(0.25 m, 1.25 m and 2.25 m below mean low water) at each site four 1 X 1 meter quadrats were randomly placed by throwing a rock with a float attached into.
Botanica Marina Vol. 39, 1996, pp. 525-531 © 1996 by Walter de Gruyter · Berlin · New York
The Effects of Wave Exposure and Depth on the Morphology of Inshore Populations of the Namibian Kelp, Laminaria schinzii Foslie R J. Molloy*»*'1 and J. J. Boltonb a
University Marine Biological Station, Millport, Isle of Cumbrae, Scotland KA280EG, U. K. Department of Botany, University of Cape Town, Private Bag, Rondebosch 7700, Cape Town, Rep. of South Africa * Corresponding author
b
Populations of Laminaria schinzii from three depths on three Namibian shores exposed to different degrees of wave action were studied to determine morphological variation with depth and water motion. Analysis of variance indicated that all of the measured morphological characteristics varied significantly with depth and water motion (as measured by plaster clod erosion) but the degree of variation (f) for individual characteristics differed markedly. Depth had a greater effect on morphology as a whole but the strongest single interaction was between water motion and blade thickness. The plant characteristics measured were analysed using discriminant analysis and the depth/site groups were well separated on the two axes of the resulting ordination with 70% of the grouped cases correctly classified. The greatest separation was along function I, which correlated most strongly with blade thickness; function II correlated with stipe length, stipe length/stipe weight and frond surface area. No significant difference in attachment strength was found between plants growing in the extremes of water motion, a decrease in frond surface area in more exposed sites resulted in a decreased resistance to water motion. Introduction In the south-western Cape region of South Africa, Laminaria pallida (Grev.) J. Ag. grows as an understorey kelp to Ecklonia maxima (Osbeck) Papenfuss down to 8m below chart datum in wave exposed conditions. In waters from 8—20 meters L. pallida is the dominant kelp (Jarman and Carter 1981). On the northern Cape coast Laminaria schinzii Foslie grows with E. maxima from Cape Columbine (32°50'S, 18° E; Fig. 1) into Namibia and ultimately becomes the dominant kelp (Anderson and Bolton 1985, Anderson et 1989, Molloy 1990). Field et al (1980) found that total kelp biomass decreased northwards along the west Cape coast from a maximum on the Cape Peninsula and suggested that decreasing water clarity may be the reason. Jarman and Carter (1981) estimated the standing crops of L. pallida and E. maxima to be 217960 and 336370 tons wet weight, respectively, between Cape Point and Cape Columbine. Due to upwelling along the western coast of South Africa and Namibia, the annual average sea temperature ranges are relatively low, considering the subtropical latitudes i.e. 12-14.8 °C for Lüderitz (Fig. 1) and 12-18.4 °C for Swakopmund, (Molloy 1990). Upwelled cool water is rich in nutrients and, because
of high rates of primary and secondary production, light penetration tends to be low except for short periods between upwelling cycles (Shannon 1985). Adult Laminaria schinzii has a hollow stipe (the only hollow stiped member of the Digitatae complex reported) and is up to 5 m in length (personal observation). The stipe is widest in the mid-region, becoming narrower at the distal and proximal ends, and is very often verrucate. At the southern end of its distribution L. schinzii occurs with L. pallida and can be difficult to distinguish (Bolton, pers. obs.). However, L. pallida has a conical stipe, widest near the
1
Present address: Department of Biology, University of Namibia, Private Bag 13301, 340 Mandume Ndemufayo Ave., Pioneerspak, Windhoek, Namibia.
Fig. 1. Kelp bed study sites on Lüderitz peninsula.
