Monitoring Phragmites australis increases from 1937 to 1976 in the Siyai Lagoon (Natal, South Africa) by means of air photo interpretation

The colonization o f the Siyai Lagoon on the north coast o f N atal by Phragmites australis was studied by means o f air photo interpretation. It was possible to locate and estim ate P. australis areas for 1957 (0,74 ha), 1965 (1,65 ha), 1969 (1,93 ha) and 1976 (2,94 ha). Phragmites australis first inhabited the shores o f the middle section o f the lagoon follow ed by rapid expansion in the lower section. The upper section was colonized only at its lower end by expansion from the middle section. It is suggested that P. australis was unsuccessful in this section because o f com petition by the Hibiscus tiliaceus—Barringtonia racemosa Lagoon Fringe Forest. This same com m unity is shading out P. australis in som e places. The notable increase in the rate o f advance o f land and littoral vegetation into the Siyai Lagoon was caused by sugar farming activities leading to erosion and sedim entation in the lagoon. A vegetation age gradient was observed from the upper section to the m outh region. The colonization o f most o f the Siyai Lagoon except the immediate m outh zone by P. australis Reedswamp and Hibiscus tiliaceus—Barringtonia racemosa L agoon Fringe Forest, can be expected before the turn o f the century. Dredging and mechanical control o f vegeta­ tion will becom e necessary if major open water spaces are to be maintained.


IN T R O D U C T IO N
An increase o f reeds was observed on air photos when studying the dune advancement at Mtunzini (Weisser et al., MS). The availability o f air photo coverage for 1937-1979 offered the possibility to m onitor the reed encroachm ent. Objectives of this work were to evaluate the adequacy o f air photos in m onitoring P. australis expansion; to provide base line data on the Siyai Lagoon; to quantify the reed encroachm ent; to establish trends and extrapolate possible developments; and to offer possible m anage m ent suggestions. The Siyai drainage system is situated in Natal at latitude 28°58' South and longitude 31°45'45" East. The Lagoon is situated amidst a well-conserved, scenically beautiful, forested-dune landscape ( Fig.   1). It has a surface o f about 8 ha, an axial length of 2,5 km and a catchm ent of about 18 km 2 (Begg, 1978). Two stream s are the main tributaries. In this work only the lagoon was studied and defined as ex tending from the confluence o f the two tributaries to the Lagoon M outh. Four sections were distinguished from south-west to north-east: upper section (from confluence to about 650 m north-east); middle sec tion (inflexion area o f watercourse); lower section (to watercourse constriction) and m outh section.
Between 1937  The potential and limitations o f the use of air photos in vegetation studies have been discussed in literature, e.g. Edwards (1972) and Weisser (1979). Concerning this study, the following points should be borne in mind. The photographs are at different scales and were not taken at the same time of year. A nother factor is the differing resolution o f the air photos com pared, the 1937 photographs having the poorest resolution. Also, the water levels o f the Siyai Lagoon varied, therefore changing the area occu pied. However, the Siyai Lagoon is situated between steep dune ridges, consequently changes in water levels have only a minor influence on the area inun dated. The reed growth varies seasonally. There is a time lag between the establishment of reeds and their detection on air photos. Sometimes P. australis col onization may be below the detection limit of the air photo taken. The density o f the P. australis stand also varies, being at its lowest in winter. This may affect its detection on the air photos.

R ESULTS A N D C O N C L U S IO N S
Field checking confirm ed that the Siyai Lagoon was a good site for studying the colonization and ex pansion of P. australis Reedswamp. The period documented in the air photos was long enough to enable the reeds to cover all the available habitat in some sections o f the river, whereas in others P. australis is still actively expanding. The observation time was also sufficient to allow succession to pro ceed, and in some areas P. australis Reedswamp is being displaced by other riverine vegetation (Fig. 1). The results deal first with the findings on P. australis, then with other aquatic or sem i-aquatic vegetation, and finally with the Siyai Lagoon.

Adequacy o f air photos to m onitor P. australis colonization
It was possible to locate and estim ate P. australis areas for 1957, 1965, 1969 and 1976 (Table 1 & Fig.  3). The resolution o f the 1937 air photos is insuffi cient to give conclusive evidence. Difficulties were encountered in the interpretation o f Job 291 (1977) and Job 329 (1979). Their scale (1:3 000) gives too few matching points in the optical field of the ZTS.

Colonization pattern
Phragmites australis first colonized the middle and lower sections o f the Lagoon. It failed to establish itself in the immediate m outh zone and in the upper region with the exception o f a 0,5 m2 patch in the confluence (June, 1980 shallow. It seems that when silting increased drastic ally in the upper section, the already present woody fringe vegetation rapidly invaded the new habitat and ousted P. australis. In the lower section o f the Lagoon, P. australis was able to establish itself and to expand in the direction of the Lagoon m outh.

