Table of Contents

• Coverage
• Dataset Description
  • Methods & Sampling
  • BCO-DMO Processing Description
• Related Publications
• Parameters
• Instruments
• Project Information
• Program Information
• Funding

Sponge collection from Autonomous Reef Monitoring Structures and field surveys:
Autonomous Reef Monitoring Structures (ARMS) are standardized sampling devices that mimic reef interstices, attracting cryptobiota colonization (Brainard et al. 2009, Knowlton et al. 2010). Standard ARMS units are comprised of an eight-tiered stack of gray Type I PVC plates (22.9 x 22. 9 centimeters (cm)), arranged in four open and four semi-closed layers (Figure S1a; Leray and Knowlton 2015), while modified ARMS units are composed of a two-tiered stack of one open and one semi-closed layer (Figure S1b; Timmers et al. 2020). Sponges were sampled from six standard ARMS deployed along the reef slope of Moku o Lo'e (Coconut Island) and from six modified ARMS (Figure S1c) hovering in the water column attached to a Moku o Lo'e intake pipe, along an adjacent reef slope. An additional 24 modified units were placed within mesocosm tanks on Moku o Lo'e that received unfiltered seawater from the same intake pipe but were exposed to future climate conditions as described in Bahr et al. (2020) (Figure S1a). The ARMS in mesocosm tanks and those at the intake pipe were retrieved for sponge subsampling every two months for two years, and sponges from the full ARMS on the reef slope were collected once upon recovery, in July 2018. The ARMS units in ensemble provided a total combined sampling surface area of 15 square meters (m²) at each period of collection. At each collection period, ARMS units were disassembled for high-resolution plate imagery and carefully examined for newly settled sponge recruits. Sponges showing unique morphological features on each plate were individually photographed, carefully subsampled, and fixed in 95 % ethanol for DNA extraction. If enough tissue was available, sponges were additionally fixed in two solutions: one containing 4% paraformaldehyde in seawater, and the other containing 4% glutaraldehyde in 0.1M sodium cacodylate with 0.35M sucrose for future histological evaluation. A total of 439 sponge samples were collected from the ARMS units.

Sponges were also collected on reef substrates along a 50-meter (m) transect line from 34 sites in Kāne'ohe Bay and one site on the Makai Pier in Waimanalo (see Table S2 for GPS coordinates and collection depth) throughout the two-year mesocosm experimental period. Collection on Kāne'ohe Bay reef sites included samples taken by global taxonomic experts during the Smithsonian-led MarineGEO biodiversity surveys in 2017. 163 marine sponges showing unique morphologies were haphazardly collected in the photic zone of the reef at a depth range of 1-16 m from within crevices, beneath coral rubble, fouling upon structures and under overhangs. Presence/absence of sponge OTUs were recorded at each surveyed site (Table S3), were photographed, and fixed in 95 % ethanol for DNA extraction. Additional sponge metadata pertaining to specimen morphology, such as color, consistency, surface, oscules, exudates, and odors was also recorded.

All samples were vouchered with the Florida Museum of Natural History at the University of Florida (UF Porifera) and the Hawaiʻi Institute of Marine Biology ("KB" or "KBOA") (Table S4). Images and associated metadata of each sponge sample are publicly available at https://www.invertebase.org/portal/ and http://specifyportal.flmnh.ufl.edu/iz/. All samples were collected under special activities collection permit (SAP) nos. 2018-03 and 2019-16 (covering the period of 13 Jan 2017 through 10 Apr 2019) issued by the State of Hawai'i Division of Aquatic Resources.