Brought to you by | provisional account Unauthenticated Download Date | 4/11/15 4:14 PM
526
F. J. Molloy and J. J. Bolton
holdfast, and is shorter than L. schinzii. An intermediate form exists in the northern cape province of South Africa, with the general appearance of L. pallida but with a hollow stipe (Bolton, pers. obs.), this intermediate form was also found on the Cape Peninsula (pers. obs.). The two species were found to hybridise and give rise to morphologically normal FI sporophytes [torn Dieck (Bartsch) and Oliveira 1993]. On the southern part of the Namibian coast, where the substratum is suitable, Laminaria schinzii occurs in dense beds interspersed with a few E. maxima plants. The beds are restricted to rocky areas, which, on the Namibian coast, occur very intermittently (Molloy 1990). The ± 80 km stretch of rocky coastline around Lüderitz is the largest expanse of rock on the 1500 km coast which, for the most part, is sandy with scattered rocky outcrops (Molloy 1990). Commercial harvesting of Laminaria schinzii has been carried out on a trial basis in the Lüderitz area. Blades were dried and exported for human consumption. Harvesting commenced in April 1987 and ended in March 1989 with a total of 137 tonnes (wet weight) harvested. The harvest was discontinued because the consumers found the texture of the blades unsuitable (Molloy 1990, Critchley et al. 1991). Estimates of standing stock for Laminaria schinzii or Eckhnia maxima from the north west Cape and Namibian coasts are not available and there are no published data of the standing stock of L. schinzii for any part of its distribution. Molloy (1990) calculated a standing stock available to the harvesting technique employed by the Lüderitz harvesters, for a 20km stretch of the coast at Lüderitz, as 300 tonnes (wet) per year. The harvesting technique involved manually cutting individual blades 2cm above the meristem and the removal of all blades from a harvested plant. Only the shallow portions of the beds (1 — 1.5 m deep) were harvested. The environment in which Namibian Laminaria grows differs from that in which most other Laminaria populations grow in that the seasonal variation in environmental variables is relatively small. Light availability varies less than at higher latitudes due to daylength and light intensity variation decreasing with decreasing latitude. Continuous upwelling ensures a relatively small water temperature range (12— 14.8 °C, running average 23 days over 10 years for Lüderitz, Molloy 1990) with associated lack of largescale seasonal nutrient limitation. This is the first ecological study of this important marine macrophyte. Materials and Methods Three sites on the Namibian coast, (Shearwater Bay, Dias Point and Grossebucht) were chosen to represent sheltered, semi-exposed and exposed habitats, respectively (Fig. 1). Shearwater Bay is sheltered from the open ocean swells (Fig. 1) by Dias Point. The prevailing south-south westerly winds (Molloy 1992)
for the area are off-shore at this site, and as a result very little wind induced wave action occurs. Dias Point is partially sheltered from swell by the position of off-shore rocks, lying to the south-south west. The Grossebucht site is open to the full force of the Atlantic swell, with an average monthly maximum of 3 m (Molloy 1992; swells in excess of 7 m are common here). The shore is steep, attaining 10m depth just 150 m off-shore, so the outer regions of the kelp bed are of little protection to the inner portion. The prevailing south-south westerly winds are offered no obstruction at the site. Because of wave exposure the site proved very difficult to work, except on very calm days. Only in-shore sites were chosen for this study as the aim was to collect ecological data of relevance to the harvesting effort which was confined to in-shore sites. Blades were harvested by hand after wading into the kelp beds when sea conditions suited; no SCUBA gear was used for harvesting. For each depth (0.25 m, 1.25 m and 2.25 m below mean low water) at each site four 1 X 1 meter quadrats were randomly placed by throwing a rock with a float attached into the bed at the appropriate depth. All plants inside the quadrats were removed by cutting just above the holdfast, so the holdfast was not included in weight measurements. The plants were broght back to the shore where the numbers of Laminaria schinzii and Ecklonia maxima were noted and stipe length and weight, frond length and weight, total weight, and frond thickness of each L. schinzii plant were measured. Surface area was calculated as: frond weight (g)/specific density/thickness (cm) 10000 To measure specific density, samples of fronds (20 per depth per site) were taken back to the laboratory. Each sample was a large piece cut mid-way along the blade. From each blade sample one 2 X 2 cm square was cut, blotted and the thickness measured with vernier calipers. The piece was then weighed on a balance accurate to 10"1 g and the specific density calculated as: weight (g)/[4 thickness (cm)] The thickness of a frond was found to be relatively uniform, except in the area of the transition zone between stipe and frond where thickness was greater and in the meristematic palm which was up to 1 mm thicker than the blades. On many specimens the stipe was already flattened up to 20 cm proximal to the region where the stipe widened to become the frond. To measure the attachment strength of the Laminaria plants, a chain (2 cm link width) was secured to the base of the stipe with 8—10 coils of 6 mm nylon rope. The chaim was run out of the kelp bed to the shore and attached to a 300 kg suspension scale. The scale was in turn coupled to a hand cranked winch that was secured to a crack in the rock with a steel
Brought to you by | provisional account Unauthenticated Download Date | 4/11/15 4:14 PM
Wave exposure and Namibian Laminaria
rod. The chain was cranked in until the holdfast lost its attachment and the reading taken. Approximately 28% of the stipes broke due to the cutting action of the rope, and these were not included in the analysis. Because of the method used no plants were more than 20 m away from the shore and only plants in l m of water were chosen. This method was not in-
tended to simulate wave action on the thalli but to give a relative measure of attachment strength. Relative exposure to water motion was measured using a plaster clod method modified from Doty (1971). Replicate hemispherical clods (3 cm diameter) were made from one plaster mix (1 volume water : 2 volumes of builders plaster) the plaster was allowed
Stipe length
Frond length
GB
SWB
SWB
Total weight
SWB
GB
Stipe weight
GB
SWB
Frond weight
SWB
527
GB
Frond surface area
GB
Stipe wt./cm length
GB
SWB
Stipe wt. / Total wt.
