Premium
Length–volume relationship of lake phytoplankton
Author(s) -
Mittler Udo,
Blasius Bernd,
Gaedke Ursula,
Ryabov Alexey B.
Publication year - 2019
Publication title -
limnology and oceanography: methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.898
H-Index - 72
ISSN - 1541-5856
DOI - 10.1002/lom3.10296
Subject(s) - phytoplankton , scaling , range (aeronautics) , exponent , linear regression , power law , volume (thermodynamics) , environmental science , cell size , scale (ratio) , regression , mathematics , statistics , biology , ecology , physics , geometry , thermodynamics , materials science , nutrient , linguistics , philosophy , quantum mechanics , composite material , microbiology and biotechnology
The shapes of phytoplankton units (unicellular organisms and colonies) are extremely diverse, and no unique relationship exists between their volume, V , and longest linear dimension, L . However, an approximate scaling between these parameters can be found because the shape variations within each size class are constrained by cell physiology, grazing pressure, and optimality of resource acquisition. To determine this scaling and to test for its seasonal and interannual variation under changing environmental conditions, we performed weighted regression analysis of time‐dependent length–volume relations of the phytoplankton community in large deep Lake Constance from 1979 to 1999. We show that despite a large variability in species composition, the V ( L ) relationship can be approximated as a power law, V ~ L α , with a scaling exponent α = 3 for small cells ( L < 25 μ m) and α = 1.7 if the fitting is performed over the entire length range, including individual cells and colonies. The best description is provided by a transitional power function describing a regime change from a scaling exponent of 3 for small cells to an exponent of 0.4 in the range of large phytoplankton. Testing different weighted fitting approaches we show that remarkably the best prediction of the total community biovolume from measurements of L and cell density is obtained when the regression is weighted with the squares of species abundances. Our approach should also be applicable to other systems and allows converting phytoplankton length distributions (e.g., obtained with automatic monitoring such as flow cytometry) into distributions of biovolume and biovolume‐related phytoplankton traits.