Sponge DNA extraction and sequencing:
Vouchered specimens were subsampled for DNA extraction using the E-Z 96 Tissue DNA Kit (Promega Bio-Tek, Norcross, GA, USA) following the manufacturer protocols. Care was taken to subsample only sponge material free of other organisms, which would contaminate the sponge DNA extract. Multiple primers were used in a stepwise fashion to successfully amplify partial fragments of both 28S rRNA and COI genes using polymerase chain reaction (PCR) (Table S1 in Timmers et al., 2020). Fragments of the COI were initially attempted with primer pairs LCO1490 / COXR1 (~1400 bp fragment) (Folmer et al. 1994), followed by primers jgLCO1490 / jgHCO2198 (Geller et al. 2013) (~648 bp fragment) within the previous PCR fragment region and a final attempt with subsequent internal primers mlCOIint / jgHCO2198 (~313 bp) (Leray et al. 2013). Similar to the approach used for the COI, amplification of 28S rRNA fragments were first attempted with primers F63mod / 1072RV (~1050 bp) (Medina et al. 2001), followed by internal primers (28S-C2-fwd / 28S-D2-rev) (450 bp) (Chombard et al. 1998) within the previous fragment, and a final attempt with 28SMycF / 1072RV. PCR reactions were carried out in a total volume of 40 microliters (µL) including the following: 14.4 µL of H₂O, 20 µL of BioMix Red (Bioline, Taunton, MA, USA) PCR Mastermix, 0.8 µL of each primer (10 mM), 3.2 µL of bovine serum albumin (BSA) (100 mg/mL), and 0.8 µL of template DNA (1 to 30 nanograms per microliter (ng/µL)). PCR products were examined on a 1% agarose gel stained with GelRed and purified using ExoFAP (Exonuclease I and FastAP - Life Technologies, Carlsbad, CA) prior to sequencing. When products showed multiple bands above the 100 bp ladder mark, products were purified using gel excision by loading 40 µL of the PCR product onto a 2% agarose gel made with 1x modified (no EDTA) TAE running them at 50 millivolts (mV). After 90 minutes, bands were excised with a sterile scalpel, loaded onto a column filter fitted inside a 1.5 milliliter (mL) centrifuge tube, and centrifuged for 10 minutes at 5,000 rpm. Sequencing reactions were performed in both directions using the Big Dye TM terminator v. 3.1, and sequencing was done on an ABI Prism 3730 XL automated sequencer at the University of Hawai'i Advanced Studies of Genomics, Proteomics and Bioinformatics sequencing facility.

Forward and reverse reads were trimmed and edited by eye using Geneious 10 (Kearse et al. 2012). Assembled and edited sequences were exported as fasta files and checked for contamination by using the BLAST (Altschul et al. 1990) function in GenBank. Sequences showing >85% sequence identity to those belonging to Porifera were kept and used for further analysis. 28S rRNA sequences for 592 samples were produced, but only 340 sequences were deposited in GenBank as many were repetitive sequences with 100% identity. When available, up to three replicate sequences per OTU were deposited and assigned accession numbers MW016037 - MW016376. 98 COI sequences were deposited in GenBank and assigned accession numbers MW059039- MW059109; MW144969-MW144988; MW143251-MW143256; MW349624 (Table S4).

Phylogenetic analysis and taxonomic assignments:
Sequences were aligned with the closest sequence relatives in the GenBank database using the ClustalW algorithm with default parameters in Geneious. Sequence KJ483037.1 Parazoanthus puertoricense was used as an outgroup for all phylogenetic topologies of partial 28S rRNA sequences and AB247348.1 Epizoanthus arenaceus was used as an outgroup for the phylogenetic topology of partial COI sequences. Bayesian inference (BI) using MrBayes version 3.2.1 (Huelsenbeck & Ronquist 2001) and a maximum likelihood (ML) framework using RaxML (Stamatakis 2006) were added to the phylogenetic analysis. The GTR substitution model and GTRGAMMA nucleotide model with 1,000 bootstrap replicates were implemented in the BI and ML analyses, respectively. The BI was run using 5 million generations sampled every 200 generations. The analysis was stopped when the standard deviation (SD) of split frequencies fell below 0.01.

Most sponge OTUs were delineated using a combination of ≥1% COI and 28S rRNA sequence divergence combined with unique morphological features and classified as distinct operational taxonomic units (OTUs). A handful of OTUs that were morphologically clearly differentiable, but had ≤ 1% sequence divergence were also recognized as distinct OTUs (Table S6). This conserved threshold was chosen based on the different rates of evolution that can exist within poriferan families and even genera which make the selection of an accurate threshold for delineating sponge OTUs arbitrary (Erpenbeck et al. 2007, Wang & Lavrov 2008, Redmond et al. 2011, Voigt & Wörheide 2016, Yang et al. 2017).

