Near-daily reconstruction of tropical intertidal limpet life-history using secondary-ion mass spectrometry

The study addresses the challenge of monitoring climate responses and life-history traits of intertidal organisms in wave- and tide-exposed tropical rocky shores. While molluscan shells serve as high-resolution archives elsewhere, a suitable high-resolution proxy for tropical intertidal climate and corresponding limpet life-history has been lacking due to complex hydrology and environmental variability in low latitudes. The authors aim to reconstruct near-daily environmental signals and life-history of the Hawaiian limpet Cellana sandwicensis by integrating oxygen isotope analysis via secondary ion mass spectrometry (SIMS) with sclerochronology, enabling seasonal growth interpretation and longevity estimation under tropical intertidal conditions.

Previous work shows mollusc shells can archive seasonal to millennial SST variations, with robust δ18O–temperature correlations at mid-to-high latitudes and in various subtidal organisms (e.g., corals, foraminifera, bivalves). However, tropical intertidal proxies are underdeveloped due to confounding salinity effects and hydrological processes. Limpet sclerochronology has been used to reconstruct temperatures and foraging seasonality in temperate systems, but tropical intertidal applications, particularly for Hawaiian Cellana spp., remain scarce. Existing studies document Cellana ecology, reproduction, and some growth patterns, but detailed, high-resolution growth and longevity estimates, especially for large individuals, are limited.

Study species and site: The yellowfoot limpet (Cellana sandwicensis) inhabits Hawaiian rocky intertidal zones. Reproductive cycles are primarily December–March with secondary activity June–August. The study site for modern specimens was Ka’alawai, O‘ahu, characterized by mixed semi-diurnal tides, trade winds, and variable wave exposure. Mean surface salinity at Ka’alawai has been reported as 25.4‰ due to submarine groundwater discharge. Specimens: Two modern live specimens (CW1, CW2) were collected June 28, 2018, at Ka’alawai. One historical museum specimen (BPBM 250851-200492) of unknown exact origin was included due to large size. Shell preparation and microstructure: Shells were cross-sectioned along the growth axis, mounted, ground, and polished. Microstructure was characterized by SEM and Raman spectroscopy, identifying layers relative to the myostracum (M): an interior aragonitic radial crossed-lamella (M−1), exterior aragonitic concentric crossed-lamellar (M+1), calcitic concentric crossed-foliated (M+2), and an outer radial crossed-foliated layer (M+3, polymorph unconfirmed). Isotope sampling targeted the calcitic M+2 layer to avoid mixing polymorphs. SIMS δ18O analysis: Polished, carbon-coated sections were analyzed with a CAMECA IMS-1280 using a Cs+ primary beam (2.5 nA; 120 s presputter). Analyses rastered a 15 µm² area (capturing 1–3 daily growth increments) with 30 cycles of 10 s integration. 16O and 18O were measured in multicollection Faraday cups. Instrumental mass fractionation was corrected using UWC-3 calcite standard measured before and after each session (n=12), with reproducibility 0.17–0.35‰ (2σ). Measurement uncertainty reported as 2σ, incorporating precision and reproducibility. Sampling spacing: CW1 mean 252 µm (55 points), CW2 288 µm (43 points) to achieve sub-weekly resolution over one annual cycle; BPBM 554 µm (56 points) across four annual cycles for sub-annual resolution. Predicted δ18O and climate reconstruction: Predicted δ18Ocalcite was computed from in situ SST (PacIOOS NS04, Waikiki Aquarium) and estimated δ18Oseawater via a salinity–δ18O relationship for the Tropical Pacific (δ18Oseawater = 0.201×SSS − 8.88). Equilibrium fractionation used Friedman & O’Neil (1977): 1000 ln α = 2.78×10^6/T^2 − 2.89; with VPDB–VSMOW conversion (Coplen et al., 1983). For the historical shell, reconstructed SSTs were modeled across salinities 24–42 psu to identify biologically plausible ranges; temperature errors were ±1.54 °C (CW1), ±1.53 °C (CW2), and ±1.60 °C (BPBM). Temporal alignment and growth features: After SIMS, carbon coats were removed and sections stained with Mutvei’s solution to reveal major annual lines, minor (circalunidian) lines, and micro-growth increments (circatidal). SEM and light micrographs were overlaid to align SIMS spots with growth features. Calendar dates between major growth lines were assigned by counting daily micro lines and using δ18O minima/maxima as anchors. Growth measurements and modeling: Daily growth was measured between micro-increments along the growth axis and horizontally (anterior–posterior) using ImageJ. Back-calculated shell lengths informed von Bertalanffy growth function (VBGF) fits: L(t)=L∞(1−e^(−K(t−t0))) with settlement length L0=0.254 mm (t0=−0.09026). Pairwise likelihood ratio tests assessed parameter differences among shells, with data truncated to 0–45 mm to standardize overlap. Statistical analyses were conducted in SAS (correlations, repeated-measures ANOVA) and R/Excel for growth modeling.

