Data from: Predator-mediated trophic connectivity between reef and oceanic habitats in Palau
Data files
Apr 29, 2026 version files 84.80 KB
-
Arnoldi_et_al_Palau_fish_SI_data_2017to2022.csv
76.22 KB
-
README.md
8.59 KB
Abstract
Effective fisheries management and marine spatial planning depend on understanding how mobile predators link reef and oceanic ecosystems through cross-habitat movements and foraging. Predator resource use shapes movement and space-use strategies, providing insight into life history and behavioral variation, the identification of critical habitats, and spatial patterns of vulnerability to fishing and other human activities. Yet cross-habitat trophic connectivity remains poorly quantified in diverse coral reef and oceanic systems. Here, we used stable isotope (δ13C, δ15N) analysis, complemented by stomach content analysis, to examine cross-habitat resource use of predatory fishes across reef and near-reef oceanic ecosystems in Palau, in the Western Pacific, evaluating patterns across species, trophic guilds, and habitat associations. Cluster analyses, isotopic niche metrics, and Bayesian stable isotope mixing models revealed asymmetric but bidirectional connectivity, with resource utilization typically spanning both habitats but skewed towards reef or oceanic depending on species groups. Reef predators showed greater variability in resource use and higher reliance on oceanic inputs than oceanic predators relied on reef resources, with broad isotopic niches consistent with mobility, water-column use, and generalist foraging strategies. In contrast, most oceanic predators largely specialized on oceanic prey, although stomach contents documenting larval and juvenile reef fishes indicate occasional incorporation of reef-origin resources. Our results highlight the disproportionate importance of oceanic production in supporting reef predator populations, particularly where physical forcing concentrates prey near reef margins. Although reef contributions to oceanic predator diets were comparatively small, even limited inputs may be ecologically meaningful for migratory species during critical life stages. Together, these findings demonstrate widespread but trait-structured and individually variable predator-mediated connectivity that challenges binary “reef” versus “oceanic” classifications. Incorporating trophic connectivity into fisheries management and marine spatial planning can better align protection with the energetic pathways and habitat utilization sustaining predators in reef–ocean systems.
Dataset DOI: 10.5061/dryad.2280gb66z
Description of the data and file structure
Authors
Natalie S. Arnoldi1*, Aaron Carlisle2, Robert B. Dunbar3, Alan M. Friedlander4,5, Yimnang Golbuu6,7, Steven Lindfield8, Fiorenza Micheli1,9
Affiliations
1 Departments of Oceans and Biology, Hopkins Marine Station, Stanford University, Pacific Grove, CA 93950, USA
2 School of Marine Science and Policy, University of Delaware, Lewes, DE 19958, USA
3 Department of Oceans, Stanford University, Stanford, CA 94305, USA
4 Pristine Seas, National Geographic Society, Washington, DC, 20036, USA
5 Hawaiʻi Institute of Marine Biology, University of Hawaiʻi, Kāneʻohe, Hawaiʻi, 96744, USA
6 Palau International Coral Reef Center, Koror, Palau, 96940, PW
7 The Nature Conservancy, Koror, Palau, 96940, PW
8 Coral Reef Research Foundation, Koror, Palau, 96940, PW
9 Stanford Center for Ocean Solutions, Stanford University, Pacific Grove, CA 93950, USA
*Corresponding author; nataliearnoldi90@gmail.com
Journal: Ecological Applications
Manuscript ID: EAP26-0093
Date of Data Collection: 2017-2022
Location: Republic of Palau
Description of the Data and File Structure
This dataset contains carbon and nitrogen stable isotope values and associated metadata for fish muscle tissue used in the associated publication.
Files and variables
File List:
- Arnoldi_et_al_Palau_fish_SI_data_2017to2022.csv: Primary dataset containing isotope values and fish metrics.
- README.md: This file.
Variables & Units
- [Species]: Scientific name of fish species.
- [Group]: Teleost (bonyfish) or chondrichthyan (cartilaginous fish).
- [Habitat]: Primary habitat association (reef vs. oceanic).
- [Sample_collection_date]: Date of collection (format: YYYY-MM-DD).
- [Year]: Year of collection (format: YYYY).
- [Season]: Boreal season (Winter, Fall, Summer, Spring).
- [FL_cm]: Fork length of fish in centimeters.
- [Lat]: Latitude in decimal degrees (WGS84).
- [Lon]: Longitude in decimal degrees (WGS84).
- [d15N]: δ15N; nitrogen stable isotope ratio (‰) relative to atmospheric N2.
- [d13C]: δ13C; carbon stable isotope ratio (‰) relative to VPDB.
- [C.N]: Carbon to nitrogen ratio (C:N) in sample.
- [d13C.]: Lipid corrected carbon stable isotope ratio.
- [Correction_eq]: Correction equation used from Logan et al. (2008); see note below.
- [Common_name]: Fish species common names.
- [Sex]: Sex of sampled fish (where possible, see below).
- [Sample_collection_group]: Group of researchers to collect each sample lead by: Dr. Steven Lindfield (collected reef fish samples from 2017-2019; n = 266) or Dr. Natalie Arnoldi (collected oceanic and some reef fish samples from 2019-2022; n = 274).
- Note*: NAs present in the FL_cm; Lat, and Lon columns represent where those variables are not available for those samples.
Methodological Information
-
Sample Processing: All tissue samples were frozen at -20 °C after collection, then dried for 48 hours at 52 °C and homogenized using a mortar and pestle. Shark tissue samples were treated with a deionized water and chloroform-methanol (2:1) solution to remove urea and lipids (Kim and Koch 2012, Bennett-Williams et al. 2022).
