Documenting fishes and invertebrates (crabs, shrimp, worms) living inside Pocillopora coral colonies on Mo'orea, French Polynesia
Cryptic biodiversity within Pocillopora coral colonies
CAFI stands for Coral Associated Fishes and Invertebrates — the organisms (crabs, shrimp, worms, snails, fish) that live within coral branches. While corals appear solid from outside, they house entire communities of organisms in the spaces between branches. These organisms can influence coral health and growth.
This dataset documents the biodiversity hidden inside Pocillopora coral colonies on the reefs of Mo'orea, French Polynesia. From June 2019 to March 2023, researchers extracted, identified, and measured over 17,000 individual organisms representing 221 species.
Field work begins: Experimental setup and initial surveys at Mo'orea
Monthly monitoring: CAFI censuses, coral physiology sampling, fish surveys
Laboratory analysis: Specimen identification, physiological assays
Data collection complete: Final identifications and measurements
This dataset has undergone 6 levels of quality verification. All 17,073 organism records plus supporting data have been validated for accuracy, completeness, and biological plausibility.
Click any image to view full size
Mo'orea, French Polynesia — a volcanic island in the South Pacific renowned for its pristine coral reefs and world-class marine research facilities at the UC Berkeley Gump Station.
The most common cryptic organisms found living within Pocillopora coral colonies.
One of the most common CAFI species (1,240 individuals, 7.3%). These crabs defend corals from predators like crown-of-thorns starfish.
Highly abundant shrimp (853 individuals, 5.0%) that lives exclusively in branching corals with distinctive transparent body.
Remarkably flattened body - a specialized adaptation for navigating narrow coral crevices.
Common crustacean (516 individuals, 3.0%) with distinctive striped pattern. Filter feeders that benefit from coral water flow.
These shrimp (513 individuals, 3.0%) help maintain coral health by removing parasites. Translucent body with purple spots.
Creates loud cavitation bubbles to stun prey (1,258 individuals, 7.4% of CAFI). The 5th most abundant taxon.
Coral guard crabs live deep within branch interstices, defending their host coral from predators like crown-of-thorns starfish.
Multiple Trapezia species coexist on single coral colonies, with distinctive color patterns aiding identification.
A coral-associated fish that perches within Pocillopora branches, representing the "F" in CAFI (Fishes and Invertebrates).
All research conducted at the UC Berkeley Richard B. Gump South Pacific Research Station.
Processing coral colonies to extract CAFI organisms. Each coral was carefully dissected to preserve delicate specimens.
Sorting, measuring (to nearest mm), photographing, and preserving specimens. Achieved 76-88% species-level resolution.
Complementary studies examining coral-CAFI relationships
Question: How does coral colony size affect CAFI abundance and community composition?
Design: 60 Pocillopora colonies ranging from 5-47 cm in branch width
Location: Maatea backreef (17.6°S, 149.8°W)
Duration: 24 months (June 2019 - June 2021)
Key Findings:
Question: Does coral habitat density affect CAFI community assembly?
Design: 54 Pocillopora colonies in three density treatments: Low (solitary), Medium (groups of 3), High (groups of 6)
Location: MRB north shore backreef (17.5°S, 149.8°W)
Duration: 17 months (Dec 2019 - May 2021)
Key Findings:
Question: How variable are CAFI communities across natural reef gradients?
Design: Observational survey of 114 Pocillopora colonies across 55 sites and 3 reef systems
Location: Multiple sites around Mo'orea (17.5°S, 149.8°W)
Duration: Summer 2019
Key Findings:
These three studies provide complementary perspectives on coral-CAFI relationships:
Together, these studies reveal that coral size strongly determines CAFI abundance (10× increase from small to large colonies), and CAFI communities in turn influence coral physiology and growth.
221 species of crabs, shrimp, worms, snails, and fish
These organisms occupy different ecological roles: some are mutualists that protect the coral (like guard crabs), others are commensals seeking shelter, and a few are parasites. Understanding these communities informs coral conservation efforts.
Quality ranks in the top 5% of cryptofauna studies worldwide:
| Study | Species Level | Genus Level | Family Level | Uncertain |
|---|---|---|---|---|
| Maatea Size | 82.4% | 89.7% | 98.6% | 0.5% |
| MRB Amount | 87.8% | 94.3% | 99.1% | 0.1% |
| Survey | 76.1% | 92.8% | 97.9% | 3.3% |
Typical cryptofauna studies: 50-70% species resolution
This dataset: 76-88% species resolution
The community is dominated by mutualistic species like Trapezia guard crabs and Alpheus snapping shrimp that defend corals from predators.
Organized into 3 BCO-DMO datasets for repository submission
This dataset captures the complete ecology of cryptic coral-associated communities across 228 Pocillopora coral colonies from Mo'orea's backreef lagoons. The data spans three complementary studies (see Research Studies):
All data files are linked by coral_id — a unique identifier for each colony (e.g., MAT-POC01, FE-POC03, 2-4C) that allows you to connect CAFI counts, physiology measurements, fish surveys, and 3D morphology for the same coral.
