In Barkley Sound, British Columbia, Canada, the wide-bladed red alga Mazzaella splendens occurs intertidally at low to intermediate wave exposure sites but not at adjacent high wave exposure intertidal sites that are occupied by the narrow-bladed sister species Mazzaella linearis. This study used morphological and biomechanical characteristics of both species to determine whether or not drag and acceleration forces could prevent blades of M. splendens from surviving at high exposure sites and hence account, in part, for their distributions For each species, these hydrodynamic forces were calculated for gametophyte and sporophyte phases and, when possible, short and long thalli. The most frequent break location when thalli were pulled by a spring-scale was the junction between the stipe and holdfast. The following predictions were made by a model that compared hydrodynamic forces which a blade would experience to the measured force required to break the stipe/holdfast junction: (1) long blades of M. splendens should not occur at high wave exposure sites, (2) within the range of wave exposure occupied by M. splendens, sporophytes should be more abundant than gametophytes when and where wave exposure is greatest and (3) long blades of M linearis are predicted to occur at high wave exposure sites. All 3 of these predictions agree with other studies of natural populations. Two predictions do not agree with field observations: (1) long M. linearis gametophytes are predicted to have greater survivorship at higher water velocities than long sporophytes but, in natural populations, sporophytes are actually more abundant when wave action is greater and (2) short blades of M. splendens are predicted to survive very high water velocities but, in reality, are absent from high wave exposure sites in Barkley Sound. The latter contradlct~on suggests that recruitment of M. splendens at high wave exposure sltes is prevented at a life history stage prior to the development of short blades.

Regulation of a Fucus djstichus L. emend. Powell population is probably mediated through a density-dependent mortality of the recruits. Blocks seeded with different densities of germlings of E distichusand cleared plots with naturally occurring recruits of different densities and with artificially thinned densities were used to monitor the effect of density on mortality and growth among recruits in a populat~on of F: distjchus in False Creek. Vancouver, British Columbia. Canada. In the first 2 mo of development, germlings on the high density blocks experienced a lower mortality than those on the low density blocks. At later stages (> 2 mo), the effect of density on mortality was reversed. Plants recruited into the cleared plots were > 1 mm in length and they also experienced a positive effect of density on mortality. In both the density blocks and cleared plots, plant mortality eventually became independent of plant density. Plant growth rate was generally not related to density but was related to plant size (length). There was a large variation in the sizes of the plants. All 3 measures of si7c inequality: zero-centered skewness, coefficient of variation, and Gini coefficient are significantly correlated. Plants in both the density blocks and cleared plots showed a slight tendency towards a more normal or positively skewed size distribution from fall 1986 to spring 1987, probably due to an increased mortality of the smaller plants during this time. The mechanism of dominance-suppression and difference in the intrinsic growth rate or timing of settlement probably all contributed to the inequality in size distribution.

Biomechanical analysis of wave-induced mortality in the marine alga Pterygophora californica MARINE ECOLOGY-PROGRESS SERIES Mar. Ecol. h o g. Ser.

Matnx models based on size and stage, corresponding to the monophasic, biphasic and triphasic life histones of algae have been constructed. Similarities and differences of these models with the Leslie matnx model, as well as with other extensions and modif~cations of it, are discussed. Two examples are given to show the utility of these models: one based on field data for Sargassum siliquosurn J. Ag. and a second, based on data for Laminaria longicruris Pyle., taken from the literature. Stable size distribution, reproductive values, sensitivity and elasticity of these models are analyzed. Many algae show discrete episodes of growth, reproduction, die back and regeneration. Hence, discrete matrix models appear to realistically describe the life history processes of these algae.

A general 9 X 9 matrix model based on recruit stages and plant size is proposed for a population of the brown alga Fucus distichus L. emend. Powell in False Creek, Vancouver, Canada. Twenty-six matrices were constructed each representing a monthly time interval covering the period from July 1985 to November 1987. Yearly matrices were also constructed as a product of monthly matrices covering periods of 1 yr. The characteristics of these matrices were evaluated by comparing the dominant eigenvalue (i.e. the population growth rate, A), the stable distribution, and reproductive values for each matrix. The relative contribution of each matrix parameter to population growth was assessed using elasticity analysis. The survival and transition of the plants among size classes contribute at least 50 % to the population growth rate. However, because E distichus plants do not exhibit vegetative regeneration, the population can only experience positive growth in the presence of recruitment. Absence of a germling bank has little effect on monthly projected population growth, but can reduce the population growth rate projected from the yearly matrix by as much as 83%. Current population size structure is very different from the projected stable distribution. The stable distribution generally has a large proportion of the plants in the recruit stage. Reproductive values are positively related to plant size but are comparable among the largest 4 size classes. Random combinations of monthly, seasonal, or yearly matrices were used to simulate population growth under various fluctuating environmental conditions. A negative growth rate was obtained in > 6 0 % of the simulation runs. In reality within the sampling period covered, the population is on the decline. It is likely that this population may recover by occasional pulses of a large number of recruits, as was observed in 1986.