Preliminary assessments of morphological characters (i.e. color, consistency, surface, oscules, and skeleton composition) were made mostly from OTUs that matched previous vouchered sponge collections in Kāneʻohe Bay (De Laubenfels 1950, Bergquist 1967, 1977, Pons et al. 2017) (Table S5). We assigned OTUs to taxonomic levels based on the placement of each barcode into the lowest well-supported clade (bootstrap support of ≥50) in the COI and 28S rRNA tree topologies. On average, taxonomic identities followed these barcode sequence identity percentages: Order (>90%), Family (>95%), Genus (>98%) and for the species above (100%). Phylogenetic topologies were first generated with only full-length amplicons for COI and 28S rRNA, then repeated with shorter sequences to maximize the inclusion of reference sequences from GenBank. Matches and identification at the species level (17 OTUs) were based on sequences and a preliminary analysis of skeleton and spicule composition, which matched sequences from vouchers in GenBank linked to a publication with a rigorous morphological assessment and description of the voucher. The remaining OTUs (including GenBank accession matches without taxonomic support) were identified as “sp.” since further morphological analysis are needed for accurate classification. In addition, species identification is impossible using molecular methods for polyphyletic groups (such as suborders, families and genera within Haplosclerida) without a complete morphological assessment of OTUs. However, the objective here is to determine species richness mostly based on molecular OTUs rather than a full species description of OTUs.

Diversity assessment:
R v.3.6.3 (R Core Team 2020) was used to visualize and analyze the molecular diversity assessments of sponges recruited on ARMS and reef substrates. Phylogenetic analyses of COI and 28S rRNA sequence data were used to prepare a taxonomy table (Table S5) for OTU classification (OTU) to the lowest level possible. An OTU distribution table (Table S2) specifying OTU presence/absence on either 'ARMS' or 'reef' substrate at each of the 35 sites was used to map sponge OTU richness using the 'ggmap v.3.0.0.901' package (Kahle & Wickham 2013). We used the specaccum function from the 'vegan v.2.5-6' (Oksanen et al. 2013) package to generate OTU richness rarefraction curves for comparison between the two substrates across the most specious sponge groups according to sponge class and order. Number of OTUs as a function of sites were used to generate rarefraction curves for reef substrate sponges and number of OTUs as a function of time points was used for ARMS as these were only present at one site. Venn diagrams were generated using the 'limma v.3.42.2' (Ritchie et al. 2015) package to determine the number of shared OTUs between the survey method types. Calculation of new OTU records were based on species comparisons to previous studies focused on Kāne'ohe Bay sponge collections.

- Imported original file "primary.csv" into the BCO-DMO system.
- Marked "NaN" as a missing data value (missing data are empty/blank in the final CSV file).
- Renamed fields to comply with BCO-DMO naming conventions.
- Replaced longitude value of "-157.82935C" with "-157.82935".
- Saved the final file as "986892_v1_hawaiian_sponge_cryptofauna.csv".

