Abstract Wave exposure is a complex set of biotic and abiotic factors, including variations in physical force, light, temperature, and nutrient levels. This study examines the distribution and morphology of an intertidal rockweed, Pelvetia compressa (Fucales, Phaeophyta) at three sites along a wave exposure gradient on the Monterey Peninsula, California, USA. Percent cover and vertical ranges were similar between the most exposed and intermediately exposed sites and much greater than at the most protected site. A discriminant function analysis of morphological data shows that the three populations may be objectively classified into three groups based on morphology. Äfter a discriminant function analysis, it is possible to accurately predict the site of origination for most of the thalli examined. Patterns between the morphological data from the extremes of wave exposure suggest that plants at the most exposed location tend to break at lower internodes, producing a "leggy" and pruned morphology compared to those at the protected site. This pruning may serve to lower drag forces and hence be an adaptation to heavy wave exposure. It is unclear from the data what factors are responsible for patterns seen at the intermediate exposure site; differences in age structure or other demographic factors may play an important role. Introduction Wave exposure is an important factor in the ecology of the rocky intertidal shore. It consists of a complex set of biotic and abiotic factors (Cousens 1982) and plays an important role in structuring intertidal communities (Dayton 1971. Sousa 1985). In addition to its effects on entire communities, wave exposure often influences the zonation of individual species along the intertidal shore (Stephenson and Stephenson 1949, Carefoot 1977, Ricketts et al. 1985). Öther effects on individual species include changes in growth rate (Cousens 1982) and changes in morphology (Denny et al. 1985, Gaylord et al. 1994, Blanchette 1994, Cousens 1982). In addition, material strengths of algal thalli may increase in high energy environments (Kraemer and Chapman 1991). Such changes reflect adaptations to stresses associated with wave exposure. In this study, I asked: Does the morphology, material strength, and distribution of Pelvetia compressa, a common intertidal alga, vary at three sites along a wave exposure gradient in a quantifiable, statistically significant way?! measured several characteristics of Pelvetia thalli at these three sites, including length, holdfast dimensions, internode dimensions, and material strength. also quantified percent cover, density, and vertical ranges of Pelvetia at each location. Secondly, I asked, if morphology, material strength, or distribution does vary at each site, can any of this variability be an adaptation to wave exposure? Materials and Methods Study Sites Ihree sites were chosen along a wave exposure gradient on the Monterey Peninsula in Pacific Grove, California, USA (Fig. 1). The most protected site was at Hopkins Marine Station (hereafter HMS) located on Point Cabrillo. The study site faces north and is protected from waves by several off-shore rock formations. The intertidal zone is horizontally broad and gently sloped. It also includes many medium to large boulder clumps, creating small "islands." Morphology data was taken immediately to the west of Agassiz Beach while the distribution was measured over a 30m by 120 m area of the intertidal. The intermediately exposed site was on the west side of Lovers Point (hereafter LVR). It is very narrow with many steep, almost vertical faces and the north/northwest facing study site is approximately 30 m long. The most exposed site was on the north face of Point Pinos, (hereafter PTP). It is a very broad and gently sloped rocky shore several hundred meters long. The gradient was established using off-shore wave heights taken near the Monterey Bay Aquarium (Monterey Bay Aquarium, unpublished data), located near the study site at HMS, as well as by regression formulae which predict wave heights near the other two study sites (Graham 1995). All measurements were made in months of April and May, 1996. The Study Organism-Pelvetia compressa Pelvetia compressa (C. Ag., De Toni), formerly known as Pelvetia fastigiata (J. Agardh, De Toni, De Cew and Silva in press) is a brown, perennial intertidal alga, commonly found either isolated or in thick aggregations on top of protected rocks in the mid-littoral belt from Coos Bay, Oregon to Lower California (Ensenada) (Setchell and Gardner 1925). In the areas studied, Pelvetia compressa is the only species of this genus; it will hereafter be referred to as Pelvetia. The fronds branch closely above an ellipsoid holdfast in a dichotomous fashion with apical growth (Lee 1980) but Pelvetia also is morphologically plastic. Multiple