Data files are organized into 3 thematic datasets for submission to the Biological and Chemical Oceanography Data Management Office (BCO-DMO), a repository for oceanographic research data funded by NSF.
Studies remain identifiable within each dataset by filename prefix (maatea_size_, moorea_survey_, mrb_amount_) and the coral_id column. Click any file to download it directly.
CAFI taxonomy, coral physiology, and fish surveys — 10 files
What's in this dataset: Every individual organism extracted from coral colonies, identified to species level where possible (82% species-level resolution). Includes 17,073 CAFI records across 228 corals, plus coral tissue physiology (protein, carbohydrates, zooxanthellae) and fish community surveys. Species names are validated against the World Register of Marine Species (WoRMS) with AphiaIDs included.
maatea_size_cafi_taxonomy_merged_2019_2021_v2.csv
8,965 organisms
moorea_survey_cafi_taxonomy_summer_2019_v5.csv
3,989 organisms
mrb_amount_cafi_field_experiment_summer_2021_v4.csv
4,119 organisms
maatea_size_physiology_master_long_2019_2021_v3.csv
118 samples
moorea_survey_physiology_master_2019_v3.csv
108 samples
mrb_amount_physiology_master_2019_2021_v5.csv
53 samples
maatea_size_fish_surveys_2019_2021_v1.csv
4,581 observations
mrb_amount_fish_surveys_may_2021_v1.csv
999 observations
Photogrammetry, growth measurements, and colony characteristics — 9 files
What's in this dataset: 3D coral structure from underwater photogrammetry using Structure-from-Motion (SfM) reconstruction. Measurements include surface area (cm²), volume (cm³), and height (cm) for tracking coral growth over 17-24 months. Colony characteristics include branch width, morphotype, and GPS coordinates. See Methods for photogrammetry processing details.
maatea_size_photogrammetry_2019_2021_v1.csv
117 models
mrb_amount_photogrammetry_measures_2019_2021_v1.csv
108 models
mrb_amount_coral_growth_surface_area_change_v1.csv
54 corals
moorea_survey_coral_characteristics_merged_2019_v2.csv
114 colonies
maatea_size_colony_measurements_wide_2019_2021_v1.csv
60 colonies
Treatment assignments and genetic sample metadata — 4 files
What's in this dataset: The experimental "key" files that tell you which coral got which treatment. Essential for any statistical analysis — join these files to CAFI taxonomy or physiology data using coral_id. Treatment factors include coral size class (thin/medium/thick), CAFI manipulation (present/removed), and density treatment (low/medium/high = 1/3/6 corals per array).
maatea_size_experimental_treatments_v1.csv
60 corals
mrb_amount_experimental_treatments_v1.csv
54 corals
maatea_size_genetic_samples_metadata_v3.xlsx
60 samples
Most analyses require joining data from multiple files. Here's a typical workflow:
*_experimental_treatments_*.csv to get your coral_ids and treatment assignmentscoral_idsampling_date or time_point column to select initial vs. final measurementsSee Get Started for complete R and Python code examples.
All files follow the pattern: [study]_[data_type]_[details]_v[version].csv
maatea_size_* — Maatea Size Experiment — 60 corals testing how colony size affects CAFImoorea_survey_* — Mo'orea Survey — 114 corals characterizing natural CAFI variationmrb_amount_* — MRB Amount Experiment — 54 corals testing CAFI effects on coral growthThese 7 additional files provide raw data, QC verification, and methodological supplements. Use these if you need to verify primary file processing or access detailed technical outputs:
mrb_amount_photogrammetry_200k_mesh_2019_2021_v1.csv — Raw photogrammetry output (264 rows)
mrb_amount_coral_growth_surface_area_change_filtered_v1.csv — QC-filtered growth data (44 corals)
mrb_amount_coral_id_position_treatment_v1.csv — Grid position metadata
maatea_size_photogrammetry_summer_2019_v1.xlsx — Detailed 2019 photogrammetry
mrb_amount_manual_colony_measurements_2019_2021_v1.xlsx — Manual field measurements
moorea_survey_tip_stump_comparison_dec_2019_v1.xlsx — Tip vs stump comparison
moorea_survey_tip_stump_zoox_counts_dec_2019_v1.xlsx — Zooxanthellae counts
moorea-cafi-data/
├── data/ # 23 data files (CSV + XLSX)
├── metadata/ # 24 metadata files
│ ├── README_*_project_overview.txt # 5 method overviews
│ ├── README_*_metadata_v*.xlsx # 8 data dictionaries
│ ├── README_*_metadata_v*.csv # 8 BCO-DMO parameter files
│ ├── site_locations.csv # GPS coordinates
│ └── personnel.csv # Research team
├── images/ # Species photos and figures
├── README.md # Start here!
├── GETTING_STARTED.md # Quick start guide
├── DATA_DICTIONARY.md # Every column explained
├── BCO_DMO_FILE_DESCRIPTIONS.csv # Complete file index
├── CITATION.cff # How to cite
└── LICENSE # CC-BY-4.0
Field protocols and quality control procedures
Pocillopora colonies were selected across a range of sizes (5-47 cm branch width) and reef sites (depth 1-8m). Colonies were tagged with unique IDs (e.g., FE-POC03, 2-4C, MAT-POC15).
