Accounting for the bin structure of data removes bias when fitting size spectra

Edwards, Andrew M.; Robinson, James P. W.; Blanchard, Julia L.; Baum, Julia K.; Plank, Michael J.

doi:10.3354/meps13230

Cited by 24 publications

(84 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use a Weibull distribution for the delay function w ( s ), and fit the shape and scale parameters using the case-specific data of reported cases and time of symptom onset (Figure 2B). In the Supplement we derive a novel multinomial likelihood function (based on [19] and [20]) to handle right-truncation. The resulting maximum likelihood estimate (with 95% univariate confidence interval) of the shape is 1.73 (1.60–1.86) and of the scale is 9.85 (9.30–10.46).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Estimating the impact of COVID-19 control measures using a Bayesian model of physical distancing

Anderson

Edwards

Yerlanov

et al. 2020

Preprint

View full text Add to dashboard Cite

Extensive physical distancing measures are currently the primary intervention against coronavirus disease (COVID-) worldwide. It is therefore urgent to estimate the impact such measures are having. We introduce a Bayesian epidemiological model in which a proportion of individuals are willing and able to participate in distancing measures, with the timing of these measures informed by survey data on a itudes to distancing and COVID-. We t our model to reported COVID-cases in British Columbia, Canada, using an observation model that accounts for both underestimation and the delay between symptom onset and reporting. We estimate the impact that physical distancing (also known as social distancing) has had on the contact rate and examine the projected impact of relaxing distancing measures. We nd that distancing has had a strong impact, consistent with declines in reported cases and in hospitalization and intensive care unit numbers. We estimate that approximately . ( . -. % CI) of contacts have been removed for individuals in British Columbia practising physical distancing and that this fraction is above the threshold of . at which prevalence is expected to grow. However, relaxing distancing measures beyond this threshold re-starts rapid exponential growth. Because the extent of underestimation is unknown, the data are consistent with a wide range in the prevalence of COVID-in the population; changes to testing criteria over time introduce additional uncertainty. Our projections indicate that intermi ent distancing measures-if su ciently strong and robustly followedcould control COVID-transmission, but that if distancing measures are relaxed too much, the epidemic curve would grow to high prevalence.

show abstract

Section: Methodsmentioning

confidence: 99%

“…Given the data h nr and these probabilities, we develop a multinomial log-likelihood function [19], adapting the approach of [20], to estimate the parameters λ and k . The log-likelihood function for λ and k , given the counts { h nr }, is where the W (·) terms depend on λ and k .…”

Section: Supplemental Methodsmentioning

confidence: 99%

Estimating the impact of COVID-19 control measures using a Bayesian model of physical distancing

Anderson

Edwards

Yerlanov

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…All statistical analyses and plots were conducted with the software R (version 4.0.0; R Development Core Team 2008). Specifically the MLE's of fish and zooplankton size spectra were performed by the provided R code and package (sizeSpectra) from Edwards et al (2017) and Edwards (2020).…”

Section: Data Analysesmentioning

confidence: 99%

Top-down and bottom-up effects on zooplankton size distribution in a deep stratified lake

2021

View full text Add to dashboard Cite

Trophic interactions in the pelagic area of lakes and the opposing effects of fish feeding (top-down) and phytoplankton biomass (bottom-up) on zooplankton communities are central topics in limnology. We hypothesized that zooplankton size distributions should be a more sensitive approach to disentangle top-down and bottom-up effects than the commonly measured zooplankton biomass. We examined zooplankton size distributions from 148 samples collected during summer months in the upper and lower pelagic layers of a deep mesotrophic lake among 13 years of sampling. Top-down effects, namely fish size and biomass, and bottom-up effects, including water temperature and total phosphorus and chlorophyll a concentrations, were considered. To add robustness to our analyses, we expressed the zooplankton size distributions as size spectra based on log-binning, as continuous size spectra and by the size diversity, a measure that has been developed to mimic taxonomic diversity indices. Among numerous regressions tested, significant top-down or bottom-up effects could rarely been detected. Our results indicate that the overall zooplankton size distribution was not significantly affected by fish predation and lake productivity measured as total phosphorus or chlorophyll a concentration. However, we found negative correlations between fish biomass and the preferred zooplankton prey, including Bosmina longirostris, Daphnia cucullata and nauplii in the upper depth layer. However, due to their small body size, low biomass and therefore relative small contribution to the zooplankton size distribution, predation on preferred zooplankton species did not translate into a statistically significant modification of the entire size spectrum. Consequently, the size spectrum seems to be relatively robust against predation effects, but might reflect the lake-wide energy availability and transfer efficiency in the food web.