apical cells may be initiated at sites of damage, so Pelvetia may appear to branch in a non-dichotomous fashion. Pelvetia produces gametes yearly in bloated reproductive structures called receptacles at the tips of ultimate dichotomies and shed their gametes at low tide (Bold and Wynne 1978). Like all fucoids, Pelvetia has only a diploid macroscopic phase and lacks alternation of generations. Since the holdfasts of adjacent Pelvetia may fuse, one morphological individual may represent not one but many genetic individuals. Therefore, following the convention of Aberg (1989) a morphologically distinct individual will be labeled a "plant" or "thallus" and not "individual". I define "morphologically distinct" as an plant whose holdfast does not contact that of any other. All plants in this study were reproductively mature and randomly selected from the middle of the Pelvetia zone. Distribution and Vertical Range Measurements 1o establish the vertical range of Pelvetia, I used a surveyor's transit and stadia rod. Heights were measured relative to a USGS benchmark (at HMS) or relative to the predicted height of the water at low tide (PTP and LVR). To measure percent cover at HMS, I sampled random points within the vertical range of Pelvetia inside a 120 m by 30 m transect. Using a 0.25 m- quadrat, I counted the percent cover and number of individuals inside each quadrat. At LVR, 0.25 m2 quardrats were placed randomly within two transects, one 30m by 2 m and the other 23 m by 2 m. At PTP, three 30 m by 2 m transects were sampled. The percent cover and densities were compared with a one-way ANÖVA followed by a Tukey's test to determine significance of differences between sites. The ANÖVA calculations for density measurement were performed upon log,, transformed data due to heterogeneity of variances. Morphology The following features of Pelvetia were measured for thalli at each site (figure 2): maximum frond length (to nearest centimeter), number of stipes per holdfast, and holdfast and stalk cross sectional area (to nearest 0.01 mm). For a randomly selected frond in each thallus, the following features were measured: internode lengths (to 0.01 mm) and areas (to 0.01 cm2) for lowest 3 internodes breaking strength of the frond (Nem2), and location of first non-dichotomous branching (hereafter referred to as "bushiness"). Twenty-five plants were measured at HMS, 22 at LVR, and 24 at PTP. Distinctness of the morphologies at the three sites was determined using discriminant analysis. The discriminant function analysis is a multivariate statistical technique which creates linear combinations of the morphology variables to maximize differences among the study sites. Next it applies this discriminant function to individual observations to classify whether a plant came from HMS, LVR, or PTP. If the morphologies are different at the three sites, the discriminant function should correctly classify each plant. If they are similar, the discriminate function will not classify them correctly. In addition, the separation of the three populations may be graphically displayed as an xy-scatter plot. The x-axis (Factor 1) represents a combination of morphological variables which maximizes differences among the sites. The y-axis (Factor 2) is a second combination of morphological variables with the second largest difference among sites but is also statistically independent from factor 1 (Afifi and Clark 1984). In addition, single factor ANÖVA's were performed on the morphological features individually to test for differences among sites. Those which were significantly different were tested with a Tukey's tests to find significant differences between sites. In addition, ten plants from each site were haphazardly removed from the site to analyze thallus areas and lengths using the computer program NIH Image 1.55 (National Center for Supercomputing Applications). Each thallus was spread upon a sheet of paper and videotaped. The image was imported into NIH Image and traced to find the area. Strength as a Material Property lo find the strength of the thallus as a material property, the breaking strengths of approximately 100 plants from each site were measured using a 20 kg spring scale and nylon webbing to hold each frond. The webbing was wrapped around the end of each frond and attached to the spring scale. The scale was then pulled until the frond broke cleanly. The cross sectional area of each break was found by measuring the dimensions of each break and the breaking strengths were computed as: =Breaking Force (Newtons))/(Cross Sectional Area of Break (m2)) These strengths were ranked in numerical order and each rank was given a probability of: probability - rank n+1 with n= total number of plants at each site. These probabilities were graphed versus their associated breaking