Harnessing solar power: photoautotrophy supplements the diet of a low-light dwelling sponge

Mixotrophy, the combination of autotrophy and heterotrophy, is widespread in marine systems and enables organisms to buffer against resource fluctuations. Many sponges harbor photosynthetic symbionts (especially cyanobacteria), yet most work has examined species in high-light environments. The contribution of autotrophy to the carbon budget of net heterotrophic, low-light dwelling sponges remains poorly quantified, and photosynthesis/respiration alone cannot determine nutrient transfer to the host. This study investigates the extent to which cyanobacterial photoautotrophy supplements the diet and respiratory demand of the Caribbean sponge Chondrilla caribensis at ~20 m depth and tests whether cyanobacteria-derived carbon and nitrogen are translocated to host cells.

Prior studies show some shallow, high-light sponges can derive up to ~80% of carbon needs from photosymbionts and be net phototrophic (P:R > 1), often with flattened morphologies to maximize light. However, cyanobacteria are common across many sponge species, including those in shaded cryptic habitats where light is low. Nutrient transfer is variable among sponge-symbiont systems; in some, symbionts enhance host growth and translocate photosynthates, while in others they do not. Stable isotope probing and NanoSIMS allow tracing of nutrient transfer at cellular resolution and have revealed metabolic integration in other photosymbiotic sponges. Quantifying both autotrophic and heterotrophic contributions to sponges’ carbon budgets has seldom been done, particularly in low-light habitats.

Location and specimens: Work was conducted at CARMABI, Curaçao (Oct–Dec 2019). Chondrilla caribensis forma hermatypica individuals were collected at 18–22 m depth (house reef), cleaned, sized to ~13 ± 4 cm² with ≥5 oscula, and acclimated 1–2 weeks at 15 m in cages under a light profile mimicking vertical surfaces at ~20 m. Healthy pumping individuals were used. Additional n = 3 specimens were used for microbiome profiling (16S rRNA gene). Photosynthesis–irradiance (P–E): Net O2 fluxes were measured ex situ (n = 4) in 2-L airtight chambers across 7 irradiance levels from ~0 to ~800 µmol photons m⁻² s⁻¹ using natural sunlight, achieved by removing layers of shade cloth and PVC covers. Net O2 rates were computed from linear slopes and expressed per cm² per hour. An exponential P–E model was fitted: P_E = P_max (1 − e^(−E/E_k)) + R_d, where R_d (dark respiration) is negative. Photokinetic parameters are in Supplementary Table 2. Daily photoautotrophy and heterotrophy: A daily vertical-surface light profile at 20 m (model adjusted with in situ irradiances) was integrated against individual P–E curves to derive hourly gross photosynthesis (P_o) and summed to daily P_o. Daily net primary productivity P = P_o + R, with R = 24 × R_d. Oxygen-to-carbon conversions used photosynthetic quotient = 1.1 and respiratory quotient = 0.75. Rates expressed as µmol C cm⁻² d⁻¹. A note that using daytime-measured R_d could underestimate daily P; separate day vs night dark incubations indicated slightly lower night respiration (−1.15 ± 0.52 vs −0.94 ± 0.50 µmol O2 cm⁻² h⁻¹). In situ natural diet incubations: Net heterotrophic uptake was measured in situ at 10 m near Buoy 1 between 10:00–12:00 over one week (n = 6 light; n = 6 dark). Sponges were in 2-L airtight chambers; water sampled at t = 0, 5, 10, 20, 40 min for DOC and live POC (LPOC). Light chambers were wrapped with shade to mimic maximal vertical-surface irradiance at 20 m (~49 ± 26 µmol photons m⁻² s⁻¹); dark chambers were opaque. Seawater-only controls (n = 3 per light condition) were run daily. DOC uptake was fit using a bi-exponential 2G model (labile + refractory fractions); LPOC removal assumed exponential clearance in a well-mixed system. Where models/regressions were not significant, uptake was set to zero. Hourly fluxes (dark and light) were each multiplied by 12 and summed to daily DOC_24 and LPOC_24. Total daily carbon uptake (mixotrophic diet) = P_o + DOC_24 + LPOC_24; SDs propagated by root-sum-of-squares. Pulse–chase stable isotope labelling: To visualize inorganic carbon and nitrogen assimilation and transfer, sponges were pulsed 6 h (starting 06:00) in 3-L airtight, stirred chambers with artificial seawater enriched to 2 mM 13C-NaHCO3 and 5 µM 15NH4Cl under low natural light (~50 µmol photons m⁻² s⁻¹ at midday). A dark pulse control was also run (6 h). After the pulse, sponges were chased 42 h in flow-through aquaria with ambient seawater. Time points: t = 0 (unlabelled), 6 h (light), 6 h (dark), 48 h (end chase); n = 4 replicates per time point. Tissue subsamples were fixed (glutaraldehyde/paraformaldehyde in PHEM buffer with sucrose) for EM/NanoSIMS; remaining tissue was stored for bulk isotope analysis. Microscopy and NanoSIMS: Fixed tissue was embedded, semithin 500 nm sections cut perpendicular to surface, placed on Si wafers, stained, and imaged by SEM (Zeiss Sigma FE-SEM, 8 kV). Regions of interest were mapped for NanoSIMS (CAMECA NanoSIMS 50). Four to eight areas per sample (typical raster 40 × 40 µm, 512 × 512 px). ROIs targeted cortex and inner sponge body: cyanobacteria, other symbiotic microorganisms, and host cells (choanocytes excluded). ROIs were delineated on 12C14N images with 31P and SEM references. Isotopic ratios were corrected using daily yeast standards and reported in delta notation relative to unlabelled controls. Enriched cells exceeded 3× SD of control ROIs. A total of 9421 cells were analyzed. Statistics: Bulk and NanoSIMS data analyzed via PERMANOVA (Primer v7 with PERMANOVA+), Euclidean distances, Type III SS under reduced model (4999 permutations). NanoSIMS biological replicates pooled per time point per ROI for analysis; pairwise tests evaluated significant differences. P–E fitting in SigmaPlot 14.5; daily irradiance integration in R 4.0.3; uptake models in Excel/SPSS; differences between light and dark diet incubations via two-tailed paired t-tests (p = 0.05).