- SIMS δ18O ranges (VPDB-corrected) in calcite M+2 layer: CW1 −5.04‰ to 7.74‰; CW2 −4.38‰ to 7.83‰; BPBM −0.57‰ to 5.02‰. Maximum δ18Ocalcite uncertainty: 0.51‰ (2σ). - Historical specimen (BPBM) exhibited sinusoidal multi-annual δ18O cycles consistent with seasonality and major annual growth lines aligned with most positive δ18O (cool season) and, once, with a most negative value. - Modern specimens: strong correlations between measured and calculated δ18Ocalcite: CW1 R²=0.71, p<0.0001 (slope 0.39); CW2 R²=0.69, p<0.0001 (slope 0.37). - Reconstructed SST from BPBM across salinities indicated best biological plausibility at SSS ≈ 42 psu, yielding SST range 15.5–38.0 °C, mean 27.1 ± 7.38 °C; sensitivity ≈ 0.84 ± 0.04 °C per psu. - Growth features: micro-increment widths 6.32–61.74 µm; averages: CW1 22.54 µm, CW2 20.67 µm, BPBM 13.69 µm. Micro lines recorded lunar/tidal cycles; spring tides produced narrow, prominent daily lines; neap tides produced wider, faint lines. - Ages estimated from isotope cycles and early ontogeny: CW1 and CW2 ≈ 2 years; BPBM ≈ 5 years. - Modern 2017 sub-monthly growth: initial shell lengths 20.75 mm (CW1), 22.89 mm (CW2). Daily growth (DG_SL): −83 to 588 µm; means 140 ± 93 µm (CW1) and 98 ± 95 µm (CW2). Monthly growth 0–5.39 mm; total 2017 growth 19.96 mm (CW1) and 15.01 mm (CW2). Monthly growth correlated with growth frequency (CW1 R²=0.81, p<0.0001; CW2 R²=0.67, p=0.0005) and DG_SL (CW1 R²=0.61, p=0.0017; CW2 R²=0.89, p<0.0001). Growth cessations occurred in December/March (primary spawning) and summer months (secondary spawning), coinciding with extreme δ18O values. - Historical BPBM annual growth declined from 23.22 mm (year 1) to 4.50 mm (year 5); mean daily growth 81.78 ± 120.55 µm; maximum daily 605 µm. - VBGF pooled fit: L∞=75.233 mm (±2.681 SE), K=0.0371 yr⁻¹ (±0.0021 SE), with no significant parameter differences among shells (pairwise likelihood ratio tests, p>0.05). Modeled lengths: 27.2 mm (1 yr), 44.5 mm (2 yr), 55.5 mm (3 yr), 62.6 mm (4 yr), 67.1 mm (5 yr).

The study demonstrates that SIMS δ18O analysis of Cellana sandwicensis shells, aligned with sclerochronological features, yields sub-weekly to near-daily resolution of environmental seasonality and growth history in a tropical rocky intertidal setting. Strong linear relationships between calculated and measured δ18O support the use of these shells for seasonal interpretations of growth and climate. However, full isotopic equilibrium with seawater remains to be validated and may be influenced by physiological vital effects, extrapallial fluid chemistry, and microhabitat (aerial exposure) differences. Modeling of the historical shell suggests evaporation-elevated salinity (~42 psu) in more aerially exposed microhabitats, with SST ranges consistent with known substratum temperatures (15–40 °C). The sclerochronological patterns capture tidal and lunar cycles, with wave exposure modulating line visibility and width. Growth dynamics show determinate growth, seasonal slowdown and cessation associated with spawning and thermal/desiccation stress, and intra-population variability likely due to genetics and microhabitat/food availability. Age at maturity was inferred at 8–9 months (~21 mm), later than previously reported, refining demographic understanding. Longevity up to 5 years places C. sandwicensis among short-lived Cellana species. These findings provide a foundation for tropical intertidal paleoecology and inform aquaculture, conservation, and fishery management of a culturally important species.

By coupling SIMS-derived δ18O with detailed sclerochronology, the study provides the first near-daily reconstruction of tropical intertidal limpet life-history and a robust method to infer seasonal growth, maturity, and longevity in Cellana sandwicensis. It establishes a pathway to reconstruct tropical intertidal climatology from molluscan archives, with applicability to archaeological records. The work refines growth parameters (VBGF), identifies thermal thresholds, and links growth cessations to spawning periods. Future research should validate isotopic equilibrium by measuring δ18Oseawater alongside shell δ18O using both SIMS and conventional GSMS, expand sampling across microhabitats and islands, quantify vital effects, and apply the approach to archaeological middens to reconstruct past intertidal climates and inform adaptive management of limpet fisheries.

- δ18Oseawater was not measured directly for modern analyses; δ18O–SSS relationships introduce potential error (~±1.5 psu variability in the Tropical Pacific). - The exact collection site/microhabitat of the historical specimen is unknown, adding uncertainty to reconstructed SSS/SST. - Potential vital effects (physiology/metabolism) and extrapallial fluid chemistry may bias isotopic equilibrium assumptions. - Microhabitat heterogeneity (evaporation, irradiance, aerial exposure) can drive intraspecific δ18O variance. - Limited sample size (three shells) and single modern site constrain generalizability. - SIMS instrumental drift was only minimally monitored between standard groups; additional standardization could improve precision.