strengths to show the probability of breakage for a plant at each site given a certain breaking strength. The mean strengths from each site were tested with a one-way ANÖVA followed by a Tukey's test to test differences between means. Results Distribution and Vertical Height The vertical range of Pelvetia was similar at LVR and PTP. At both these sites, Pelvetia extends both higher and lower in the intertidal height range than at HMS (Table 1). LVR and PTP also had similar percent covers and density; at both sites, density and percent cover were greater than at HMS (Fig 3). ANOVA calculations indicated significant differences (p « 0.05) among the three sites (Table 2 & 3). A Tukey's tests showed significant differences of density and percent cover (p « 0.05) between all pairwise combinations of sites except between LVR and PTP. Morphology Compared to HMS plants, the plants at PTP appear tougher and "leggier. The first several internodes are longer and most branching does not begin until the 4th or 5th internode at PTP. These long internodes often have scars where a branch had been broken off. Plants at LVR, however, appear to have much new growth with very regular branching. -Discriminant Function Analysis The discriminant function correctly classified most of the plants from each site (Table 4). The individual features which were significantly different (p « 0.05) among the sites were internode lengths, internode areas, bushiness, and breaking strength (Table 5). In general, internode length increases from HMS to LVR to PTP while internode areas are similar at HMS and PTP but less at LVR. PTP plants begin to "bush" at lower internodes than those at HMS and LVR (Fig. 4). These features would create plants with thin internodes at LVR and plants with long internodes at PTP. These results confirm qualitative observations about the plants at each site. The separation of the three populations based upon morphology may be illustrated as an xy-scatter plot (Fig. 5). PIP plants had a lower ratio of plant area to length than those at HMS (Table 6). At-test between HMS and PTP showed a significant difference between the mean ratios of area to length at the 95% confidence interval. -Breaking Strength as a Material Property The cumulative probability of breakage versus breaking strength curve illustrates that LVR plants are noticeably weaker than those at HMS and PTP (Fig. 6). The PTP plants have the strongest material of all three sites. In addition, a one way ANÖVA on the mean breaking strengths showed significant differences (p « 0.05) among the three sites (Table 7). All pairwise differences in means were significant by a Tukey's test (p«0.05) Discussion Pelvetia Distribution The differences in vertical range of Pelvetia among the three sites is not unexpected. Intertidal zones shift upwards along the shore and broaden when exposure increases (Lewis 1964, Ricketts et al. 1985, Carefoot 1977). In addition, the range found for Pelvetia at HMS agrees with that found by Doty in 1946 who studied intertidal zonation at HMS (Doty 1946). It was unexpected that HMS would have such a low percent cover and density compared to LVR and PTP since Pelvetia has been reported as a protected intertidal species (Setchell and Gardner 1925). This pattern, however, may still be related to the degree of protection of a shore. Cousens (1982) describes exposed locations as "less extreme and less variable" (p. 194) due to temperature and nutrient buffers created by greater water turnover. It has been shown that the abundance of other fucoids varies through time with environmental conditions, usually due to desiccation stress (Gunnill 1980). In addition, it has been noted that the Pelvetia population at HMS has declined sharply in the last 20 years (Watanabe, Baxter personal communication). It is possible that desiccation stress at HMS may have killed a large percentage of the population and that such environmentally-dependent variability may be an important factor in Pelvetia's ecology. The lower wave exposure at HMS, therefore, may have contributed to a decrease in abundance through desiccation stress. Percent cover and density at LVR and PTP may be similar for several reasons. The actual wave exposure at these two sites may be very similar or there may be a threshold wave exposure which allows Pelvetia to reach a maximal abundance, after which its abundance becomes limited by other factors such as competition for space. Morphology The discriminant function analysis showed that the three populations are morphologically distinguishable since most plants were accurately classified. In addition, single factor ANÖVA's showed that several morphological features as well as breaking strengths varied significantly among the sites. Since these three populations are statistically distinguishable, I