Settlement behaviour of larvae of the Stripey Snapper, <Emphasis Type="Italic">Lutjanus carponotatus</Emphasis> (Teleostei: Lutjanidae)

Larval behaviour is important to dispersal and settlement, but is seldom quantified. Behavioural capabilities of larval Lutjanus carponotatus in both offshore pelagic and reef environments at Lizard Island, Great Barrier Reef were observed in situ to determine if they were sufficient to influence dispersal. Offshore, larvae swam with higher directional precision and faster on the windward side of the island (28 cm.s⁻¹) than on the leeward side (16 cm s⁻¹). Most larvae swam directionally. Mean swimming directions were southerly in the windward area and northerly in the leeward area. Larvae avoided the surface and remained mostly between 3–15 m. Larvae released near reefs were 2–3 times faster swimming away from reefs (19 cm s⁻¹) than swimming toward or over them (6–8 cm s⁻¹). Speed swimming away was similar to that offshore. Of 41 larvae released near reefs, 73% reached the reef, 59% settled, and 13% of those reaching the reef were eaten. Larvae settled onto hard and soft coral (58%), topographic reef features (29%) and sand and rubble (13%). Settlement depth averaged 5.5 m (2–8 m). Before settling larvae spent up to 800 s over the reef (mean 231 s) and swam up to 53 m (mean 14 m). About half of the larvae interacted with reef residents including predatory attacks and aggressive approaches by residents and aggressive approaches by settling larvae. Settlement behaviour of L. carponotatus was more similar to a serranid than to pomacentrids. Settlement-stage larvae of L. carponotatus are behaviourally capable, and have a complex settlement behaviour.     

Use of substratum ripples for flow refuging by Atlantic cod, Gadus morhua

The ability to maintain position in a current without actively swimming (station-holding) was measured on substratum ripples for Atlantic cod, Gadus morhua, a bentho-pelagic fusiform species. The current velocities tested ranged from 0–111 cm sec^-1. Ripples were sinusoidal, with twelve combinations of ripple wavelength (10, 25, 50, 125 cm) and ripple amplitude (1.0, 2.5, 5.0 cm). Ripple wavelengths were chosen to approximate 0.5, 1.0, 2.0 and 5.0 times fish total length. The potential of ripples to locally retard current and thereby provide a refuge from the flow was measured as a velocity ratio, u_trough/u_free-stream, where u_trough is the flow velocity measured at a height of 0.5 cm from the bottom of a trough and u_free-stream the flow velocity measured at a height of 10 cm above ripple crests. Cod usually swam steadily above substratum ripple crests in the free-stream flow. They used substratum ripples to hold station on only 3 of the 12 ripples tested by refuging from the flow in the ripple troughs (flow refuging). These ripples had wavelengths approaching twice the body length, with ripple amplitudes that produced velocity ratios of 0.44–0.65, thus providing at least a 35% flow reduction in the troughs. In addition, these ripples were only used at intermediate velocities starting at 49–78 cm sec^-1 and ending at 81–109 cm sec^-1 depending on the ripple morphology, suggesting there may be costs involved in flow refuging, probably in stability control. Flow refuging on substratum ripples could dramatically impact the physiology and ecology of cod in high current velocities by providing areas of retreat for energetic savings, but also offering opportunities for enhanced feeding and migration.     

Selection of diurnal refuges by the nocturnal squirrelfish, <Emphasis Type="Italic">Holocentrus rufus</Emphasis>

We examined the diurnal refuges occupied by the nocturnal squirrelfish, Holocentrus rufus, to describe refuges and the behavior associated with their use and to determine which, if any, refuge characteristics were selected. We tagged 21 H. rufus on two sites on a fringing reef in Barbados, West Indies, identified the refuges they used (n = 57), measured ten characteristics of each refuge and the surrounding microhabitat, and monitored their refuge use for 4 weeks. To evaluate refuge selection, we measured the same characteristics on a comparable number of unused potential refuges (n = 67) on the same reefs and used classification tree models to determine which characteristics separated used from unused refuges. Each fish used 1–9 refuges, which did not overlap among individuals and were defended against intrusion by conspecifics and some heterospecifics. Fish with more than one refuge frequently moved among them. There was strong site fidelity with no immigration of untagged fish or emigration of tagged fish on either reef during the study period and no additional refuges being occupied over the 4-week period. Refuges were primarily holes, open at one or two ends, which varied in size, distance from the reef edge, entrance orientation, and vertical relief at the entrance. Holes used as refuges differed significantly from unused holes mainly in characteristics related to the vertical position of their entrance, but the classification tree models differed for the two sites. This study provides the first detailed information on characteristics of daytime refuges used by a nocturnally active reef fish and the first evidence of selectivity of refuges. It suggests that the abundance and characteristics of holes on reefs could influence the density of H. rufus on natural reefs.     

