Premium
Cladocera assemblages from reservoirs in Sri Lanka and their relationship to measured limnological variables
Author(s) -
Yatigammana S. K.,
Cumming B. F.
Publication year - 2017
Publication title -
lakes and reservoirs: research and management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.296
H-Index - 39
eISSN - 1440-1770
pISSN - 1320-5331
DOI - 10.1111/lre.12183
Subject(s) - cladocera , subfossil , canonical correspondence analysis , bioindicator , ecology , taxon , environmental science , species richness , biology , zooplankton , paleontology , holocene
Abstract The potential of using surface‐sediment assemblages of Cladocera as bioindicators for reservoirs in Sri Lanka was assessed for their subfossil remains, along with contemporary physical and chemical measurements from each reservoir. The reservoirs span five climatic regions, from extremely arid environments to tropical montane forests, as well as three orders of magnitude changes in many physical and chemical variables. In total, although the remains of 39 Cladocera taxa from 21 genera were identified, only 31 taxa from 37 sites were present at sufficiently high abundances to assess their relation to measured environmental variables. Canonical correspondence analysis ( CCA ) identified surface area, maximum depth and chloride as the three most important measured environmental variables that could account for the variation in the cladoceran assemblages. Taxa such as Chydorus sphaericus , Alona aff. verrucosa and Leydigia acanthocercoides were more abundant in generally deeper, larger reservoirs, whereas Alonella excisa , Euryalona orientalis , Notoalona globulosa and Chydorus eurynotus were more abundant in shallow smaller reservoirs. Although there was a strong separation between climatic zones in terms of factors related to specific conductance, this factor only appears marginally important in separating cladoceran assemblages. Quantitative inference models developed to assess the strength of inferring environmental variables using partial least squares regression and calibration were all relatively weak, with jackknifed coefficient of determination values of 0.40, 0.28 and 0.27 for surface area, maximum depth and chloride, respectively. These results, in conjunction with large differences in eigenvalues between constrained and unconstrained ordinations, suggest that unmeasured environmental variables are also important in structuring cladoceran assemblages.