hypothesized that there may be biological reasons for this separation. To elucidate the biological reasons behind the morphological differences, first examined differences between HMS and PTP, the two sites at the extremes of wave exposure and asked if the variability in morphology and breaking strength could be adaptations to wave exposure. There were no significant differences in internode area between these two sites so the plants do not grow thicker fronds at PTP to adapt to wave exposure. Plants at PTP are stronger than those at HMS; for a given breaking strength, PTP plants are less likely to break than those at HMS. Changes in algal material properties as a response to wave exposure have also been studied in Egregia menziesii (Kraemer and Chapman 1991); the strengths of young Egregia thalli were greater when grown in high energy situations than those grown in low energy situations. Increasing the material strength of the algal tissue, therefore, may be one adaptation of PTF plants to wave exposure. Another feature which significantly differs between HMS and PTP is internode length. There could be several explanations why internode lengths at PTP are approximately twice those at HMS. First, the internodes could grow to longer lengths at PTP and branch less. Second, the plants at PTP could prune and break off branches due to wave forces, making a longer internode out of two or more shorter internodes. The second explanation may be more likely because l observed many long internodes at PTP which were actually the result of two shorter internodes separated by a scar where a branch had been broken off. In addition, pruning could be an adaptation to wave forces. The drag force is proportional to the amount of area on an object. It is possible that Pelvetia adapts to wave forces by reducing area on which drag may act while maintaining the same internode thickness. This adaptive characteristic has been seen in other fucoids (Blanchette 1994) and mechanical limits to size due to wave forces have been postulated for other intertidal algae (Gaylord et al. 1994). Pruning may reduce the area on a thallus and hence reduce the amount of drag if the plants did not regain the pruned area by growing more at the tips of the thalli. If the plants do lose area at PTP relative to those at HMS by pruning, the ratjo of area to length should be less at PTP than at HMS. This is true (Table 6); hence. the differences in internode lengths between the two sites may be related to pruning as an adaptation to wave forces. Though the reduction of area per length from HMS to PTP seems fairly small, this amount may be enough to significantly reduce the drag forces on plants at PTP (see Appendix 1). Cousens (1982) found that the morphology of Ascophyllum nodosum at an intermediately exposed site was not necessarily intermediate compared to sites at the extremes of exposure. Many morphological features at LVR are not intermediate compared to those at HMS and PTP. Internode lengths are intermediate compared to those at HMS and PTP; this result agrees with the pruning hypothesis. If the plants prune as an adaptation to wave forces and the wave forces are intermediate at LVR compared to HMS and PTP, then the plants at LVR should prune an intermediate number of times. The average internode length should be between those of HMS and PTP. Internode areas at LVR were less than those at HMS and PTP (Fig. 4b) and breaking strengths were much less (Fig. 6, Table 7). These results are difficult to understand in terms of wave forces. If wave forces are greater at LVR than at HMS, why would the thalli at LVR grow weaker and thinner than those at HMS? There could be several explanations for the patterns at LVR. Öther factors of wave exposure, such as temperature or nutrient levels, specific to the amount of exposure at LVR may contribute to these patterns. The nutrient levels may allow for more new growth at LVR, creating thinner and weaker internodes, In addition, there may be other causes which are unrelated to exposure. The reason for the morphologies at LVR may be specific to the population at LVR¬ perhaps the plants are of a different age structure or different gene pool. The results do not clearly suggest one possibility over another; further study should be conducted on the population at LVR to determine what factors control the morphologies and breaking strengths. In conclusion, this study showed that the distribution, abundance, and morphology of Pelvetia are distinct at these three sites along a wave exposure gradient and that some of these differences may be adaptations to wave exposure. The factors contributing to the distinct features of LVR plants remain unclear. When conducting an ecological study of wave exposure, one must remember that exposure consists of many factors whose