• Photosynthesis–irradiance: C. caribensis showed a typical photosymbiont-bearing P–E response with steep initial slopes and low saturating irradiance (E_k = 57 ± 28 µmol photons m⁻² s⁻¹; model fits R² ≥ 0.907), indicating efficiency at low light.
• Energetics at 20 m: Daily P:R = 0.35 ± 0.08 (net heterotrophic). Gross primary productivity was 9.4 ± 4.1 µmol C cm⁻² d⁻¹; daily respiratory demand 18.3 ± 6.8 µmol C cm⁻² d⁻¹. Photosynthetic carbon potentially supplied 52 ± 11% of respiratory demand. Net primary productivity was negative (−8.8 ± 4.4 µmol C cm⁻² d⁻¹; −392 ± 195 µmol C g DW⁻¹ d⁻¹).
• Heterotrophic diet: DOC dominated heterotrophic uptake at 127 ± 63 µmol C cm⁻² d⁻¹ (98% of heterotrophic organic C), while LPOC contributed ~2% (light vs dark differences not significant). Total daily carbon uptake combining photoautotrophy and heterotrophy was 139 ± 63 µmol C cm⁻² d⁻¹, with photoautotrophy contributing 7 ± 3% of total uptake.
• Microbiome: 16S rRNA profiling showed dominance of Chloroflexi (37%), Gammaproteobacteria (16%), Alphaproteobacteria (9%); the cyanobacterium Candidatus Synechococcus spongiarum comprised ~6% of the community. Cyanobacteria were more abundant and smaller in the cortex (0.84 ± 0.16 µm width) than inner body (1.28 ± 0.13 µm).
• Single-cell isotope tracing: After a 6 h 13C-bicarbonate pulse, >99% of cyanobacteria were 13C-enriched; bulk tissue 13C-enrichment increased significantly. Host cells in both cortex and inner body became 13C-enriched during the light pulse, evidencing translocation of photosynthates. A fraction of other symbionts assimilated 13C (43% cortex; 9% inner body), but to much lower levels than cyanobacteria. Cyanobacterial 13C-enrichment persisted through 48 h chase; host and other symbiont 13C increased further between 6 and 48 h, indicating continued transfer. Phagocytosis of enriched cyanobacteria by host cells was observed, but only ~21% of 13C-enriched host cells contained visible cyanobacteria, implying additional transfer via released photosynthates.
• Dark controls: Holobiont-level dark 13C fixation occurred but was on average 4.4× lower than light; at single-cell level, 13C-enrichment did not significantly exceed background for any cell type in the dark pulse; no host cells enriched in the dark.
• Ammonium assimilation (15N): >90% of cyanobacteria and other symbionts, and >97% of host cells became 15N-enriched within 6 h, with symbiotic microorganisms showing higher enrichment than cyanobacteria. 15N assimilation did not differ between light and dark, indicating microbially mediated ammonium assimilation not dependent on light. Some enriched symbionts were observed inside host cells, indicating transfer of microbially assimilated nitrogen.