show abstract

“…There are a number of methods to fit the size-spectra exponent; however, the use of various calculations of b has been demonstrated to produce conflicting or biased estimates (Edwards et al 2017), limiting the comparability of results across studies. Following recommendations regarding the precision and accuracy of widely utilized methods, and recent investigations of reef-fish size-spectra (Robinson et al 2017;Edwards et al 2020), we used the ''MLEbins'' methodology. MLEbins extends the maximum likelihood estimation of b to explicitly account for speciesspecific size-bin structures and has been demonstrated to remove biases and accurately estimate b even for lowresolution data (Edwards et al 2020), such as those collected by UVC (in comparison to large-scale fisheries-dependent surveys) or those that are collected using predetermined size-bins for length estimates, as is the case for fish \ 35 cm here.…”

Section: Response Variablesmentioning

confidence: 99%

The importance of biophysical context in understanding marine protected area outcomes for coral reef fish populations

Fidler

Andradi-Brown²,

Awaludinnoer³

et al. 2021

Coral Reefs

View full text Add to dashboard Cite

Multiuse marine protected areas (MPAs) that utilize a mosaic of no-take and sustainable use zones are an increasingly common tool for promoting fish population recovery while minimizing socioeconomic conflict. However, significant gaps remain in our understanding of the effects of multiuse MPAs on reef-fish populations, and how protection from fishing may interact with biophysical gradients to drive ecological outcomes. Here, we examine changes in fish assemblages inside and outside of two multiuse MPAs in Raja Ampat, Indonesia, over the first four to five years after their implementation. We use linear models to assess the impact of protection status, anthropogenic pressure, and biophysical gradients on changes in four biological metrics: abundance, biomass, mean size, and size-spectra. In addition, we use multivariate analyses to assess whether fish assemblages in protected and unprotected reefs diverged after MPA establishment. We find that both fish abundance and biomass were driven equally or more by benthic characteristics than protection status and that these environmental impacts were decoupled from the effects of MPAs. Contrary to expectations, increases in abundance were more pronounced in unprotected reefs, where fish assemblages diverged from protected reefs through an increase of herbivorous parrotfishes (Subfamily Scarinae) likely driven by increased algal production on hard-bottom reef areas. However, mean size and size-spectra improved in sustainable use zones compared to unprotected reefs, demonstrating that non-exclusionary management can provide ecological benefits, especially when combined with nearby no-take areas. Increased mean size was also associated with lower human population sizes, suggesting a strong influence of fishing pressure, which is mediated by MPAs. Our findings suggest that multiuse MPAs can quickly achieve some fisheries management objectives and that these effects are likely to increase over longer timescales. However, biophysical context can have significant impacts on reef-fish assemblages that are distinct from the effects of overlying fisheries management. Keywords Marine protected areas Á Marine conservation Á Biophysical context Á Coral reef fisheries Á Indonesia Topic Editor Andrew Hoey

show abstract

Accounting for the bin structure of data removes bias when fitting size spectra

Cited by 24 publications

References 44 publications

Estimating the impact of COVID-19 control measures using a Bayesian model of physical distancing

Estimating the impact of COVID-19 control measures using a Bayesian model of physical distancing

Top-down and bottom-up effects on zooplankton size distribution in a deep stratified lake

The importance of biophysical context in understanding marine protected area outcomes for coral reef fish populations

Contact Info

Product

Resources

About