Dataset: Integrative taxonomy of introduced Haplosclerida and three new species of Haliclona sponges from Hawai'i based on samples collected from a variety of habitats on O'ahu from 2016 to 2022

Sample collection:
Sponges were photographed in situ and collected from a variety of habitats including: Autonomous Reef Monitoring Structures (ARMS), floating docks, pier pilings, marine mammal pens, and shallow (0.3 to 3 meters (m)) natural reef environments in Kāne'ohe Bay, Ke'ehi Harbor and Shark's Cove on the island of O'ahu, Hawai'i (Fig. 1). Collections from ARMS were conducted monthly throughout a 2-year sampling period (July 2016 through June 2018) following methods in (Timmers et al. 2020) and Vicente et al. (2022). Sponge species were mapped using the ggmap v.3.0.901 package (Kahle & Wickham 2013) in R v.4.3.2 (R Core Team 2023). Field observations of morphology, color, consistency, surface, and oscules for each specimen were recorded. Samples were preserved in 95% ethanol and when enough material was available, were also fixed in 4% paraformaldehyde (PFA) for 24 hours and then transferred to 70% ethanol. Type and other specimens were deposited in the Florida Museum of Natural History (catalogue numbers beginning with acronym UF) in Florida, USA, and the Bernice Pauahi Bishop Museum (catalogue numbers beginning with acronym BPBM) in O'ahu, USA. Samples from Kāneʻohe Bay were collected under special activities collection permits SAP2018-03 and SAP2019-06 (covering the period of January 13, 2017, through April 10, 2019) issued by the State of Hawai'i Division of Aquatic Resources. Samples from Pūpūkea Marine Life Conservation District were collected under permit SAP2023-20 (covering the period of March 11, 2022, through March 10, 2023). Samples from Kewalo and Ke'ehi harbors were collected by the Division of Aquatic Resources Aquatic Invasive Species team, who do not require special activities permits.

DNA extraction, sequencing, and assembly:
Subsamples of sponge tissue (30 milligrams (mg)) were removed from type material preserved in 95% ethanol and were processed for DNA extraction. Two different approaches were utilized for DNA extraction and downstream analysis. First, we followed protocols in Vicente et al. (2022) for DNA extractions, polymerase chain reactions (PCR) for amplifying partial fragments of both 28S rRNA and COI genes and for Sanger sequencing. Forward and reverse reads were assembled, trimmed, and edited by eye using Geneious 10 (Kearse et al. 2012). Sequences were checked for contamination using the online BLAST server (Altschul et al. 1990) and results that showed at least 85% sequence identity to sponges were used for subsequent analysis. All assembled chromatograms resulted in >90% high-quality base pair reads with a mean Phred quality score 40. Second, total DNA from holotypes of Haliclona (Gellius) loe BPBM C1523, Haliclona (Reniera) kahoe BPBM C1539, Haliclona (Rhizoniera) pahua BPBM C1518, vouchers BPBM C1519, BPBM C1510 for Haliclona (Soestella) caerulea and Gelliodes conulosa, respectively, were extracted with a phenol-chloroform method modified from Saghai-Maroof et al. (1984) and used directly for library preparation with the Illumina True-Seq PCR free kit and low-coverage of whole genome sequencing on Illumina NovaSeq 6000 at the Iowa State University DNA facility. A detailed protocol for mitochondrial and ribosomal genome sequences for these samples can be found in Lavrov et al. (2025).

Phylogenetic analysis:
Haplosclerida sequences within GenBank closely resembling (>90% sequence identity) new sequences in our study were selected for the phylogenetic analysis. These only included sequences from species associated with voucher specimens associated with a peer-reviewed publication authored or coauthored by taxonomists. ClustalW with default parameters was used for aligning partial and complete 28S rRNA and COI sequences. Alignments consisted of 300 bp of the 28S and 480 bp of the COI gene sequence. RaxML (Stamatakis 2006) included Geneious 10 was used for maximum likelihood (ML) analysis with the GTR+GAMMA model of nucleotide substitution, 100 starting maximum parsimony trees, and 1,000 bootstrap replicates. Resulting bootstrap values of >50 from the ML posterior probabilities are shown on the tree. Phylogenetic trees were rooted on Ephydatia fluviatilis OX175335.1 and ON000190.1 for 28S and COI, respectively. Sequences of holotypes and other specimens for each species were deposited to GenBank under accession numbers: MW016123, MT452542, MT586742, MW016124, MW059074, MW143255, MW016168, MW059064, MW059059, MW016133, MT586743, MW016360, MW016153, MW016155, MW016154, MW059075, MW016154. All accession numbers pertaining to each species are also available in Table S1.

Sectioning and spicule preparation:
Sponge pieces (3 to 5 cubic millimeters) containing both ectosomal and choanosomal tissue fixed in either 4% PFA or 95% ethanol were transferred to 70% ethanol. Sponge pieces were dehydrated in an alcohol series of 35%, 50% and 70%, 100% and embedded in paraffin. Sections >100 micrometers (μm) thick were cut perpendicular to the surface through the ectosome and choanosome with a microtome. In specimens where the ectosome was specialized, tangential sections across the sponge surface were made at 100 μm thickness. Small pieces were also boiled in nitric acid for 1 to 2 minutes or until solution turned clear. Spicules were let to settle, and the acid was discarded. Spicules were then rinsed two times with distilled water to remove the acid; water was then changed to 95% ethanol for storage. Spicules were suspended and a few drops were observed under light microscopy, photographed, and measured using ImageJ (Abràmofff et al. 2005) http://imagej.nih.gov/ij/). Fifty oxeas and, if present, ten sigmas per species were measured [lengths and widths, expressed herein as minimum–mean [±1 standard deviation (SD)]–maximum length / width in μm (n)]. A few drops of the spicule suspension were added to a stub, air dried, and imaged under a Hitachi S-4800 FESEM Scanning Electron Microscope (SEM) at the Biological Electron Microscope Facility at the University of Hawai'i Mānoa.

Summarizing morphological characters of congeneric comparative material:
Sponge species found throughout the Hawaiian Archipelago are shared with the Central Indo-Pacific (Australia, Philippines, the Mariana Archipelago), Temperate Australasia (New Zealand), Temperate Northern Pacific (Japan), the Eastern Mexican Pacific, Caribbean, Mediterranean and the Northeast Atlantic (Bergquist 1967; Carballo et al. 2013; de Laubenfels 1950; van Soest et al. 2021). We used the World Porifera database to include a summary of morphological characters for 51 Haliclona spp. closely resembling the new species in this study. These species were selected from geographic locations with shared species with the Hawaiian Archipelago. Species from the Temperate Atlantic, Black Sea, Arctic, or Southern Ocean were considered improbable species due to geographic barriers and temperate climates. Species from these ecoregions were therefore disregarded as comparative material.