The global decline of coral reefs has led to the emergence of novel communities, including macroalgal-dominated states, that can be both prevalent and persistent, making understanding the dynamics of these emergent communities vital. In the South Pacific, emergent stands of Turbinaria ornata, a secondary foundation species, create an associational refuge that protects understory algae from herbivory. Our objective was to utilize a trait-based framework to examine intraspecific trait variation of T. ornata individuals and to explore if this variation differs across life stage and with position within a canopy-forming stand. We compared four functional traits related to the ecological strategies of resistance to herbivory and physical disturbance of large and recruit T. ornata thalli. We collected individuals in two community contexts - both within and at the edge stands (n= 22, N=88). Trait distributions varied significantly with life stage but not with community context. Small recruits were less tough, less strong, and less compact than larger canopy-forming thalli, indicating higher vulnerability to herbivory and physical disturbance. Our findings demonstrate that the associational refuge provided by large, canopy-forming individuals is likely crucial for recruit protection and therefore persistence of the stand. This refuge may ultimately contribute to the expansion and stability of this emerging secondary foundation species.

To explore shifts in T. ornata trait distributions between recruits and large canopy-forming thalli and whether these are mediated by position in a T. ornata stand (hereafter referred to as community context), we collected macroalgal individuals in a two-factor design. The first factor, life stage, included either recruits (8 cm ± 0.29 SE) or canopy-forming thalli (27 cm ± 0.62 SE) (with 44 individuals per life stage). For the second factor, community context, replicate thalli were randomly selected from either inside or the edge of 22 T. ornata stands (n=88 individuals). Inside individuals were collected from the middle of each stand, while edge individuals were collected within 1 cm of the perimeter, where the stand transitions to open reef. Notably, recruits were never found completely outside of stands. To limit environmental variability, we selected stands of similar depth (1-2 m), similar size (approximately 1-2 m diameter), and avoided damselfish territories. To ensure stands were not resampled, we collected from west to east across the reef, each collection day starting at the previous day’s endpoint. All collections were made during the dry season, June – July 2022.

We measured four traits: tensile strength, toughness, dry weight: wet weight ratio, and wet weight: height ratio. We excluded any direct measures of size, as we intentionally sampled different life stages based on size. Tensile strength imputes resistance to physical disturbance (Ryznar et al., 2021) while toughness imputes resistance to herbivory (Bittick et al., 2016). Two traits are general resistance traits with dry weight: wet weight ratio as a measure of energy allocated to physical structure (Mauffrey et al., 2020; Cosca et al., 2024) and wet weight: height as an estimate of compactness, or greater allocation to height compared to mass (Steneck & Dethier, 1994; Smith et al., 2025a).

Traits were measured within 12 hours of collection (as in Cosca et al., 2024, and Smith et al., 2025a; b). Tensile strength was quantified as the force required to break the thallus using a spring scale (N) (Ryznar et al., 2021). Toughness was measured as the weight required to penetrate the macroalgal blade with a penetrometer (±0.01 g) (Bittick et al., 2016). Wet weight was determined by placing thali in a salad spinner to remove a consistent amount of water and weighing on an Ohaus scale (±0.01 g). Thalli were dried to a constant mass in a drying oven and reweighed on an Ohaus scale (±0.01 g). Dry weight: wet weight (g/g), and wet weight: height (g/cm) were calculated by division. Height was measured from the base of the holdfast to the tip of the thallus with a ruler.  

We tested for differences among traits with a two-factor PERMANOVA (Bray-Curtis dissimilarity, 999 permutations) where factors were life stage, community context, and their interaction. Trait distributions were visualized using Principal Component Analysis (PCA) with ellipses delineating the 50% confidence interval of the standard deviation. Trait contributions to each principal component were calculated from the squared loadings and compared to the null expectation of equal contributions, calculated as 100% divided by the number of traits.  

As the PERMANOVA failed to find differences due to community context, we only compared differences due to size with univariate analysis. To evaluate differences between recruits and large thalli for individual traits, the normality of each trait was assessed using Shapiro-Wilk tests and homogeneity of variances with Levene’s test. Tensile strength met both assumptions, and means were compared with a Student’s t-test. As wet weight: height and toughness had unequal variances but normal distributions, we compared means with Welch’s t-test. Dry weight: wet weight violated both assumptions and was analyzed using Mann-Whitney rank-sum tests.