Tidal activity pattern and feeding behaviour of the ophiuroid Ophiocoma scolopendrina on a Kenyan reef flat

Ophiocoma scolopendrina exhibits a distinctive pattern of feeding activity on intertidal reef platforms off Kenya. With the first wave of the flooding tide, dense aggregations of these ophiuroids (up to 320 m⁻²) engage in a 1–2 min burst of surface-film feeding, vigorously sweeping the air-water interface and associated sea foam with the ventral surface of 2–4 arms. Suspension feeding (with arms extended in the water column) is the primary feeding mode throughout the rest of the tidal cycle (involving 25–65% of the population at a time), while bottom feeding (with arms extended along the substratum) is infrequent (<10%). Field experiments showed that surface-film feeding is regulated by water depth and can be triggered by suspended particles. This feeding mode appears to be an adaptation to the intertidal habitat, enabling the ophiuroids to exploit a nutrient-rich surface film during a temporal refuge (low tide) from fish predation. Dense populations of O. scolopendrina may represent an important trophic link between producers of particulate organic material and higher-level consumers in coral reef environments.Tara Oak and Robert E. Scheibling contributed equally to this paper. The order of authorship is alphabetical     

Hyporheic Production is Substantially Greater than Benthic Production for a Common New Zealand Caddisfly

Life-history and production of Olinga feredayi in both benthic and hyporheic stream habitats were investigated in a pristine Waikato, New Zealand, forest stream over two years to investigate the contribution of hyporheic habitat to total secondary production. O. feredayi had a univoltine life-history with adult emergence occurring from November to March. Larvae with case lengths < 2 mm were present on most dates suggesting delayed egg hatching. Benthic densities were inversely related to maximum peak daily flow in the month prior to sampling, and positively related to the dry mass of particulate organic matter present in samples. Reach-average benthic production calculated by the size-frequency method was 0.024 g DM m⁻² year⁻¹. Hyporheic production was 4.276 g DM m⁻³ year⁻¹ and 6.462 g DM m⁻³ year⁻¹ in colonisation baskets set at 15–30 cm and 30–45 cm within the substratum, respectively, 2.3–3.4 times greater than production in surface baskets (0–15 cm). Averaged out over the reach scale, it was estimated that 96% of annual secondary production of O. feredayi occurred in hyporheic habitats >10 cm below the streambed surface. Our study clearly demonstrates that only sampling benthic habitats can lead to gross under-estimation of population-level annual production, and provides evidence for the role of the hyporheos as a source of secondary production that may partly account for the Allen Paradox.     

Energy metabolism of underwater swimming in river-otters (Lutra lutra L.)

We used a still-water swim channel in conjunction with open-flow oxygen and carbon dioxide respirometry to examine the energy requirements of river-otters (Lutra lutra L.) swimming voluntarily underwater in Neumünster Zoo (Germany). While at rest on land (5 °C), river-otters had a respiratory quotient of 0.77 and a resting metabolic rate of 4.1 W kg⁻¹. This increased to an estimated 6.4 W kg⁻¹ during rest in water (11–15 °C) and to 12.3 W kg⁻¹ when the animals were feeding in the channel. River-otters swimming under water preferred a mean speed of 0.89 m s⁻¹, and their energy requirements attained 11.6 W kg⁻¹. Cost of transport, however, was minimal at 1.3 m s⁻¹ and amounted to 0.95 J N⁻¹ m⁻¹. Accepted: 3 November 1997     

Phenotypic plasticity for skeletal growth,density and calcification of <Emphasis Type="Italic">Porites lobata</Emphasis> in response to habitat type

