PROJECT
Diagnosing stock-assessment model misspecification with machine learning
Simulation-estimation experiments testing whether multivariate diagnostics can identify structural sources of stock-assessment misspecification.
About
Quantitative fisheries science at the intersection of stock assessment, marine ecology, statistical modelling, machine learning, and reproducible scientific computing.
I am a quantitative fisheries scientist working at the intersection of stock assessment, marine ecology, statistical modelling, machine learning, and reproducible scientific computing. I study how environmental variability, population processes, and fishing shape exploited marine populations by combining fishery observations, scientific surveys, environmental information, and statistical, spatial, Bayesian, and stock-assessment models.
I am based at the Instituto del Mar del Perú (IMARPE), where I work on quantitative analyses of pelagic resources and fisheries. I also turn recurring analytical problems into reproducible methods and open research software for transparent, reusable, and auditable analysis.
A recurring goal is to connect data and models to evidence-based fisheries management across species, ecosystems, and scales of observation.
My work is organized around six connected research interests:
- Stock assessment & population dynamics: productivity, recruitment, mortality, selectivity, model diagnostics, and structural uncertainty.
- Environmental variability: climate and oceanographic forcing, pressure-state relationships, thresholds, and regime change.
- Reproduction, growth & condition: maturity, spawning, growth, condition, and life-history variation beyond biomass alone.
- Spatial ecology & biodiversity: species distributions, habitat suitability, SDMs, associated fauna, biodiversity, and bycatch.
- Fisheries dynamics & management: fleet and season dynamics, resource availability, biological risk, and adaptive management.
- Scientific computing: statistical methods, reusable software, data infrastructure, and reproducible research workflows.
Selected research
SOFTWARE
oceancube
Validated R infrastructure for representing, inspecting, subsetting, extracting, visualizing, and transforming multidimensional ocean data cubes for reproducible marine analysis.
TALK
Multivariate Health Index of the anchovy: Understanding the dynamics of small pelagic fish under environmental variability
A multivariate stock-health framework integrating complementary biological, demographic, spatial, and reproductive signals for highly variable small-pelagic populations.
Recent News and Posts
Statistical distributions for fisheries and marine ecology
Updated Sep 9, 2026 - Statistical ecology - 8 min read
A practical guide to choosing probability distributions for common response variables in fisheries and marine ecology.
Contact
For research collaboration, scientific discussion, or professional enquiries, see the Contact page.