (Chao - from 21 to 70% higher than S_obs, Jack-knife – 23-58%), calculated by randomization for increasing the row of all samples. Based on relative error (RE) and squared relative deviation (SRD) the estimators' accuracy, depending on biotopes peculiarities and sampling efforts, was evaluated. As it was shown, all estimators gave overestimated results of the S_exp (1.3 to 1.8 times) under consideration of a few samples (4-6). With the increase of station number up to 15-20, Chao and Jack-knife estimators showed decreases in RE and SRD values. Estimator S_∞ quite precisely evaluates S_exp parameter after the first 7-8 permutated samples. The parameters of generalized log-dependence between number of samples (1 to 93) and ratio (%) of species richness, registered in near-shore water areas of SW Crimea (433 species and infraspecific taxa) were determined. Detection of about 50% of all species registered in SW Crimea on sandy/muddy substrates within 5-45 m depth range required consideration of not less than 10 randomized stations. The detection of 67% of total species richness requires studying of 20 or more stations. The detection of 80% of total species richness requires about 40 stations (assuming the equal probability of occurrence of species in samples. Application of this dependence can be recommended for prognostic evaluation of percent share of diatom species richness in relation to different sampling efforts under exploration of ecologically similar and coastal areas of the Black Sea not studied earlier.