A reciprocal transplant experiment (RTE) of the reef-building coral Porites lobata between shallow (1.5 m at low tide) back reef and forereef habitats on Ofu and Olosega Islands, American Samoa, resulted in phenotypic plasticity for skeletal characteristics. Transplants from each source population (back reef and forereef) had higher skeletal growth rates, lower bulk densities, and higher calcification rates on the back reef than on the forereef. Mean annual skeletal extension rates, mean bulk densities, and mean annual calcification rates of RTE groups were 2.6–9.8 mm year⁻¹, 1.41–1.44 g cm⁻³, and 0.37–1.39 g cm⁻² year⁻¹ on the back reef, and 1.2–4.2 mm year⁻¹, 1.49–1.53 g cm⁻³, and 0.19–0.63 g cm⁻² year⁻¹ on the forereef, respectively. Bulk densities were especially responsive to habitat type, with densities of transplants increasing on the high energy forereef, and decreasing on the low energy back reef. Skeletal growth and calcification rates were also influenced by source population, even though zooxanthella genotype of source colonies did not vary between sites, and there was a transplant site x source population interaction for upward linear extension. Genetic differentiation may explain the source population effects, or the experiment may have been too brief for phenotypic plasticity of all skeletal characteristics to be fully expressed. Phenotypic plasticity for skeletal characteristics likely enables P. lobata colonies to assume the most suitable shape and density for a wide range of coral reef habitats.     

Uptake of picophytoplankton,bacterioplankton and virioplankton by a fringing coral reef community (Ningaloo Reef,Australia)

We examined the importance of picoplankton and virioplankton to reef trophodynamics at Ningaloo Reef, (north-western Australia), in May and November 2008. Picophytoplankton (Prochlorococcus, Synechococcus and picoeukaryotes), bacterioplankton (inclusive of bacteria and Archaea), virioplankton and chlorophyll a (Chl a) were measured at five stations following the consistent wave-driven unidirectional mean flow path of seawater across the reef and into the lagoon. Prochlorococcus, Synechococcus, picoeukaryotes and bacterioplankton were depleted to similar levels (~40% on average) over the fore reef, reef crest and reef flat (=‘active reef’), with negligible uptake occurring over the sandy bottom lagoon. Depletion of virioplankton also occurred but to more variable levels. Highest uptake rates, m, of picoplankton occurred over the reef crest, while uptake coefficients, S (independent of cell concentration), were similarly scaled over the reef zones, indicating no preferential uptake of any one group. Collectively, picophytoplankton, bacterioplankton and virioplankton accounted for the uptake of 29 mmol C m⁻² day⁻¹, with Synechococcus contributing the highest proportion of the removed C. Picoplankton and virioplankton accounted for 1–5 mmol N m⁻² day⁻¹ of the removed N, with bacterioplankton estimated to be a highly rich source of N. Results indicate the importance of ocean–reef interactions and the dependence of certain reef organisms on picoplanktonic supply for reef-level biogeochemistry processes.     

Small scale vertical gradients of Arctic ice algal photophysiological properties

Photosynthetic parameters of phytoplankton and sea ice algae from landfast sea ice of the Chukchi Sea off Point Barrow, Alaska, were assessed in spring 2005 and winter through spring 2006 using Pulse Amplitude Modulated (PAM) fluorometry including estimates of maximum quantum efficiency (F _v/F _m), maximum relative electron transport rate (rETR_max), photosynthetic efficiency (α), and the photoadaptive index (E _k). The use of centrifuged brine samples allowed to document vertical gradients in ice algal acclimation with 5 cm vertical resolution for the first time. Bottom ice algae (0–5 cm from ice–water interface) expressed low F _v/F _m (0.331–0.426) and low α (0.098–0.130 (μmol photons m⁻²s⁻¹)⁻¹) in December. F _v/F _m and α increased in March and May (0.468–0.588 and 0.141–0.438 (μmol photons m⁻²s⁻¹)⁻¹, respectively) indicating increased photosynthetic activity. In addition, increases in rETR_max (3.3–16.4 a.u.) and E _k (20–88 μmol photons m⁻² s⁻¹) from December to May illustrates a higher potential for primary productivity as communities become better acclimated to under-ice light conditions. In conclusion, photosynthetic performance by ice algae (as assessed by PAM fluorometry) was tightly linked to sea ice salinity, temperature, and inorganic nutrient concentrations (mainly nitrogen).     