relative influence may change at different exposure levels (Cousens 1982). Most commonly associated with wave exposure are mechanical factors (Denny et al. 1985, Denny 1988) or wave-borne projectiles (Shanks and Wright 1986). In addition, there are biological factors associated with water motion such as decreased desiccation stress or temperature and nutrient levels. At a very exposed location, hydrodynamic forces create a very extreme mechanical environment while temperature and nutrient buffers due to increased water turnover create a less extreme environment (Cousens 1982). The ultimate impact upon the ecology of intertidal organisms is due to a combined effect of these factors. In addition. without conducting long-term demographic studies or reciprocal transplant experiments, one cannot disregard demographic factors or site-specific characteristics in a study of exposure and morphology/distribution. Age structures of the populations at each site may be different, causing widespread differences in morphology. Since the growth rates of Pelvetia may be negative (Gunnill 1979), it is difficult to ascertain growth rates and ages. An age structure could only be determined by following plants over a number of years. One way to determine if the environment is responsible for the morphological variability in Pelvetia would be to perform reciprocal transplant experiments between protected and exposed sites. Such an experiment has been conducted on Fucus gardneri (Blanchette 1994); the results supports the hypothesis that wave forces directly influence plant morphology. A reciprocal transplant experiment should be done with Pelvetia in conjunction with a discriminant function analysis This would show whether transplanted thalli from site "a" looked more like those at site "a" or those at the site of transplantation, site "b" after a certain amount of time. If the discriminant function classifies thalli transplanted from site "a" to site "b" as site "b" plants, it would further strengthen the claim that wave exposure strongly influences the morphology of Pelvetia. Acknowledgments Tam indebted to Dr. James Watanabe for his assistance with the discriminant function analysis and with the experimental design, as well as for his endless patience and good-humor. In addition, I thank Dr. Mark Denny for his frequent and lucid explanations of biomechanics and wave forces. The following people have also been instrumental in my research: Dr. J. Wible, B. Gaylord, Dr. C. Blanchette, Dr. J. Connor, Dr. P. Silva, Dr. S. Brawley, S. J. Liaw, J. Strharsky, E. Leydig, and C. Patton. Literature Cited Aberg, P. (1989). Distinguishing between genetic individuals in Ascophyllum nodosum populations on the Swedish west coast. Br. Phycol. J. 24: 183- 190. Afifi, A. A. and V. Clark. (1984). Computer Aided Multivariate Analysis. Belmont, California: Lifetime Learning Publications. Agardh, C. (1924). Systema Algarum. vol. 38. Lunda: Literis Berlingianis. Agardh, J. (1841). In historiam algarum symbolae. Linnaea. 15: 1-50. Bell, E. C., and M.W. Denny. (1994). Quantifying "wave exposure": a simple device for recording maximum velocity and results of its use at several field sites. J. Exp. Mar. Bio. Ecol. 181: 9-29. Blanchette, C. A. (1994). The effects of biomechanical and ecological factors on population and community structure of wave-exposed, intertidal macroalgae. Oregon State University. Bold, H. C. and M. J. Wynne. (1978). Introduction to the algae. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Carefoot, T. (1977). Pacific seashores: A guide to intertidal ecology. Seattle: University of Washington Press. Cousens, R. (1982). The Effect of Exposure to Wave Action on the Morphology and Pigmentation of Ascophyllum nodosum (L.) Le Jolis in South-Eastern Canada. Botanica mar. 25: 191-195. Dayton, P. K. (1971). Competition, disturbance, and community organization: the provision and subsequent utilization of space in a rocky intertidal community. Ecol Monogr. 41: 351-389. De Cew, T. and P. Silva. (in press). De Toni, J. B. (1895). Sylloge Algarum. vol 3. pp. 215-216. Patavii: Sumptibus auctoris. Denny, M. (1995). Predicting physical disturbance: mechanistic approaches to the study of survivorship on wave-swept shores. Ecol. Monogr. 65(4): 371-418. Denny, M. W., et al. (1985). Mechanical limits to size in wave-swept organisms. Ecol. Monogr. 55: 69-102. Denny, M.W. (1988). Biology and the mechanics of the wave-swept environment. Princeton: Princeton University Press. Doty, M. S. (1946). Critical tide factors that are correlated with the vertical distribution of marine algae and other organisms along the Pacific coast. Ecology. 27: 315-328. Gaylord, B., C. A. Blanchette, and M.W. Denny (1994). Mechanical consequences of size in wave-swept algae. Ecol. Monogr. 