Dataset: Coral turf allelopathy experiments

We utilized turfs from territories of Stegastes nigricans and Stegastes punctatus to assess: i) turf effects on corals, ii) differential impacts based on turf species composition, and iii) variance in effects on the coral Acropora pulchra versus Porites rus. Species composition of turfs in territories of S. nigricans versus S. punctatus was assessed by collecting turf haphazardly from territories (n = 12) of each damselfish species spread across an approximately 0.07 km² region in both March and September 2022. Samples from S. nigricans territories were collected using a 1.3 cm grommet punch and from S. punctatus territories by cutting a similarly sized area from territories on A. pulchra bases. Turfs were stored in 4% formalin in seawater, and species composition quantified following methods of Diaz-Pulido and McCook (2002). See Related Datasets section below for dataset 949552 on turf cover species composition.

In March 2022 (mean water temp = 28.6°C; Leichter et al. 2023), we tested the impacts of the macroalgae Sargassum pacificum and Amansia rhodantha, turfs from S. nigricans or S. punctatus territories, plastic macroalgal mimics, turf mimics, and no contact controls on the common corals A. pulchra and P. rus (n = 12 for each coral x treatment). Macroalgae were chosen due to their high abundance and differing impacts on corals (Rasher and Hay 2010; Longo and Hay 2017; Clements et al. 2020).

For our assays, seven 6 to 8 cm length coral fragments were collected from 12 individuals of both A. pulchra and P. rus in the study area. The fragments were attached with epoxy into the cut-off necks of plastic bottles (Z-spar Splash Zone Compound, Kopcoat, Pittsburgh, Pennsylvania, USA – see Clements et al. 2020), and attached to inverted bottle caps secured to a metal rack ~20 cm above the substrate to limit abrasion from sediments (Figure S1, Altman-Kurosaki et al., 2024). Treatments were assigned using a randomized block design such that each block (coral colony) had one algal treatment affixed to a different fragment from that colony, thus minimizing potential variance associated with individual colonies.

Algal treatments and mimics were applied to corals by zip-tying algae or controls directly to the corals. Turf treatments from S. nigricans and S. punctatus territories consisted of ~1.3 cm² pieces of substrate with natural turfs attached. Turf mimics (controls for effects of substrate) were substrates of similar size, sterilized in bleach, rinsed thoroughly with filtered seawater, and dried at 60°C for three days prior to use in the field. Macroalgal treatments were thalli ~6 cm in length while controls for the physical presence alone were similar sized plastic aquarium plants resembling Sargassum. After 13 days, treatment effects on corals were assessed via in situ pulse-amplitude-modulated (PAM) fluorometry directly at the site of coral-algal contact. PAM fluorometry quantifies the photosynthetic efficiency of coral endosymbionts (Y, or effective quantum yield) which is a correlate of coral fitness. PAM readings were conducted between 0900–1400 hours and were blocked in time, with each block consisting of a random replicate from each coral x algal treatment, and with readings for each coral x algal treatment within a block randomly interspersed through time. Readings took two days, with half of the blocks being assessed each day. PAM fluorometer readings did not show any significant differences based on the date of assessment, thus allowing the measurements from the two separate days to be combined for further analyses

In March 2022, we tested for algal allelopathy by extracting the surface chemicals from both of the turf communities using methods of De Nys et al. (1998). Twenty milliliter volumes of turf algae collected from S. nigricans or S. punctatus territories were spun for 30 seconds in a mesh bag and blotted with paper towels to remove excess water. Algae were then vortexed for 30 seconds in 50 mL of hexane, the hexane decanted, and then evaporated under a fume hood for 8.5 hours, leaving the lipid soluble surface extracts. Surface extracts from each turf type were separately resuspended in 1 mL of methanol before being added to phytagel strips following the methods outlined in Rasher and Hay (2010). Control strips were made with 500 mL of methanol but no algal extract. Strips were refrigerated for less than 12 hours prior to deployment in the field.

Effects of extracts were assessed in situ by zip-tying phytagel strips directly to fragments of A. pulchra and P. rus for 24 hours in the field. New coral fragments were collected, placed on racks, and blocked by colony (as above) so that a control strip and strips with extracts from each turf type were blocked by individual coral colony for each of the two coral species (n = 12 blocks per treatment). Treatment effects on photosynthetic efficiency were measured after 24 hours using PAM fluorometry, with assessments conducted between 0900–1400 hr. PAM readings were compared between coral x algal treatments using beta regression with coral colony ID as a random effect. 

We hypothesized that seasonal outcomes of coral-algal interactions could be due to seasonal variance in the production of allelochemicals by algae or resistance to those chemicals by corals. Given this, the surface extract assay was repeated with newly collected coral fragments in austral winter from September 25–28, 2022, when temperatures averaged 26.3°C (i.e., 2.3°C cooler than during the austral summer experiments in March). To account for differences in turf allelopathy due to seasonal changes, we tested the allelopathy of turf extracts using from turf that was freshly collected in September (austral winter) versus extracts from turf that had been collected in March (austral summer) and preserved at -80°C. If extracts from turf that were collected in March had a reduced effect in September and/or if extracts from turf that was collected in September had a lower impact on corals relative to the March turf, then this could suggest that allelopathic effects of turfs on corals are reduced in winter months. PAM readings were compared between coral species x treatments using beta regression with coral colony ID as a random effect.