Silicon uptake by sponges: a twist to understanding nutrient cycling on continental margins

Abstract

About 75% of extant sponge species use dissolved silicon (DSi) to build a siliceous skeleton. We show that silicon (Si) uptake by sublittoral Axinella demosponges follows an enzymatic kinetics. Interestingly, maximum uptake efficiency occurs at experimental DSi concentrations two orders of magnitude higher than those in the sponge habitats, being unachievable in coastal waters of modern oceans. Such uptake performance appears to be rooted in a former condition suitable to operate at the seemingly high DSi values characterizing the pre-Tertiary (>65 mya) habitats where this sponge lineage diversified. Persistence of ancestral uptake systems causes sponges to be outcompeted by the more efficient uptake of diatoms at the low ambient DSi levels characterizing Recent oceans. Yet, we show that sublittoral sponges consume substantial coastal DSi (0.01-0.90 mmol Si m(-2) day(-1)) at the expenses of the primary-production circuit. Neglect of that consumption hampers accurate understanding of Si cycling on continental margins.

Figure 1. Relationship between DSi uptake and DSi availability during initiation of experiment I.

The uptake-rate response by the sponges (µmol Si per h and ml of sponge) was linearly related to DSi concentration in the experimental bottles, whenever DSi availability ranged from natural values (1.6 µM) to 200 µM. Crosses are 3 further treatment steps (20, 70 and 100 µM DSi) conducted as a test after allowing sponges to rest for 5 days at natural DSi concentration. Note that all 3 test responses fell within the 95% prediction interval of the previously calculated regression equation.

Figure 2. Summary of sponge uptake responses during experiments I and II.

The solid line indicates the course of average uptake rates (µmol Si per h and ml of sponge ± s.d.) by "healthy" (i.e., non-fluorine poisoned) sponge sets in response to experimental DSi concentrations during the first phase of the hexafluorosilicate-based experiment I (yellow circles) and through the metasilicate-based experiment II (blue triangles). Crosses are 3 treatment steps (20, 70 and 100 µM DSi) conducted as a test after allowing sponges to rest for 5 days at natural DSi concentration prior to initiating the second phase of experiment I. The dashed line indicates average (± s.d.) uptake rates during the second phase of experiment I (i.e., DSi >200 µM), in which sponges became poisoned by high concentrations of fluorine released from sodium hexafluorosilicate.

Figure 3. Summary of individual uptake data.

Some variability was noticed in the individual uptake responses during the "healthy" phase of experiment I (a) and through experiment II (b). Green and blue lines indicate small (< 9ml) and large (> 9 ml) individuals of Axinella damicornis, respectively. Red lines indicate individuals of Axinella verrucosa or Axinella polypoides. Note that the sponge individuals used for experiments plotted in "a" and "b" graphs are different.

Figure 4. Relationship between sponge size and uptake.

A weak negative relationship between uptake rate (µmol Si h⁻¹ ml⁻¹) at 200 µM DSi and sponge size (ml) was detected, grossly fitting an inverted, first order, polynomial regression. The highest uptake rate corresponded to individual 6 of Exp. I.

Figure 5. DSi uptake model for Axinella.

Michaelis-Menten kinetics function modeling the relationship between Axinella spp. average uptake rate and ambient DSi in the experimental bottles, obtained after pooling data of experiments I and II.