-
Analyses: Stable carbon and nitrogen isotope ratios and elemental concentrations were measured at the Stanford Stable Isotope Biogeochemistry Laboratory using a Thermo Delta V Advantage IRMS coupled to a Thermo FlashSmart elemental analyzer via a ConFlo IV interface, with analytical precision < 0.1 ‰ for δ13C and δ15N. Samples with C:N ratio > 3.5 were arithmetically corrected for lipid bias (Post et al. 2007, Wells et al. 2008) using equations from Logan et al. (2008).
-
Lipid correction equations from Logan et al. (2008) Appendix 1:
o Fish muscle tissue (“L1a.fish.muscle”): δ13C’ – δ13C = [a*(C:N)+b]/[(C:N)+c]
§ a = 7.415 ± 0.5576
§ b = -22.732 ± 1.5722
§ c = 0.746 ± 0.5734
o Tuna muscle tissue (Thunnus thynnus; “L2.tuna.muscle”): ”): δ13C’ – δ13C = P – [(P*F)/(C:N)]
§ P = 7.489 ± 0.1439
§ F = 3.097 ± 0.0138
o Variables:
§ [δ13C’]: Lipid corrected carbon stable isotope ratio (‰)
§ [δ13C]: Carbon stable isotope ratio (‰)
§ [C:N]: Carbon to nitrogen ratio in sample
§ [a, b, c, P, F]: Parameter estimates (± SE) for models fit to a dataset of marine fish muscle samples (See Logan et al. (2008) Table 1).
- Sex identification: Lindfield and associates performed histology on many fish sampled and were thus able to identify sex. Arnoldi and associates did not conduct any histological sampling and were thus unable to identify sex of bony fishes they sampled. Sex ID’s are provided for sharks where possible.
Works Cited
-
Bennett-Williams, Joshua, Christina Skinner, Alex S. J. Wyatt, Rona A. R. McGill, and Trevor J. Willis. 2022. “A Multi-Tissue, Multi-Species Assessment of Lipid and Urea Stable Isotope Biases in Mesopredator Elasmobranchs.” Frontiers in Marine Science 9 (March). https://doi.org/10.3389/fmars.2022.821478.
-
Kim, Sora L., and Paul L. Koch. 2012. “Methods to Collect, Preserve, and Prepare Elasmobranch Tissues for Stable Isotope Analysis.” Environmental Biology of Fishes 95 (1): 53–63.
-
Logan, John M., Timothy D. Jardine, Timothy J. Miller, Stuart E. Bunn, Richard A. Cunjak, and Molly E. Lutcavage. 2008. “Lipid Corrections in Carbon and Nitrogen Stable Isotope Analyses: Comparison of Chemical Extraction and Modelling Methods.” The Journal of Animal Ecology 77 (4): 838–46.
-
Post, David M., Craig A. Layman, D. Albrey Arrington, Gaku Takimoto, John Quattrochi, and Carman G. Montaña. 2007. “Getting to the Fat of the Matter: Models, Methods and Assumptions for Dealing with Lipids in Stable Isotope Analyses.” Oecologia 152 (1): 179–89.
-
Wells, R. J. David, James H. Cowan, and Brian Fry. 2008. “Feeding Ecology of Red Snapper Lutjanus Campechanus in the Northern Gulf of Mexico.” Marine Ecology Progress Series 361 (June): 213–25.
Code/software
All analyses for this study were conducted in R (v. 4.2.3; (R core team 2023)).
Packages used:
-
Post-hoc pairwise comparisons were conducted using Tukey’s HSD implemented in the multcomp package in R (Hothorn, Bretz, and Westfall 2008).
-
Hierarchical cluster analysis on species-mean δ13C and δ15N values were conducted using the factoextra package in R (Kassambara and Mundt 2021).
-
Partitioning around medoids (PAM) clustering to individual δ13C and δ15N values for all predator species with n ≥ 5 was implemented using the cluster package in R (Maechler et al. 2019)
-
The Stable Isotope Bayesian Ellipses in R (SIBER) package (Jackson et al. 2011) was used to estimate both maximum-likelihood and Bayesian isotopic niche metrics for predatory fish species with n ≥ 5.
-
Bayesian SIMMs were implemented using the mixSIAR package in R (Stock et al. 2018) for predator species with n ≥ 10.
References
Hothorn, Torsten, Frank Bretz, and Peter Westfall. 2008. “Simultaneous Inference in General Parametric Models.” Biometrical Journal. Biometrische Zeitschrift 50 (3): 346–63.
Jackson, Andrew L., Richard Inger, Andrew C. Parnell, and Stuart Bearhop. 2011. “Comparing Isotopic Niche Widths among and within Communities: SIBER - Stable Isotope Bayesian Ellipses in R: Bayesian Isotopic Niche Metrics.” The Journal of Animal Ecology 80 (3): 595–602.
Kassambara, A., and Fabian Mundt. 2021. “Factoextra: Extract and Visualize the Results of Multivariate Data Analyses, R Package Version 1.0. 7. 2020.” Preprint.
Maechler, M., P. Rousseeuw, A. Struyf, and M. Hubert. 2019. “Cluster: Cluster Analysis Basics and Extensions.” R Package Version.
R core team. 2023. R: A Language and Environment for Statistical Computing (version 4.3.3). R Foundation for Statistical Computing, Vienna, Austria. Computing RFfS.
Stock, Brian C., Andrew L. Jackson, Eric J. Ward, Andrew C. Parnell, Donald L. Phillips, and Brice X. Semmens. 2018. “Analyzing Mixing Systems Using a New Generation of Bayesian Tracer Mixing Models.” PeerJ 6 (June): e5096.