64: 287-313. Graham, M. (1995). Regulation of the shallow limit of giant kelp, Macrocystis pyrifera, at three sites along the Monterey Peninsula. San Jose State University. Gunnill, F. C. (1979). The effect of host distribution on the faunas inhabiting an intertidal alga. DAI. 40: 4012-4444. Gunnill, F. C. (1980). Demography of the intertidal brown alga Pelvetia fastigiata in Southern California, USA. Mar Biol. 59: 169-179. Kraemer, G. P. and D. J. Chapman. (1991). Biomechanics and alginic acid composition during hydrodynamic adaptation by Egregia menziesii (Phaeophyta) juveniles. J. Phycol. 27 (1): 47-53. Lee, R. E. (1980). Phycology. Cambridge: Cambridge University Press. Lewis, J.R. (1964). The ecology of rocky shores. London: Hodder & Stoughton. Longuet-Higgins, M. S. (1952) On the statistical distribution of the heights of sea waves. J. Mar. Res. 11: 245-266. Ricketts, E. F. et al. (1985) Between Pacific Tides. Stanford: Stanford University Press. Setchell, W. A. and N. L. Gardner. (1925). The marine algae of the Pacific coast of North America. Berkeley: University of California Press. Shanks, A. L., and W. G. Wright. (1986). Ädding teeth to wave action: the destructive effects of wave-borne rocks on intertidal organisms. Oecologia. 69: 420-428. Sousa, W.P. (1984). Disturbance and patch dynamics on rocky intertidal shores. In The ecology of natural disturbance and patch dynamics, (ed. S. T. A. Pickett and P. S. White), pp101-124. New York: Academic Press. Stephenson, T.A., and A. Stephenson. (1949). The universal features of zonation between tide-marks on rocky coasts. J. Ecol., 37: 289-305. Appendix 1. Determination of maximum wave heights and drag force Hydrodynamic forces on a wave-swept shore consist of lift, drag, and acceleration. The lift force acts perpendicular to the direction of flow and is proportional to area. Drag is also proportional to area but acts in the direction of flow. Acceleration also acts in the direction of flow but is proportional to an object's volume. The effects of lift and acceleration are usually small in relation to drag (Denny 1995, Gaylord et al. 1994) which is why only drag is considered in these calculations. The relative exposure of the three study sites was determined by visual estimation of wave height and offshore wave heights at each location as determined from wave height data and regression formulae (Monterey Bay Aquarium, unpublished results, Graham 1995). To estimate relative wave forces at the three sites lused these off-shore wave heights. Extrapolating from off-shore wave heights to degrees of in-shore exposure, however, is very complex and involves making empirical measurements of topography at each site (Denny 1995). In addition, local variations in topography might make exposure highly variable on a scale of meters within the same site (Bell and Denny 1994). Taking the average maximum significant wave height at the Monterey Bay Aquarium from 1987-1992 as the maximum significant wave height at HMS, applied this value to a Graham's (1995) regression formula to find the maximum height at PTP. From this maximum wave height, I found the am.. d-H (1) 272 ams is not the standard deviation of surface elevation but only the root-mean¬ square of the amplitude of waves. Using the statistics of extremes (Longuet¬ Higgins 1952),I found the maximum wave amplitude: dmax - (n(f 1) Ddms (2) (Longuet-Higgins 1952) 2I() with f=1/9 s" (Denny 1995), 0 = .5772, and t = 8 hours since the measurements made for the off-shore wave heights at the Monterey Bay Aquarium were taken every 8 hours. The maximum wave velocity for a turbulent bore running up the beach is: (3) Imax — 8 (20ma) with una= maximum water velocity and g= 9.81 m:s2. For PTP, the maximum wave velocity was 9.55 m.s" while at HMS, the maximum wave velocity was 7.22 m.s. The maximum drag force was calculated as: Fa =3puPSpAp (4) (Denny 1995) with p=density of seawater, 1024 kg-m“, Sap =the shape coefficient which is a constant for each organism at a particular orientation to flow = 0.18 (Gaylord, personal communication), ßy is the velocity coefficient = 1.7 (Gaylord, personal communication), and Ap, is the maximum projected area A PTP plant with an average length per area of 0.319. m’/m and an average length of 0.368 m would experience a drag force of 501.4 N. If it had the same average length per area as HMS plants (.4199 m2/m), it would experience of a force of 660.0 N. An average length plant at HMS (0.396 m) would experience 441.5 N. Without the reduction in area per length, a PTP plant experiences 1.49 times the drag force as a plant at HMS. Äfter the reduction in area per length, the PTP plant experiences only 1.14 times the drag force as a plant at HMS Tables Table 1. Heights of Pelvetia (feet above MLLW) Site Lower Limit Upper Limit Range (ft above (ft above (ft above MLLW MLLW MLLW