The Importance of Water Flow for Culture of Dysidea avara Sponges

The principal aim of this research was to understand the optimal water flow requirements for the marine sponge Dysidea avara, so that flow parameters could be set for ex situ culture in tanks, or in situ culture in the sea. The ultimate goal was to help insure the success of large-scale aquaculture of this sponge for obtaining biomass for extraction of its valuable terpenoid natural product avarol.

Process development and economics of aquaculture for natural product chemicals
In Chapter 2, results from two successful in-sea aquaculture feasibility projects and one partially successful ex situ tank culture project were presented. The steps involved in full process development for each of the three examples (a bryozoan, an ascidian, and a sponge) and their attendant costs and economics were detailed. Results showed that in-sea
aquaculture for the bryozoan and the ascidian for their anti-cancer natural products could be
profitable at the projected market prices for the drugs. However, for the case history example
presented for sponge culture in controlled environment tanks, the results were less positive or
conclusive. Slow growth and low chemical yields, followed by the drug being dropped by the
pharmaceutical company sponsor, left aquaculture of the source sponge redundant.

In-sea investigations of ambient flow regimes – Chapter 3
From the results of our field investigations on the N.E. coast of Spain, we learned that Dysidea avara inhabits primarily low flow habitats protected by rock cover or depth. Proximal flow speeds recorded within 2-4 cm of specimens at three depths were mostly oscillatory in form
and averaged only 1.6 cm/s in calm seas, 5.9 cm/s during infrequent storms, and 2.6 cm/s
over all seasons, depths and sea conditions.
Sponge morphologies at the three depths differed. Near-surface (4.5 m) forms were
globular, without tall processes – but with large oscula. Mid-depth (8.8 m) sponges had
tall, thick protuberances each topped with a large osculum. Deeper (14.3 m) forms had tall,
thin, heavily-conulated “spiky” tree-shaped protuberances with narrow basal attachments and small oscula not placed at the upper tips of the protuberances. We speculated that the variation in morphotypes were responses to diminishing proximal flows with depth.

Flow tank studies of morphological effects on flow patterns close to sponges – Chapter 4
The field observations were applied to the design a recirculating flow tank, which was used together with particle tracking velocimetry (PTV) and laser-illuminated high speed photograph to reconstruct near instantaneous flow fields around live specimens. We sought to better understand how the morphology in this sponge interacts with flow passing close to its surfaces, to gain a better overall understanding of how flow in nature helps to sculpt the body forms we encountered in the field, and how body forms, in turn, effect food particle capture in this species.
Flow results showed induced mixing at the sponge surface due to the complex surface morphology, with a high degree of variation between experiments. Regions of increased vorticity and larger-scale vortices were formed, changing and moving from moment to moment over the sponge surfaces. We speculated on how morphology-induced mixing could help trap food particles by increasing the collision events between particles and the sponge, contributing to increased food particle influx. We also discussed how the morphology in this species reflects the prevailing flow regimes recorded at each of the three depth stations in the field study.

Modeling work
Computational fluid dynamic simulations were initiated to model the effects of mm-scale conules on fine-scale flow over the sponge surface, and to study how varying the size, spacing and other aspects of the conules effected influx into the sponge. Preliminary results for two fields of 13 conules (0.4 and 0.8 mm tall) set on separate modeled pieces of flat sponge showed increased pressures on the leading flanks of the conules and reduced pressures on their leeward flanks increasing in magnitude with conule height. Influx simulations showed sensitivity to conule height, the total area over which the pumping pressure was applied,
and degree of blocking in the computational “tank.” Future simulations are planned to test other conule spacing and heights, the combined effects of conule spacing and height, and the spacing between conulated sponge protuberances in various orientations.

