Doing more with less? Balancing sampling resolution and effort in measurements of protistan growth and grazing‐rates

Abstract

The dilution‐method has been key in establishing the role of protistan‐grazing in marine foodwebs. Yet its laborious application limits the sampling‐resolution achieved. We assessed the reliability of an abbreviated method known as the 2‐point by analyzing 77 dilution‐experiments performed using 4–5 dilutions in diverse biotic and abiotic conditions. Our aim was to inform practitioners on how experimental design and nonlinear feeding behaviors affect the accuracy of 2‐point rate‐estimates. We found good agreement between rate‐estimates of both phytoplankton growth (μ) and grazer‐induced mortality (g) from either method, even though the comparison included experiments with nonlinear feeding‐responses. The accuracy of 2‐point estimates was similar to the inherent standard deviation of dilution‐series estimates (± ∼ 0.1 d−1). Nonlinear feeding‐responses did not alter overall rate‐estimates, negating the need for more than two dilution‐levels. Decreasing dilution resulting in an increase in biomass in the dilute treatment from 10% to 40% increased the median difference between 2‐point and dilution‐series estimates 3‐fold and increased 2‐point estimates' variance 2‐fold, both for μ and g. Recognition of these biomass vs. accuracy tradeoffs enables practitioners to choose whether to procure more biomass at the expense of constraining estimate variance. Using duplicate bottles at each dilution level doubled average accuracy of 2‐point estimates. The reduction in effort and water‐needs afforded by the 2‐point design facilitates acquisition of higher‐resolution data of predation‐rates across seasons, latitudes, and in response to diverse environmental conditions in the ocean, which is critically needed to decipher abiotic and biotic drivers of protistan‐grazing and to parameterize protistan herbivory in global biogeochemical models.

Seasonal similarity in rates of protistan herbivory in fjords along the Western Antarctic Peninsula

“…To verify that this critical assumption was met, the linear regression of k vs. the dilution factor for each DS experiment was tested for deviations from linearity using ANOVA on the residuals of the regression at an alpha level of 0.05 (Zar ). We determined μ and g according to Morison and Menden‐Deuer () as follows: If no deviation from linearity was detected, we tested the null hypothesis that the regression slope = 0. If the regression slope was significantly different from 0, the rates were estimated from the linear regression coefficients ( g from the negative slope and μ from the y ‐intercept) following Landry and Hassett ().…”

Methodological review and meta‐analysis of dilution assays for estimates of virus‐ and grazer‐mediated phytoplankton mortality

“…Technically, only two dilution levels are required to estimate growth and mortality rates, but most experiments incorporate replicate bottles at four to six dilution levels (Landry 1993). The use of a "2-point method" has been promoted recently in the standard dilution assay literature as a reliable, accurate and easier approach for estimating phytoplankton growth and grazer-induced mortality (Chen 2015;Morison and Menden-Deuer 2017), though such an approach has yet to be attempted for estimating virus-induced mortality. If they are occurring, treatment effects are expected to be largest between the most and least dilute treatments.…”

Exploring nonlinear functional responses of zooplankton grazers in dilution experiments via optimization techniques

“…Our results on the optimization of the choice of dilution factors contribute to the long‐standing discussion on constraining and optimizing the number of dilution incubations (Li et al ; Morison and Menden‐Deuer and citations therein). We find that the use of only two dilution points is acceptable only if we have a very accurate estimate of the half‐saturation constant of the functional response.…”

“…Finally, we show that the choice of dilutions for grazing experiments is important; better selection of dilution factors can largely decrease the error in the approximation of the nonlinear functional response. The latter finding may contribute to the practical question about a possible decrease in the number of dilutions required in a given experiment (Morison and Menden‐Deuer ).…”

“…Even when we have detailed simulation models, there will likely be a need to show a bulk validation of the emergent description of “microzooplankton predation” vs. “prey abundance” against field data series. Obtaining an accurate, simple, and reliable method of quantifying the functional response of microzooplankton, via the bulk interaction description, is of interest due to pronounced impact of grazers on primary production (Beckett and Weitz ; Calbet and Saiz ; Li et al ; Morison and Menden‐Deuer ), coupled with the appreciation that outputs from simulations are sensitive to this functional description in models (Mitra et al ).…”