HMS 4.15 3.09 1.06 4.00 LVR 5.62 1.62 PTP 1.94 5.9 3.96 Table 2. Analysis of variance of percent cover of Pelvetia at HMS, LVR, and PTP Source of Variation of MS value 4111.87 6.86 0.002 Between Groups Within Groups 111 599.015 Total 113 Table 3. Analysis of variance of density of Pelvetia at HMS, LVR, and PTP. Due to heterogeneity of variances as determined by the Cochran’s test, the ANÖVA calculations were performed upon log,, transformed data. Source of P-value MS Variation Between Groups 1.86 9.92 60.001 111 Within Groups 0.19 113 Total Table 4. Predicted sites of origination based upon morphological characteristics. The elements in each row originated at the same site; these elements are placed in columns which are the predicted sites of origination. Site Predicted Total PI LVR HMS Actual 18 HMS 21 LVR 18 21 PTF 20 21 Total 23 Table 5. Single factor ANÖVA calculations on morphology data VARIABLE DE MS Length 110340 2.525 089 Error 60 43.703 Stipes/Holdfast 35.244 1.938 153 60 Error 18.185 Internode 1 Length 1056.866 6.003 004 60 Error 176.048 Internode 1 Area 1047.565 3.258 045 Error 60 321.566 2086.457 9.500 Internode 2 Length 0.000 60 Error 219.636 Internode 2 Area 1301.424 5.293 008 60 Error 245.857 Internode 3 Length 1484.934 8.499 007 Error 60 174.709 1484.934 8.499 Internode 3 Area 001 Error 174.709 528 Holdfast Area 787495.018 592 Error 1490496.927 3.868 026 Bushy 11.263 60 Error 2.912 797193.379 1.044 358 Stalk Area 60 763711.483 Error 133375E414 006 Breaking Strength 5.650 60 236054E113 Error Table 6. Mean ratio of area to length for 10 plants from each site. Site Ratio (mim?) HMS 4199 LVR 3279 PTP 319 Table 7. One way ANÖVA on mean breaking strengths SUMMARY Groups Average Variance HMS 103 1.82E112 2468543 2.67E111 LVF 116 1133765 PTF 114 2880860 2.3E112 ANOVA Source of MS F P-value Variation Between 9.56Et13 66.19 60.001 Groups 330 1.44E+12 Within Groups Tot. 332 Figure Legends Figure 1. Map of study sites along Monterey Peninsula. The inset shows a closeup of the peninsula with the study sites labeled. Figure 2. Morphology measurements a. Measurements made upon each thallus. (a) the number of stipes emerging from each holdfast. (b) dimensions of the cross sectional area of the stalk. (c) dimensions of cross sectional area of holdfast. (d) length of thallus. b. Measurements made upon randomly selected frond in each thallus. The internode lengths and cross sections were measured. An internode is defined as the distance from one fork to the next. Internodes are numbered with integers. The internode numbered "0" is the one which emerges from the holdfast before any branching begins. Figure 3. Distribution of Pelvetia at three sites. Significant comparisons (ps 0.05 as determined by a Tukey's test) are marked by a """ while non-significant comparisons are marked "ns". Error bars illustrate +- one standard error. a. Percent cover of Pelvetia at three sites. b. Density of Pelvetia in number of individuals per m2. Figure 4. Morphological features with significant univariate F-tests. Significant differences (p20.05 as determined by a Tukey's test) are marked by a * while non-significant differences are marked by "ns". Error bars illustrate +/- one standard error. (a) Lengths of internodes 1-3 at HMS, LVR, PTP in millimeters. (b) Cross-sectional areas of internodes 1-3 at HMS, LVR, PTP in millimeters?, (c) Bushiness at HMS, LVR, PTP. The internode number on the y-axis was determined by the convention illustrated in figure 2. Figure 5. Scatter plot of the separation of three populations based upon morphology and breaking strength. The x-axis (factor 1) is a combination of variables which maximizes differences among sites. The y-axis (factor 2) is a second combination of morphological variables with the second largest difference among sites but is also statistically independent from factor 1. Figure 6. Cumulative probability curve of breakage versus breaking strength. The applied stress (breaking strength) is shown on the x axis with the probability of breaking at that stress shown on the y axis. This curve illustrates that for a given stress, plants at LVR are much more likely to break than those at PTP or HMS. Scale in miles 1/4 1/2 3/4 PTE S 10' LVR Figure 1. Map of Monterey Peninsula HMS Monterey Peninsula 122° W 50 Figure 2. Morphology measurements a. Measurements on each thallus a. 6 —c— b. Measurements on each frond 2.0 0.5 00 20 15 40 30 20 10 Figure 3a. Percent cover Lovers Pt Pinos □Percent Cover Hopkins Figure 3 b. Density (number of individuals per m“2) Pt Pin UDensity (Number of Individuals per m'2) Lovers Hopkins Figure 4a. Internode Lengths Internode Figure 4 b. Internode Areas Intemode HOPKINS ELOVERS OPT PINOS EHOPKINS ELOVERS OPT PINOS 5 Hopkins Figure 4 c. Bushiness Lovers Pt Pinos DBUSIN C 90 . 9 2 ape o2 z10e C I aaa- 3 8 5 o geo