Ecological and Statistical Consultancy

Bede Ffinian Rowe Davies Consultancy bedeffinian@gmail.com

Statistical Experience

My statistical experiences have ranged from:

• modelling decadal oceanographic changes over ocean basin scales to predict marine invasive species

• assessing changes in acoustic complexity index (a univariate measure of ecosystem noise), to assemblage level changes (multivariate abundance) to predictive spatial modelling of macroalgal species from satelite imagery

• abundance and diversity of taxa of importance, to whole ecosystem functional trait analysis.

The methods used have mostly been applied within the coding language {r}, with models being applied of varying complexity, in both frequentist and bayesian frameworks as well as Machine Learning algorithms, from one or two fixed effects to more complex mixed effects, hierarchical regression, General Additive, Random Forest, XGBoost, Neural Network classification models. These models, like the data, have been varying in distributions and characteristics with models using:

• gaussian distributions

• poisson distributions

• beta distributions

• gamma distributions

• bernoulli distributions

• binomial distributions

• zero-inflated/one-inflated/zero-altered/one-altered alternatives.

These analysis methods have been used to provide visiualisations of predicted scenarios across many of the projects.

Consultancy Fieldwork Experience

I have assisted in Fisheries Assessment cruises as a Benthic Taxonomist across the Greenlandic and Icelandic Seas.

During this cruise on the RV Tarajoq my main duties were to:

• Identify

• Quantify

• Record

All Benthic taxa caught as bycatch during fisheries assessments of Alantic Cod and Shrimp.

Statistical Experience

My statistical experiences have ranged from:

• modelling decadal oceanographic changes over ocean basin scales to predict marine invasive species

• assessing changes in acoustic complexity index (a univariate measure of ecosystem noise), to assemblage level changes (multivariate abundance) to predictive spatial modelling of macroalgal species from satelite imagery

• abundance and diversity of taxa of importance, to whole ecosystem functional trait analysis.

The methods used have mostly been applied within the coding language {r}, with models being applied of varying complexity, in both frequentist and bayesian frameworks as well as Machine Learning algorithms, from one or two fixed effects to more complex mixed effects, hierarchical regression, General Additive, Random Forest, XGBoost, Neural Network classification models. These models, like the data, have been varying in distributions and characteristics with models using:

• gaussian distributions

• poisson distributions

• beta distributions

• gamma distributions

• bernoulli distributions

• binomial distributions

• zero-inflated/one-inflated/zero-altered/one-altered alternatives.

These analysis methods have been used to provide visiualisations of predicted scenarios across many of the projects.

Consultancy Fieldwork Experience

I have assisted in Fisheries Assessment cruises as a Benthic Taxonomist across the Greenlandic and Icelandic Seas.

During this cruise on the RV Tarajoq my main duties were to:

• Identify

• Quantify

• Record

All Benthic taxa caught as bycatch during fisheries assessments of Alantic Cod and Shrimp.