Metabolic rates in an anadromous clupeid, the American shad (Alosa sapidissima)

To assess the energetics of migration in an anadromous fish, adult American shad (Alosa sapidissima) were swum in a large respirometer at a range of speeds (1.0–2.3 body lengths (BL) s⁻¹, 13–24 °C). Metabolic rate (M_O2) was logarithmically related to swimming speed (Bl s⁻¹; r ² = 0.41, slope = 0.23 ± 0.037) and tailbeat frequency (beats × min⁻¹; r ² = 0.52, slope = 0.003 ± 0.0003). Temperature had a significant effect on metabolic rate (r ² = 0.41) with a Q₁₀ of 2.2. Standard metabolic rate (SMR), determined directly after immobilization with the neuroblocker gallamine triethiodide, ranged from 2.2–6.2 mmolO₂ kg⁻¹ h⁻¹ and scaled with mass (W) such that SMR = 4.0 (±0.03)W^{0.695(±0.15)}. Comparison of directly determined and extrapolated SMR suggests that swimming respirometry provides a good estimate of SMR in this species, given the differences in basal activity monitored by the two methods. Overall, American shad metabolic rates (M_O2 and SMR) were intermediate between salmonids and fast-swimming perciforms, including tunas, and may be a result of evolutionary adaptation to their active pelagic, schooling life history. This study demonstrates variability in metabolic strategy among anadromous fishes that may be important to understanding the relative success of different migratory species under varying environmental conditions. Accepted: 3 March 1999     

The effect of substratum type on aspects of swimming performance and behaviour in shortnose sturgeon Acipenser brevirostrum

The swimming performance and associated swimming behaviour (i.e. substratum‐skimming, station‐holding and free swimming) were assessed in shortnose sturgeon Acipenser brevirostrum during critical swimming and endurance swimming tests over a rough and a smooth substratum. It was hypothesized that the addition of a rough substratum in the swimming flume may provide a surface for the A. brevirostrum to grip and offer an energetic advantage. Substratum type did not affect the critical swimming performance, but A. brevirostrum consistently performed more bottom behaviours (i.e. substratum‐skimming and station‐holding) while on a smooth substratum. Acipenser brevirostrum had little contact with the rough substratum until the velocity was >1 body length s^?1. Endurance swimming time was significantly lower for A. brevirostrum over the rough bottom at the highest velocity (30 cm s^?1) which may be attributed to the observed increase in free swimming and decrease in bottom behaviours. During endurance swimming, the rough substratum was mainly used at intermediate velocities, suggesting that there may be a stability cost associated with being in contact with the rough substratum at certain velocities.     

Distribution and sediment production of large benthic foraminifers on reef flats of the Majuro Atoll,Marshall Islands

The distributions and population densities of large benthic foraminifers (LBFs) were investigated on reef flats of the Majuro Atoll, Marshall Islands. Annual sediment production by foraminifers was estimated based on population density data. Predominant LBFs were Calcarina and Amphistegina, and the population densities of these foraminifers varied with location and substratum type on reef flats. Both foraminifers primarily attached to macrophytes, particularly turf-forming algae, and were most abundant on an ocean reef flat (ORF) and in an inter-island channel near windward, sparsely populated islands. Calcarina density was higher on windward compared to leeward sides of ORFs, whereas Amphistegina density was similar on both sides of ORFs. These foraminifers were more common on the ocean side relative to the lagoon side of reef flats around a windward reef island, and both were rare or absent in nearshore zones around reef islands and on an ORF near windward, densely populated islands. Foraminiferal production rates varied with the degree to which habitats were subject to water motion and human influences. Highly productive sites (>10³ g CaCO₃ m⁻² year⁻¹) included an ORF and an inter-island channel near windward, sparsely populated islands, and a seaward area of a reef flat with no reef islands. Low-productivity sites (<10 g CaCO₃ m⁻² year⁻¹) included generally nearshore zones of lagoonal reef flats, leeward ORFs, and a windward ORF near densely populated islands. These results suggest that the distribution and production of LBFs were largely influenced by a combination of natural environmental factors, including water motion, water depth, elevation relative to the lowest tidal level at spring tide, and the distribution of suitable substratum. The presence of reef islands may limit the distribution and production of foraminifers by altering water circulation in nearshore environments. Furthermore, increased anthropogenic factors (population and activities) may adversely affect foraminiferal distribution and production.     

Developmental phase-dependent photosynthetic responses to ultraviolet-B radiation: damage, defence, and adaptation of primary leaves of wheat seedlings

Alterations in photosynthetic capacity of primary leaves of wheat seedlings in response to ultraviolet-B (UV-B; 280–320 nm; 60 μmol m⁻² s⁻¹) exposure alone and in combination with photosynthetically active radiation (PAR; 400–800 nm; 200 μmol m⁻² s⁻¹) during different phases of leaf growth and development were assessed. UV-B exposure resulted in a phase-dependent differential loss in photosynthetic pigments, photochemical potential, photosystem 2 (PS2) quantum yield, and in vivo O₂ evolution. UV-B exposure induced maximum damage to the photosynthetic apparatus during senescence phase of development. The damages were partially alleviated when UV-B exposure was accompanied by PAR. UV-B induced an enhancement in accumulation of flavonoids during all phases of development while it caused a decline in anthocyanin content during senescence. The differential changes in these parameters demonstrated the adaptation ability of leaves to UV-B stress during all phases of development and the ability was modified in UV-B+ PAR exposed samples.     