Oscular outflow in-the-sea and in the flow tank – Chapter 5
The flow tank and PTV methods were also used to visualize and calculate outflow velocities and rates from individual sponge oscula in relation to varying background velocities. Oscular outflow velocities ranged from (12.7–130.4 mm/s) and volumetric flow rates from 0.004-0.013 ml/ml sponge/s (includes any inducted inter-cellular free water). Outflow velocities varied with changes in background flow speeds, but no synchrony or significant patterns were seen. Sponges were shown to first slow down then cease pumping activity altogether when background flows reached the surprisingly low range of 7.5-12 cm/s (depending on the individual tested and experiment). Shut-down occurred for other reason than elevated background flows, and sometimes sponges shut-down for no apparent reason. A similar shutdown response in the same background flow speed range was also seen in the field with wild sponges.
During storms when current speeds were elevated and when there was a significant amount of sediment in the water column, most oscula on most sponges were closed. In calm weather 66% of oscula were open and flowing strongly; 25% were moderately flowing, and 8% were not flowing at all. Measurements taken from 1-5 days following a storm showed only 24% of oscula were strongly flowing, 75% moderately flowing, and only 1% not flowing at all.
Both the 3-month laboratory-acclimated sponges and the freshly-collected specimens produced outflow velocities in the flow-tank that was within the range measured for sponges in situ. A wide range (1.1-17.7 cm/s) in oscular outflow velocities were recorded for wild sponges. One large specimen in situ produced an estimated daily outflow rate of 1,172 L/day.
Calculated average outflow rates averaged 0.008 ml (ml sponge)–1s–1, which was somewhat lower than that reported in the literature for other sponge species.
We also sought to detect if flow through the aquiferous systems of the flow-tank specimens was enhanced by induction from ambient flow (as had been demonstrated for other sponge species and other benthic invertebrates). Flow-tank results showed relatively “flat” outflow curves over time, and we did not detect any evidence of induced flow through this sponge.

Over-sized ostia – a new discovery for Dysidea avara – Chapter 6
The flow tank and PTV tools were also applied to investigate unusual dermal membrane openings termed “over-sized ostia” (OSO), which we observed in our laboratory-held sponges. As the sponges remodeled their body forms to adjust to the new tank environment (a well documented process for sponges in general) they converted some oscula (outlets) into functioning over-sized (i.e., 1.5-2.5 mm dia.) inlets through internal re-plumbing of their inhalant-exhalant canals. With PTV we recorded tracer beads being taken into the OSO at high rates. Functioning OSO were also found on freshly-collected specimens brought into a flow-tank set-up in Spain, supporting the conclusion that development of OSO was not purely a response to specimens being held in tanks for extended periods. The full adaptive significance for D. avara forming and using the oversized inlets remains speculative pending future investigations. Based on our estimates of increased influx per unit sponge surface area, however, we theorized that the over-sized inlet structures allowed the sponge to exploit food
resources at an increased rate.

General Discussion and Conclusions
In the General Discussion and Conclusions Chapter, the role of aquaculture for future production of sponge natural products is discussed in the context of competing technologies. Given that current evidence for avarol biosynthesis implicates the sponge as the true source and not a microbial symbiont, aquaculture of Dysidea avara to yield the natural product is projected as the only viable large-scale production technology for the near-term.
I discuss some interesting un-answered questions generated during the thesis project and possible future directions for sponge flow research. I also discuss and present the very interesting and insightful preliminary results obtained from computational flow dynamic modeling simulations of the effects of mm-sized surface “bumps” called conules, on flow over and influx into modeled ostial fields placed onto a simulated piece of sponge surface.
I suggest future directions for continuing this very useful methodology to help us better elucidate the effects of specific morphology features on flow close to the sponge surface.
The General Conclusions recommend the most appropriate range and form for aquaculture flow regimes derived from the thesis results (i.e., maximum magnitude 7-10 cm/s, in an oscillating flow cycle). Key results from the individual chapters are discussed, and an overall recommendation for locating an in-sea Dysidea avara “aquapharm” is given.