NSF Award Abstract:
Coral reefs are among the most species-rich ecosystems on the planet, occupying only about 1% of the seafloor, but housing more than a quarter of known marine biodiversity. Sometimes called the rainforests of the sea, coral reefs have great intrinsic biological, cultural and economic value. Nearly a billion people across the planet rely on coral reef ecosystems as a significant source of their diet, and the annual economic benefits of coral reefs are estimated to be around $9.9 Trillion USD. Thus, the global decline of coral reefs by an estimated 30-50% since the 1980s is of considerable concern as scientists struggle to understand whether species are being lost before they are even discovered. While coral reefs are spectacularly diverse, the majority of this biodiversity actually lives hidden deep within the three-dimensional framework of the reef itself. This hidden (or cryptic) community of organisms are both dramatically understudied and fundamentally important for the persistence of coral reefs. Sponges are a dominant group among these cryptic organisms within the reef which provide food from the bottom of the food chain and help sustain coral reef biodiversity. Despite the vital ecological role of sponges on coral reefs, little is known about their diversity, abundance or species ranges across the Indo-Pacific. For example, the most striking marine biodiversity gradient on the planet is described from several of the visibly dominant groups on coral reefs, including corals and reef fishes. From the global hotspot of species richness in the Indo-Pacific Coral Triangle there is a sharp eastward decline in species numbers to more remote oceanic islands in the Central Pacific, such as the Hawaiian Archipelago. However, no survey to date has evaluated whether the diversity of poorly known cryptic coral reef species, such as sponges, show the same pattern as the visible species that dominate the surface of the reef. Summer training modules introduce at-risk Pacific Islander youth to coral reef biodiversity to recruit and train a new generation of sponge taxonomists. Identification guides are being produced to help resource managers in establishing a baseline of sponge diversity, which allows resource managers to identify and protect native species, improves detection of alien species introductions and serves as a tool for monitoring changes in the ecosystem in response to human impacts. The work is being disseminated widely through scientific literature, public and professional presentations, popular press articles, and an educational display about sponges and coral reef biology in collaboration with the Waikīkī Aquarium.

This important knowledge gap is addressed by analyzing an existing backlog of standardized sampling devices (ARMS) collected from throughout the Pacific Ocean to determine whether sponges that live largely unseen within the reef framework follow the same diversity gradient as has been previously reported for fish and corals. By integrating taxonomy with multi-locus DNA barcoding and metabarcoding, this project is documenting species richness and biodiversity patterns among the cryptic sponge community across five ecoregions spanning over 10,000 km of the tropical Pacific. These collections include many new species and are providing vouchered DNA barcodes to existing reference databases that currently include fewer than 1% of sponge species across the planet. Sponges are a rich source for pharmaceutical development, so discovery of new species also provides opportunity for exploration of natural products from both the sponges and culturable microbes associated with them. By examining sponge species occurrence and diversity along both environmental and anthropogenic gradients in each ecoregion, the data also address whether coral reef sponges can serve as indicators of human impacts. Collectively, these results are transforming our knowledge of tropical Pacific sponge biodiversity, species ranges, and providing much-needed reference barcodes to global sequence databases. By determining whether sponges show the same Indo-Pacific richness gradient as reported in fishes and corals, this project is testing how well generalizations made from the visible subset of species that live on the surface of coral reefs apply to rest of coral reef biodiversity. This study is greatly advancing our knowledge of Pacific coral reef sponges and will ultimately inform the scale over which vital ecological roles performed by this understudied taxon, such as the production of nutrients at the bottom of the food chain, are acting across the Pacific.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Description from NSF award abstract:
The objective of this Research Coordination Network project is to develop an international network of researchers who use genetic methodologies to study the ecology and evolution of marine organisms in the Indo-Pacific to share data, ideas and methods. The tropical Indian and Pacific Oceans encompass the largest biogeographic region on the planet, the Indo-Pacific. It spans over half of the Earth's circumference and includes the exclusive economic zones of over 50 nations and territories. The Indo-Pacific is also home to our world's most diverse marine environments. The enormity and diversity of the Indo-Pacific poses tremendous logistical, political and financial obstacles to individual researchers and laboratories attempting to study the marine biology of the region. Genetic methods can provide invaluable information for our understanding of processes ranging from individual dispersal to the composition and assembly of entire marine communities.

The project will:
(1) assemble a unique, open access database of population genetic data and associated metadata that is compatible with the developing genomic and biological diversity standards for data archiving,
(2) facilitate open communication and collaboration among researchers from across the region through international workshops, virtual communication and a collaborative website,
(3) promote training in the use of genetic methodologies in ecology and evolution for researchers from developing countries through these same venues, and
(4) use the assembled database to address fundamental questions about the evolution of species and the reservoirs of genetic diversity in the Indo-Pacific.

The network will provide a model for international collaborative networks and genetic databasing in biodiversity research that extends beyond the results of this Research Coordination Network effort.