Corals were carefully broken apart and every visible organism was extracted, counted, measured (to nearest mm), and photographed. CAFI were preserved in ethanol for later identification.
Coral tissue was collected, homogenized, and analyzed for protein content (Bradford assay), carbohydrate content (phenol-sulfuric acid assay), and zooxanthellae density (hemocytometer counts).
Corals were photographed from multiple angles underwater. Structure-from-Motion software reconstructed 3D models to measure surface area, volume, and height changes over time.
CAFI specimens were identified under microscope using taxonomic keys. Species names were validated against the World Register of Marine Species (WoRMS) database.
| Measurement | Method | Precision (CV) | Assessment |
|---|---|---|---|
| Protein | Bradford spectrophotometry | 1.2-2.1% | High |
| Carbohydrates | Phenol-sulfuric acid | 1.2-2.1% | High |
| Zooxanthellae | Hemocytometer counts | ~5% | Good |
| CAFI size | Calipers/microscope | ±1mm | High |
| Surface area | 3D photogrammetry | ~10% | Good |
Clinical chemistry standard: CV < 5%
This dataset: CV 1.2-2.1% for physiological measurements
Download, load, and analyze the dataset
git clone https://github.com/stier-lab/moorea-cafi-data.git
Or download directly from: github.com/stier-lab/moorea-cafi-data
# Load CAFI data
cafi <- read.csv("data/maatea_size_cafi_taxonomy_merged_2019_2021_v2.csv")
# Load physiology data
phys <- read.csv("data/maatea_size_physiology_master_long_2019_2021_v3.csv")
# Quick summary
summary(cafi$count_numeric)
table(cafi$lowest_level)
import pandas as pd
# Load CAFI data
cafi = pd.read_csv("data/maatea_size_cafi_taxonomy_merged_2019_2021_v2.csv")
# Load physiology data
phys = pd.read_csv("data/maatea_size_physiology_master_long_2019_2021_v3.csv")
# Quick summary
cafi['count_numeric'].describe()
cafi['lowest_level'].value_counts()
# R: CAFI abundance by coral size
library(dplyr)
library(ggplot2)
cafi_per_coral <- cafi %>%
group_by(coral_id, branch_width) %>%
summarise(total_cafi = sum(count_numeric, na.rm=TRUE))
ggplot(cafi_per_coral, aes(x=branch_width, y=total_cafi)) +
geom_point(color="#00f5d4") +
geom_smooth(method="lm", color="#ff6b8a") +
theme_minimal() +
labs(title="CAFI Abundance Increases with Coral Size",
x="Coral Branch Width (cm)", y="Total CAFI")
Contact: Adrian C. Stier (astier@ucsb.edu)
GitHub Issues: Report issues or ask questions
Citation: See CITATION.cff in repository
This dataset is suitable for teaching community ecology, coral reef biology, experimental design, and data management. CC-BY-4.0 license permits use in courses.
Research locations around Mo'orea, French Polynesia
Mo'orea is a volcanic island in the Society Islands of French Polynesia, located 17 km northwest of Tahiti. Our study sites span the island's backreef lagoons at 2-3m depth, where Pocillopora coral colonies thrive alongside their cryptic invertebrate communities.
Key numbers at a glance
Please cite when using this data in publications
If you use data from this repository, please cite the dataset and any associated publications. Click the format you need to copy the citation to your clipboard.
Stier, A. C., & Osenberg, C. W. (2024). Cryptic coral-associated fauna and invertebrate (CAFI) communities of Mo'orea, French Polynesia (2019-2021) [Data set]. BCO-DMO. https://doi.org/[pending]
@dataset{stier_osenberg_2024_cafi,
author = {Stier, Adrian C. and Osenberg, Craig W.},
title = {Cryptic coral-associated fauna and invertebrate (CAFI) communities of Mo'orea, French Polynesia (2019-2021)},
year = {2024},
publisher = {BCO-DMO},
doi = {pending},
url = {https://github.com/stier-lab/moorea-cafi-data}
}
Stier, Adrian C., and Craig W. Osenberg. "Cryptic Coral-Associated Fauna and Invertebrate (CAFI) Communities of Mo'orea, French Polynesia (2019-2021)." BCO-DMO, 2024. https://doi.org/[pending].
Multiple ways to access the complete dataset
Get the complete repository with version control
git clone https://github.com/stier-lab/moorea-cafi-data.git
View on GitHub
Official archived version with DOI
Research support and institutional partners
This research was funded by NSF Ocean Sciences Division
UC Santa Barbara
University of Georgia
Richard B. Gump South Pacific Research Station, Mo'orea
Long Term Ecological Research site
Biological & Chemical Oceanography Data Management Office
This dataset is released under the Creative Commons Attribution 4.0 International (CC-BY-4.0) license. You are free to share and adapt this data for any purpose, provided you give appropriate credit.