We dissolved plaster forms in seawater to examine the effects of surface roughness and flow conditions on masstransfer rates. Plaster blocks with varying roughness were dissolved under both steady and oscillatory flows between 7 and 43 cm s Ϫ1 yielding calcium mass-transfer coefficients (S Ca ) that varied from 0.5 to 3 m d
Ϫ1. S Ca measured in a flume was 30-40% greater under oscillatory flow than under steady flow at flow speeds Ͻ10 cm s Ϫ1 ; this difference decreased with increasing flow speed. Plaster blocks with added millimeter-, centimeter-, and decimeter-scale roughness were dissolved under oscillatory flows between 2 and 26 cm s Ϫ1 on three different coral reef flats. Plaster dissolution rates in the field were linearly proportional to surface area regardless of the roughness scale. Variation in S Ca across different reef environments was less affected by whether the flow was steady or oscillatory (a factor of 1.2-1.6) than it was by flow speed alone (a factor of 4.7).Water motion is an important factor regulating the exchange of metabolites between reef communities and their surrounding environment. Rates of nutrient uptake (Atkinson and Bilger 1992; Thomas and Atkinson 1997), photosynthetic production (Dennison and Barnes 1988;Carpenter et al. 1991), and nitrogen fixation (Williams and Carpenter 1988;Carpenter et al. 1991) by coral and algae increase with greater water motion. This occurs as a result of changes in the rate at which dissolved compounds are physically transferred across concentration boundary layers adjacent to the surfaces of autotrophic organisms. Prior studies of metabolite mass transfer have been conducted using reef communities with morphologies incorporating multiple scales of roughness (10 Ϫ3 to 10 Ϫ1 m); however, little attention has been paid to the relative importance of varying roughness scales to total rates of metabolite mass transfer to reef communities. It is well known that autotrophic reef organisms exhibit morphologies that incorporate multiple scales of roughness (Kaandorp 1999). However, the success of correlating nutrient mass-transfer rates to bottom drag or frictional dissipation have led some investigators to suggest that small roughness scales are unimportant to total rates of metabolite mass transfer (Baird and Atkinson 1997;Hearn et al. 2001). In addition, most studies of water motion on coral and algal metabolism have been conducted under steady, unidirection-1 To whom correspondence should be addressed. Present address: MSB 205, 1000 Pope Road, Honolulu, Hawaii 96822 (falter@ hawaii.edu).
AcknowledgmentsThanks to Jim Fleming and Eric Hochberg for suggestions and help in doing field work.