This study demonstrates that even in low-light habitats (~20 m, vertical surfaces), photoautotrophy by cyanobacterial symbionts can substantially supplement a net heterotrophic sponge’s metabolism, potentially covering about half of daily respiratory carbon needs. However, heterotrophy—especially DOC consumption—dominates the total carbon budget, consistent with mixotrophs favoring one pathway and with trade-offs in maintaining dual nutritional modes. Single-cell analyses confirm metabolic integration: photosynthetically fixed carbon by cyanobacteria is transferred to host cells via at least two pathways—phagocytosis of symbionts and uptake of released photosynthates—with the latter likely prevalent given many enriched host cells lacked engulfed cyanobacteria. The persistence of high cyanobacterial 13C through 48 h and relatively modest host enrichment suggest slower or limited translocation compared with Symbiodiniaceae-based symbioses in shallow corals/sponges, potentially reflecting low-light conditions and cyanobacterial storage dynamics. Ammonium assimilation was widespread across the microbiome and host, largely light-independent, highlighting the broader microbial community’s role in nitrogen cycling and its contribution to host nutrition. Overall, the findings address the knowledge gap for low-light, net heterotrophic, photosymbiont-bearing sponges and indicate that mixotrophy provides resilience against variable food supply and may influence reef biogeochemistry.

Photoautotrophy measurably supplements the diet and respiratory demand of the low-light dwelling sponge Chondrilla caribensis, despite overall net heterotrophy. Quantifying both autotrophic and heterotrophic carbon fluxes reveals that cyanobacteria-derived carbon can support up to ~52% of respiratory needs and ~7% of total daily carbon uptake at 20 m. Single-cell imaging confirms translocation of photosynthates to host cells. The study advocates comprehensive assessment of mixotrophy (autotrophy plus heterotrophy) across sponge species and environments to understand trophic plasticity, ecosystem productivity, and biogeochemical impacts. Future work should determine prevalence and magnitude of mixotrophy across light regimes and test environmental drivers (irradiance, temperature, food availability) on the balance of nutritional modes.

• Respiration used to compute daily net primary productivity was measured as dark respiration during the day; this may underestimate daily net P due to diel variation (slightly lower rates at night were observed).
• In situ heterotrophic uptake rates showed high variability; some replicates exhibited no net DOC or LPOC uptake, potentially reflecting analytical sensitivity, variable ambient food quantity/quality, pumping variability, or physiological state.
• Diet incubations were conducted over limited time windows (midday) and at 10 m depth with shaded/light treatments to mimic 20 m vertical-light conditions; full diel and seasonal variability was not captured.
• NanoSIMS provides 2D snapshots; some engulfed symbionts may not be visible in plane of section, and quantifying absolute fluxes of translocation was beyond scope.
• Study focused on a single species and depth/light environment; results may not generalize across morphotypes, locations, or higher-light habitats where P:R may differ.