The recovery and dynamics of submerged aquatic macrophyte vegeta tion in the Wilderness lakes , southern Cape

Between 1979 and 1981, the submerged aquatic macrophyte vegetation in the Wilderness lakes died back significantly, and in some areas disappeared altogether. This study documents the senescent phase and describes the recovery of the plant populations between May 1982 and May 1983. In two lakes, namely Langvlei and Eilandvlei, the plant biomass approximately doubled between the winters of 1982 and 1983. Seasonal changes in species composition are documented and possible factors accounting for the collapse and recovery of the plant populations are discussed.


INTRODUCTION
Several authors (Howard-Williams 1980; Jacot Guillarmod 1982; Weisser & Howard-Williams 1982) have studied the aquatic macrophyte communities in the Wilderness lakes.Their work, which was conducted prior to a major die-back of aquatic plants throughout the lakes, provide base-line data for describing vegetation changes arising from a combination of seasonal and successional processes (Miles 1979).
Between 1975 and 1979, submerged macrophytes were an important biological component of the Wilderness lakes, producing in excess of 750 t of dry plant mass per annum (Howard-Williams 1980).The most important species were Potamogeton pectinatus, Chara globularis, Lamprothamnium papulosum and Ruppia cirrhosa.The aquatic macrophyte population beds in this system died back significantly between 1979 and 1981.This study describes the senescent phase and documents seasonal changes in species composition and biomass which occurred during part of the recovery period (1982)(1983).

STUDY AREA
The Wilderness lakes comprise a segmented barrier lagoon system lying parallel to the southern Cape coast.There are three interconnected lakes (Rondevlei, Langvlei and Eilandvlei) linked to the Wilderness Lagoon and the Touws River by a narrow winding channel known as the Serpentine (Figure 1).The Wilderness Lagoon is shallow, varying in depth from 1 m in its lower reaches to 3.5 m in the vicinity of the Serpentine entrance.The lakes are all deeper than the lagoon, with a maximum depth of 6.5 m in Eilandvlei, 4.0 m in Langvlei and 6 f Rondevlei.there are generally few replacement taxa.This hypothesis is supported by the current study which found that, during the senescent phase, no new macrophyte species invaded the previously occupied littoral zone.Recolonization potential was limited to five species and contrasts to the situation pertaining in most terrestrial environments e.g.tropical forests.

Eilandvlei
The rapidity with which vegetation changes proceed in brackish waters is noteworthy.In this study community 'succession' was most prevalent in Eilandvlei (Figure 5).Nevertheless, in both Eilandvlei and Langvlei, seasonal changes in mean standing crops followed a similar pattern and were of the same order of magnitude.Mean biomass per square metre approximately doubled between the winter of 1982 and the winter of 1983 in both lakes.
However, the rate of expansion of plant beds towards the centre of the lake in Langvlei was more rapid than in Eilandvlei, a much deeper system.Average water trans parencies (Secchi disc) in Langvlei were 1.7 m and in Eilandvlei 1.2 m over the period April 1982 to January Evidence from Langvlei suggests that aquatic vegeta tion proceeds through cycles; from an aquatic macrophyte dominated system through a phytoplankton dominated system and back to a macrophyte dominated system.However, the dominant submerged plants at the beginning of the study (Characeae) were not well represented at the end of the study period, when P pectinatus was dominant.The cause(s) of most of the submerged macrophytes dying in the Wilderness lakes has not been identified and may well differ from one part of the system to another.According to Weisser & Howard-Williams (1982) the highwater level management policy, planktonic algal blooms, development of periphyton on submerged plants and up rooting of plants by wave action during strong winds may all have contributed to declines in submerged macrophytes within the system.It is interesting to note that a similar collapse in aquatic macrophytes was recorded from nearby Swartvlei in 1979 (Davies 1982;Whitfield 1984), and 10 years later these Characeae and P pectinatus beds had still not fully recovered (A.K. Whitfield pers.obs.).In Natal, Breen & Weisser (1986) recorded the disappearance of previously abundant Potamogeton schweinjurthii and P pectinatus in Lake Mzingazi, but could not identify any definite fector(s) which causal the die-back.Breen & Weisser (op.dr.) did, however, suggest that rapid changes in lake water levels may have contributed to the collapse.This investigation has shown that aquatic plant commu nities of the Wilderness lakes are highly variable, both on a spatial and temporal scale, and that this dynamism is due to a complex interplay of environmental and biotic factors.The study has also provided base-line in formation which will facilitate future monitoring of the system.In conclusion we would like to recommend regular vegetation mapping and aerial photographic coverage of the study area as a valuable tool in the future management of the Wilderness lakes.
whole lake system (91 km2) is situated on a series of Quaternary sands known as the Wildemess-Knysna embayment.Immediately inland of this embayment is a 200 m high Tertiary platform (foothills of the Outeniqua Mountain range) consisting of folded sediments of the Table Mountain Sandstone group, which forms part of the catchment area of streams and rivers that feed the lakes and Wilderness Lagoon.The lake system is estuarine in nature and only periodically open to the sea at the Touws River Mouth.An important characteristic of the Wilderness lakes is its reversed salinity gradient (Allanson & Whitfield 1983).Rondevlei, which only receives fresh water from precipi tation and is the lake farthest from the sea, has the highest salinity, whereas Eilandvlei, which is also influenced by fresh water input from the Touws River has the lowest sa linity.Evaporation at Rondevlei (1 100-1 200 mm per an num) generally exceeds precipitation and this leads to increasing salinity during dry periods.Heavy rainfall years have the opposite effect.A series of floods in 1981 resulted in a dilution of salt present in all three lakes.At Rondevlei the salinity decreased from 16 g kg1 to 10 g kg1 between September 1980 and September 1981, whereas over the same period in Langvlei and Eilandvlei salinity decreased from 10 g kg ' to 5 g kg1 and 9 g k g1 to 4 g kg1 respectively (Allanson & Whitfield 1983).The waters throughout the system are generally well oxy genated.though low oxygen concentrations « 2 mg I1) are occasionally recorded in deeper areas of all three lakes during the summer months (December-February).