Rainbow trout <Emphasis Type="Italic">Oncorhynchus mykiss</Emphasis> energetic responsesto pulsed flows in the American River,California, assessed by electromyogram telemetry

Although rainbow trout Oncorhynchus mykiss within the American River, California, apparently exhibit minimal upstream or downstream movements in response to hydroelectric-power-generation-related pulsed flows, the associated energetic costs are unknown. We implanted rainbow trout (n = 9, ≥30 cm SL) with electromyogram (EMG)-sensor-equipped radio transmitters to assess the swimming behavior and associated energetic costs associated with their responses to pulsed flows. Using laboratory calibrations in a Brett-type swimming respirometer, the trouts’ swimming speeds and oxygen consumption rates were estimated for their in-river EMG data, through a complete hydroelectric power-generation river pulsed-flow sequence (pre-pulse, increasing flow, peak, and decreasing flow stages), on several (mean: 3.2) sampling dates. Using a mixed-linear model, we found that fish swimming speed estimates increased during the increasing flow stage, while the associated mean oxygen consumption rates also increased at this stage. At river flows near the usual peak (>44 m³s⁻¹), swimming speeds and movement rates decreased, possibly due to the fish using the river’s habitat complexities as hydraulic cover. We conclude that rainbow trout incur increased swimming-related energetic costs during increasing flows and, potentially, decreased foraging opportunities at high flows.     

High Metabolic Rates in Beach Cast Communities

Metabolic hotspots at land–water interfaces are important in supporting biogeochemical processes. Here we confirm the generality of land–aquatic interfaces as biogeochemical hot spots by extending this concept to marine beach cast materials. In situ atmospheric pCO_2, from a respiration chamber (10 cm in diameter and 20 cm high) inserted into wrack deposits, was determined using a high-precision (±1 ppm) non-dispersive infrared gas analyzer (EGM-4, PP-systems) at 1 minute recording intervals. The wrack deposits supported high metabolic activities, with CO₂ fluxes averaging (±SE) 6.62 ± 0.88 μmol C m⁻² s⁻¹, compared to median value of 0.98 μmol C m⁻² s⁻¹ (mean 2.21 ± 1.25 μmol C m⁻² s⁻¹) for bare sand adjacent to deposits. Wrack metabolic rates ranged 40-fold across beaches, from a minimum of 0.57 ± 0.22 μmol C m⁻² s⁻¹ to a maximum of 20.8 ± 5.04 μmol C m⁻² s⁻¹, both derived from beaches with deposits dominated by Sargassum. Rates tended to increase significantly (F test, P < 0.05) from the shoreline to reach maximum rates at about 10 m from the shoreline, declining sharply further from the shoreline, and increased with increasing thickness of the deposits (maximum about 10 cm deep), declining for thicker deposits. Wrack differing in composition had similar metabolic rates, although deposits consisting of a mixture of seagrass and algae tended to show somewhat higher rates. Our results show a meter square of wrack deposit supports a metabolic rate equivalent to that supported by 3 m² of living seagrass or macroalgal habitat. In wrack, the marine environment provides organic material and moisture and the land environment provides oxygen to render wrack ecosystems an efficient metabolic reactor. Intense wrack metabolism should also be conducive to organismal growth by supporting the development of a cryptic, but diverse wrack-based food web.     

Evidence of central and peripheral gravireception in the ciliate Loxodes striatus

Effects of the density of the external medium on gravireception in Loxodes striatus were investigated using Percoll solutions. With increasing density, the swimming rates changed from prevailing in the downward direction to prevailing in the upward direction. A cellular density of 1.036 g cm⁻³ was determined measuring direction and speed of sedimenting immobilized cells at different accelerations and medium densities. Viscosity increases by Percoll were measured and taken into account. At 30% air saturation Loxodes maintained a negative gravikinesis of approximately −27 μm s⁻¹ at external densities corresponding to cellular density (±0.02 g cm⁻³). Negative gravikinesis decreased gradually to −9 μm s⁻¹ with the density difference rising from 0.020 to 0.036 g cm⁻³ (=normal). The data indicate the existence of central gravireception, presumably by the Müller organelle, to generate in swimming Loxodes a constant value of gravikinesis and a bimodal gravitaxis. Peripheral gravireception occurs, in addition to central gravireception, when the transmembrane density difference exceeds 0.02 g cm⁻³. Peripheral gravireception can neutralize, in part, gravikinesis as raised by the central gravireceptor. We hypothesize that both central and peripheral gravireception of Loxodes guide vertical locomotion in gliding and swimming cells. Accepted: 26 May 1998     

