Anthropogenic stressors are predicted to alter biodiversity and ecosystem functioning worldwide. However, scaling up from species to ecosystem responses poses a challenge, as species and functional groups can exhibit different capacities to adapt, acclimate, and compensate under changing environments. We used a naturally acidified seagrass ecosystem (the endemic Mediterranean Posidonia oceanica) as a model system to examine how ocean acidification (OA) modifies the community structure and functioning of plant detritivores, which play vital roles in the coastal nutrient cycling and food web dynamics. In seagrass beds associated with volcanic CO2 vents (Ischia, Italy), we quantified the effects of OA on seagrass decomposition by deploying litterbags in three distinct pH zones (i.e., ambient, low, extreme low pH), which differed in the mean and variability of seawater pH. We replicated the study in two discrete vents for 117 days (litterbags sampled on day 5, 10, 28, 55, and 117). Acidification reduced seagrass detritivore richness and diversity through the loss of less abundant, pH-sensitive species but increased the abundance of the dominant detritivore (amphipod Gammarella fucicola). Such compensatory shifts in species abundance caused more than a three-fold increase in the total detritivore abundance in lower pH zones. These community changes were associated with increased consumption (52-112%) and decay of seagrass detritus (up to 67% faster decomposition rate for the slow-decaying, refractory detrital pool) under acidification. Seagrass detritus deployed in acidified zones showed increased N content and decreased C:N ratio, indicating that altered microbial activities under OA may have affected the decay process. The findings suggest that OA could restructure consumer assemblages and modify plant decomposition in blue carbon ecosystems, which may have important implications for carbon sequestration, nutrient recycling, and trophic transfer. Our study highlights the importance of within-community response variability and compensatory processes in modulating ecosystem functions under extreme global change scenarios.

Study sites and environmental parameters

We utilized two discrete CO2 vents near the Castello Aragonese islet at Ischia Island (Italy, 40°43'57" N, 13°57'52" E) (Fig. 2A). The first site was located at 2-3 m depth on the north side of the islet (referred henceforth to as North Castello), characterized by sloping rocky reefs and continuous tracts of P. oceanica beds surrounding the reef (Fig. 2B). The second site was located approximately 500 m away from North Castello and 150 m from the nearest shore at a depth of 3-6 m (referred to henceforth as Vullatura). This site was characterized by soft-bottom habitats (sandy or dead Posidonia matte) surrounded by seagrass patches growing over 2-3 m tall pinnacles and ridges of Posidonia matte (Gambi et al., 2020; Mecca et al., 2020) (Fig. 2C). At North Castello, we established three pH zones (ambient, low, and extreme low pH; depth 3-3.5 m) delineated by previous studies based on seawater carbonate chemistry analyses and in situ pH monitoring (Hall-Spencer et al., 2008; Kroeker et al., 2011). At Vullatura, we established two study zones (low and extreme low pH) where the intensity of CO2 bubbling was comparable to that of acidified zones in North Castello. The ambient zone for Vullatura was established approximately 200 m away from the other study zones, at similar depth and surroundings (5.5-6 m) (Fig. 2A). We monitored the in situ seawater pH and temperature using autonomous sensors (Satlantic seaFET^TM Ocean pH sensor). We simultaneously deployed two to three individual sensors in different pH zones and rotated these sensors around all six zones from June to October 2017.

 Effects of ocean acidification on detritivory and seagrass decomposition rate

We conducted an in situ decomposition experiment by deploying litterbags containing seagrass leaf detritus. During May 2017, live Posidonia shoots (including aged and young green leaves) were collected from 3-5 m depth at San Pietro point > 2 km away from the study vents and accessible from the Ischia Marine Center of the Stazione Zoologica Anton Dohrn (40°44'48" N, 13°56'40" E). In the laboratory, we thoroughly rinsed all collected seagrass leaves in a freshwater bath to remove excess epiphytes and epifauna. Seagrass leaves were then dried at 60°C to constant mass. We selected leaves with a minimum amount of epiphytic materials and haphazardly grouped selected leaf fragments into 10 g aliquots. We constructed litterbags (20 cm × 20 cm) out of fiberglass insect screen with 2 mm openings to minimize the loss of detritus fragments without limiting detritivore colonization. We randomly assigned litterbags into one of six groups (n = 48 per group) for deployments in each of the six study zones. All litterbags were labeled and received a clump of pre-weighed aliquots of leaf fragments.

In early June 2017, we deployed a total of 288 litterbags in the field (Fig. 2C) via SCUBA. In each zone, we chose relatively flat seafloor surfaces composed mainly of dead Posidonia matte at similar distances away from existing seagrass beds. The deployment area was directly adjacent (< 3-4 m) to the position in which pH sensors were anchored to the bottom. The water depth differed between North Castello and Vullatura, with the former vent sitting at shallower depths close to the shoreline, but within each vent site, we selected areas with similar depth at 3-3.5 m and 5.5-6 m, respectively. Prior to the deployment, we haphazardly grouped litterbags into clumps of four and firmly anchored these clumps to the bottom using aluminum and steel rods.

Litterbags were retrieved after 5, 10, 28, 55, and 117 days of deployment. On each retrieval day, we collected eight litterbags per study zone (16 litterbags per zone on the last sampling date) and carefully enclosed each litterbag in a plastic bag. All bags were placed on ice until being transported to the laboratory. In the laboratory, we gently washed all litterbags with seawater and removed seagrass detritus and animals within. Seagrass detritus was thoroughly rinsed to remove all animals, sediments, and other organic materials. On the day-55 collection, seagrass leaves began showing numerous feeding marks and a greater extent of mechanical fragmentation. To quantify the relative intensity of detritivore feeding and mechanical fragmentation, we counted the number of all seagrass leaf fragments > 1 cm in the widest length that had visible bite marks (referred to henceforth as the “Detritivory Index”). After these initial processing steps, seagrass detritus was dried at 60°C to constant mass. We analyzed a subset of dried seagrass samples (n = 8 per study zone per date for sampling day 10, 55, and 117) for organic carbon and nitrogen content using a Carlo Erba CN Elemental Analyzer. Animal components from the samples were collected using a fine sieve (400 μm) and preserved in 4% formalin.

Effects of ocean acidification on seagrass detritivore community structure

To identify and enumerate seagrass detritivores, we randomly chose four litterbag samples per study zone per sampling date (day 28, 55, and 117). Posidonia detritus is colonized by various invertebrates, but most of these species do not directly consume seagrass leaves and instead function as predators, grazers, scavengers, suspension/deposit feeders, or detritivores of algal materials (Remy et al., 2018; Gallmetzer et al., 2005). We, therefore, focused on species known to directly ingest dead seagrass leaves. We counted the gammarid amphipod Gammarella fucicola, which are highly abundant and important shredders of P. oceanica detritus (Remy et al., 2018; Lepoint et al., 2006; Wittmann et al., 1981). We also identified and counted other seagrass detritivores, including non-amphipod crustaceans (isopod Zenobiana prismatica and decapod Galathea bolivari), snails (Bittium reticulatum), and urchins (Psammechinus microtuberculatus). It has not been confirmed whether G. bolivari actively consumes seagrass detritus. However, considering the species is closely related and ecologically similar to Galathea intermedia, which occupies the same habitat (i.e., seagrass litter) and consumes Posidonia detritus (Remy et al., 2018), we classified it as a potential seagrass detritivore.