FIGURE 1 .
FIGURE 1.-Map of the Wilderness lakes showing localities mentioned in the text.
LakeSystem prior to the die-back were mapped according to a method described byWeisser & Howard-Williams (1982).During the senescent phase(1979)(1980)(1981), the degree of aquatic macrophyte die-back in each part of the system was visually monitored by boat at monthly intervals.The recovery phase was more intensively monitored and sampling commenced in May 1982.Six littoral transects were positioned at approximately equidistant intervals around Eilandvlei and five around Langvlei.The limits of each transect were defined as the edge of the emergent zone and the deepest point at which macrophyte growth was recorded.A submerged macrophyte sampler de scribed by Howard-Williams & Longman (1976) was used to sample a 0.0625 m2 area at variable intervals along the length of each transect.Samples were returned to the laboratory for sorting and drying.Results are expressed as g m 2 dry weight.The total species biomass along each transect was established by multiplying the mean bi omass for each plant species by the total transect length.The above procedure was repeated in May, August and November 1982, and in February and May 1983.In this study, the period December-February is regarded as summer, March-May as autumn, June-August as winter and September-November as spring.The standing biomasses of submerged macrophyte popu lations in Rondevlei and Wilderness Lagoon were too low to sample effectively with the macrophyte collector.There fore only changes in species composition in these two sections of the Wilderness lakes were recorded.Past distribution patterns of aquatic macrophytes in the Wilderness lakes were obtained from direct inspection of historical aerial photographs using a Wild Aviopret Stereo scope (Weisser & Stadler 1983).
In 1975 most of Eilandvlei (Figure 2) supported a littoral plant community dominated by Potamogeton pectinatus, which occurred in water up to 3 m deep.At depths greater than 3 m, no macrophytes were recorded.Zones of Characeae and Ruppia cirrhosa occurred on the landward side of the P pectinatus zone (Figure 2).Between 1976 and 1978, the P. pectinatus zone (Figure 3) receded by almost 40 m at the southeastern side of the lake.This decline continued during 1979, so that by January 1980 no mature plant canopy was evident.A series of se vere floods in 1981 caused silt-laden waters (400-500 mg dry mass I*1) to enter Eilandvlei from the Duiwe River, increasing water turbidity.The recovery of submerged aquatic plant beds com menced during 1982 with Najas marina occupying large areas previously colonized by P. pectinatus (Figures 2 & 4).Changes in species composition were most apparent in the northeastern section of Eilandvlei.In May 1982 the mean dry biomass (six sampling transects around the lake) of N. marina was 635 g m 2 (S.D. ± 1032), that of P pectinatus 109 g m 2 (S.D. ± 81) and Characeae 4 g m 2 (S.D. ± 5).The average transect length in Eilandvlei FIGURE 2 .-Distribution of submerged aquatic macrophytes in Eilandvlei during 1975 (adapted from Weisser & Howard-Williams 1982).Dashed line = limit of study area; stippled areas = Potamogeton pectinatus; areas with vertical lines = Characeae; oblique lines = Ruppia cirrhosa\ horizontal lines = Najas marina.

FIGURE 4 .
FIGURE 4. -Distribution of submerged aquatic macrophytes in Eiland vlei during 1982.Areas with vertical lines = Najas marina\ stippled = Potamogeton pectinatus.vegetation of Rondevlei comprised a peripheral band of the dominants Ruppia cirrhosa and R pectinatus (Weisser & Howard-Williams 1982) extending to a depth of 1.3 m.In March 1979 the submerged macrophytes of Rondevlei had diminished considerably compared to March 1978, and the few scattered remaining plants did not reach the surface.Aerial photographs taken during December 1981 (Job 391, No. 291/3) confirmed the macrophyte senescent phase.This situation persisted until 1982 when evidence of a recovery was recorded.In May 1982 sparse patches of R. cirrhosa covered approximately 30% of the littoral zone to a depth of about 1 m.By May 1983 these patches, together with R pectinatus, had expanded to cover approxi mately 90% of the littoral zone.