GB
GB
SWB
Frond thickness
Plaster clod erosion
Site
SWB
Fig. 2. Measured morphlogical characteristics for Laminaria schinzii. Stipe length, frond length, total weight, stipe weight, frond weight, frond surface area, stipe weight/cm length, stipe weight/total weight and frond thickness. Number of plants: Shearwater Bay 0.25 m, n = 72; 1.25 m, n = 22; 2.25 m, n = 5. Dias Point 0.25 m, n = 46; 1.25 m, n = 22; 2.25 m, n = 29. Grossebucht 0.25 m, n = 87; 1.25 m, n = 22; 2.25 m, n = 8. Percentage plaster clod erosion is also included (n = 5 for each depth at each site). Standard errors on all bars. Brought to you by | provisional account Unauthenticated Download Date | 4/11/15 4:14 PM
528
F. J. Molloy and J. J. Bolton
to set for 24 h and then oven dried (80 °C) for a further 24 h and the dry wt. measured. The clods were then glued to concrete flagstones (5 per stone) using epoxy glue. At the 0.25 m and 1.25 m sites all kelp plants within a 3 -4 m radius of the flagstones were cleared to prevent stipes and fronds coming into contact with the clods, this was not necessary at 2.25 m as the floating kelps remained upright. All flagstones were put in position on the same day. As a control, a flagstone with 5 clods was immersed in a drum of sea water and re-weighed the following day. The clods were removed after 24 h, oven dried (80 °C) for 24 h and the dry weights taken. General linear model ANOVA of plant characteristics against depth and water motion (plaster clod erosion) was performed with Minitab. The latter is the appropriate analysis of variance when the number of observations per site is not equal as in this investigation. The predictors used in the discriminant analysis were those that scored the highest f values in the general linear model ANOVA i.e. stipe length, frond surface area, stipe length/stipe weight and frond thickness. Predictors, such as stipe weight and frond weight, were excluded to avoid covariance. Stipe weight, frond weight and total weight were incorporated as stipe weight/total weight. The analysis was performed using 'Statistical Package for Social Sciences Personal Computer version -f' (SPSSPC-l·). At the 0.25 m sites many juvenile plants were present, and to avoid a large bias towards the juveniles, plants with a stipe length of less than 25 cm were excluded from the analysis.
Fronds at 1.25 m tended to be longer and lighter at Dias Point, shorter at Grossebucht and heavier at Dias Point. At 2.25m the tendency was towards longer and heavier fronds at Shearwater Bay and shorter and lighter ones at Grossebucht. Frond surface area tended to be larger at 2.25 m and smaller at 0.25 m at all sites. At 0.25 m it tended to be larger at Dias Point and smaller at Shearwater Bay. At l m and 2.25 m frond surface area was larger at Shearwater Bay and smaller at Grossebucht. At all depths frond thickness tended to be greatest at Grossebucht and least at Shearwater Bay. All plaster clods at Grossebucht were eroded away, indicating that these were the most exposed sites. At the 0.25 m Dias Point site 72.9 ± 10.3% erosion occurred. Erosion at Shearwater Bay 0.25 m and 1.25 m and Dias Point 1.25m was similar at 45.9 ± 1.4%, 45.1 ± 0.75% and 45.1 ± 2.3%, respectively. Erosion was least at Shearwater Bay 2.25 m and Dias Point 2.25m at 36.3 ± 1.1% and 34.7 ± 1.5% respectively (Fig. 2). In the general linear ANOVA model (Table I) all interactions were significant (p < 0.05) meaning that all of the measured plant characteristics varied significantly with different depths and degrees of exposure (as measured by plaster clod erosion). However, the Table I. General linear model ANOVA variance ratio f values for interactions between measured plant characteristics and depth and water motion (as measured by plaster clod erosion). Significance is indicated. Parameters
Depth
Water motion
Stipe length
f 223 P