Phragm ites australis Reedswamp area changes: 1937-1976
The P. australis surface increases at the Siyai Lagoon are summarized in Table 1 and represented graphically in Fig. 3. Whereas from 1937 to 1957 there was an increase o f only 0,47 (?) ha in the area covered by P. australis, the area increased by 2,20 ha between 1957 and 1976. Therefore expansion of P. australis in the Lagoon accelerated after 1957 (Fig.  3). The curve will flatten out as the available habitat dim inishes. H ow ever, with the n o rth -eastw ard advancement o f the m outh, new areas suitable for P. australis colonization are being created (Fig. 2).

Probable causes o f reedswamp increase
Reedswamp expansion is often a natural process in some lagoons. Conspicuous increases of reeds have been reported in lakes after lowering of water levels (e.g. Kopf, 1964, in Weisser 1970Bjork, 1974) or when siltation increases (Weisser, 1978). Begg (1980) states that reed encroachm ent is due to the Lagoon becoming shallower and less saline. We agree that sedim entation and consequent reduction in depth are the chief causes o f the rem arkable increase in P. australis. No records are available for assessing the influence of the salinity factor.
The sedim entation o f the Siyai Lagoon has in creased strikingly, becoming critical with the advent o f sugar farm ing in the catchment area (c. 1946 fide Begg, 1978). This must be considered as the main reason for the deterioration of the Siyai Lagoon. G arland (pers. com m .) considers the clearing and cultivation o f Swamp Forest and Cyperus papyrus Swamp (c. 1955) as being especially detrim ental, because o f the con seq uent elim in atio n o f the 'sponge' function o f the vegetation.

Succession
In some areas o f the middle section of the Lagoon, a replacement o f P. australis Reedswamp by the Hibiscus tiliaceus-Barringtonia racemosa Lagoon Fringe Forest was observed, probably by shading. This com m unity seems to be the following phase in the land reclam ation process in this Lagoon. This corresponds with successional schemes proposed (Edw ards, 1967;see also W ard, 1980).

Reduction o f open water by increase o f woody shore vegetation
An expansion of the Lagoon Fringe Forest form ed mainly by Hibiscus tiliaceus and Barringtonia

Age gradient o f woody shore vegetation
In the upper section of the Lagoon, the woody fringe vegetation consists o f dense tall trees (about 9 m), whereas the woody vegetation downstream is less dense, lower and formed by young trees. In the lower section, the woody shore vegetation is only 2 to 3 m high, discontinuous and limited to isolated seedlings o f Hibiscus tiliaceus and Barringtonia racemosa. Therefore, an age gradient for the Lagoon fringe vegetation from the upper Lagoon section to the Lagoon m outh can be assumed (see also Weisser et al., MS).
A few patches of dead woody fringe vegetation were noted in the lower section o f the Lagoon. The preceding high water levels have been suggested as the cause by G arland (pers. com m .).

Area changes o f the Siyai Lagoon
The surface covered by the Lagoon including reeds, but excluding the Hibiscus tiliaceus-Barringtonia racemosa Lagoon Fringe Forest has increased from about 6,06 ha (1937) to 7,87 ha (1976). This increase is mainly due to accretionary processes in the mouth region, which has advanced about 740 m northwards , therefore oc cupying areas previously covered by dunes (Figs 2 &  3).
The area of open water has decreased since 1937, mainly owing to the reedswamp encroachm ent and the expansion of the Lagoon fringe vegetation.  (Table 1  & Fig. 3).

Sedimentation
The sedim entation has been heaviest in the upper zone o f the Lagoon. The dense P. australis stand in the middle zone probably acts as a sediment trap partly protecting the lower section of the Lagoon from sedimentation.
If sedim entation is not controlled, the accelerated filling up o f the bed will continue. This could even tually lead to the water breaking through and form ing another bed.

Prognosis
Because o f sedim entation, the colonization by P. australis Reedswamp and Hibiscus tiliaceus-Barringtonia racemosa Lagoon Fringe Forest of most o f the Siyai Lagoon, except the immediate m outh zone, can be expected probably before the turn o f the century. Any management action tending to diminish sedi m entation, e.g. erosion control upstream , should be encouraged. Reeds could be controlled locally by repeated underw ater cutting. The opening o f a new river course may occur in the long term owing to fill ing up o f the existing Lagoon bed. Dredging and mechanical control of vegetation will probably become necessary if m ajor open-water spaces are to be m aintained at the Siyai Lagoon.