Physiological differences in the growth of <Emphasis Type="Italic">Sargassum horneri</Emphasis> between the germling and adult stages

The effects of temperature, irradiance, and daylength on Sargassum horneri growth were examined at the germling and adult stages to discern their physiological differences. Temperature–irradiance (10, 15, 20, 25, 30°C × 20, 40, 80 μmol photons m⁻²s⁻¹) and daylength (8, 12, 16, 24 h) experiments were carried out. The germlings and blades of S. horneri grew over a wide range of temperatures (10–25°C), irradiances (20–80 μmol photons m⁻²s⁻¹), and daylengths (8–24 h). At the optimal growth conditions, the relative growth rates (RGR) of the germlings were 21% day⁻¹ (25°C, 20 μmol photons m⁻²s⁻¹) and 13% day⁻¹ (8 h daylength). In contrast, the RGRs of the blade weights were 4% day⁻¹ (15°C, 20 μmol photons m⁻²s⁻¹) and 5% day⁻¹ (12 h daylength). Negative growth rates were found at 20 μmol photons m⁻²s⁻¹ of 20°C and 25°C treatments after 12 days. This phenomenon coincides with the necrosis of S. horneri blades in field populations. In conclusion, we found physiological differences between S. horneri germlings and adults with respect to daylength and temperature optima. The growth of S. horneri germlings could be enhanced at 25°C, 20 μmol photons m⁻²s⁻¹, and 8 h daylength for construction of Sargassum beds and restoration of barren areas.     

Light-Dependency of Growth and Secondary Metabolite Production in the Captive Zooxanthellate Soft Coral Sinularia flexibilis

The branching zooxanthellate soft coral Sinularia flexibillis releases antimicrobial and toxic compounds with potential pharmaceutical importance. As photosynthesis by the symbiotic algae is vital to the host, the light-dependency of the coral, including its specific growth rate (μ day⁻¹) and the physiological response to a range of light intensities (10–1,000 μmol quanta m⁻² s⁻¹) was studied for 12 weeks. Although a range of irradiances from 100 to 400 μmol quanta m⁻² s⁻¹ was favorable for S. flexibilis, based on chlorophyll content, a light intensity around 100 μmol quanta m⁻² s⁻¹ was found to be optimal. The contents of both zooxanthellae and chlorophyll a were highest at 100 μmol quanta m⁻² s⁻¹. The specific budding rate showed almost the same pattern as the specific growth rate. The concentration of the terpene flexibilide, produced by this species, increased at high light intensities (200–600 μmol quanta m⁻² s⁻¹).     

The effect of water exchange on bacterioplankton depletion and inorganic nutrient dynamics in coral reef cavities

We studied the effect of water exchange on the depletion (or accumulation) of bacterioplankton, dissolved organic matter and inorganic nutrients in small open framework cavities (50–70 l) at 15 m depth on the coral reef along Curaçao, Netherlands Antilles. The bacterioplankton removal rate in cavities increased with increasing water exchange rates up to a threshold of 0.0045 s⁻¹, reaching values of 50–100 mg C m⁻² total interior cavity surface area (CSA) per day. Beyond the threshold, bacterioplankton removal dropped. The cryptic community is apparently adapted to the average water exchange in these cavities (0.0041 s⁻¹). Dissolved inorganic nitrogen (DIN), nitrate + nitrite (NO _x ) in particular, accumulated in cavity water and the accumulation decreased with increasing water exchange. Net NO _x effluxes exceeded net DIN effluxes from cavities (average efflux rate of 1.9 mmol NO _x vs. 0.8 mmol DIN m⁻² interior CSA per day). The difference is ascribed to net ammonium losses (NH₄) in cavities at reef concentrations >0.025 μM NH₄, possibly due to enhanced nitrification. Dissolved inorganic phosphate accumulated in cavities, but was not related to water exchange. The cryptic biota in cavities depend on water exchange for optimization of consumption of bacterioplankton and removal of inorganic nitrogen. Coral cavities are an evident sink of bacterioplankton and a source of NO _x and PO ₄ ³⁻ .