Statistical analysis for environmental and marine data

Ecological data rarely behaves like the examples in a statistics textbook. Counts with many zeros, unbalanced designs, samples taken at the same site over and over. I choose methods that fit the data you actually have, check their assumptions, and write down why each choice was made.

What I can do

Design before fieldwork: how many sites and replicates you need to detect the effect you care about (power analysis), and how to lay them out.
Hypothesis tests and models: linear and generalised linear models, and mixed models for nested or repeated designs.
Community analysis: ordination (PCA, nMDS) and multivariate tests such as PERMANOVA, for questions about whole assemblages rather than single species.
Species richness: estimating how many species an area holds from incomplete surveys, with species accumulation curves and richness estimators.
Reporting: figures, tables and a methods section ready for a report, a reviewer or a journal.

How a project runs

A short call about the question, the data you have and when you need it.
A look at a sample of the data, then a written proposal with the approach, its limits and the time it will take.
The work itself, with the decisions written down as I go.
Delivery and a walk-through of the results, plus the files and code.

What you get

A written report with results, figures and their interpretation.
A methods section you can reuse, with assumptions and limits stated.
The Python code, so the analysis can be rerun when new data comes in.

Background

My degree thesis (presented in 2015) estimated how many sea anemone species live in Venezuelan shallow waters from 50 surveys, using species accumulation models. I have kept using statistics since, in an ecology laboratory and later in manufacturing, where I trained as a Lean Six Sigma Green Belt.

Tell me about your data.

A few lines are enough: the question, the data you already have and when you need it. I reply in English, Portuguese or Spanish.

Write to me