In conducting the survey we focused on the two most basic parameters in pond aquaculture: soil quality and water supply. We did not consider it necessary to conduct biological sampling, e.g. for fry abundance, since the project KEN/80/018 has been, and still is, conducting such surveys as part of its programme. Furthermore, it is felt that fry availability within a given locality, while convenient, is not absolutely essential since these can be transported from one area to another without much difficulty or can be produced in a hatchery. Neither was the natural fertility or productivity of any given area considered, since in pond culture it is standard practice to use fertilizers to enhance productivity or provide supplemental feed where such productivity is inadequate. A different set of criteria was used for identifying possible hatchery sites.
An area is considered suitable if its soil can be used to construct structurally sound and reasonably impermeable dykes as well as pond bottom without resorting to the use of extraneous materials such as clay or synthetic water-proofing membranes, and if it can be watered up to a depth of at least 0.75 m at mean high-tide level without pumping (suitable for extensive systems), or without entailing a high energy cost for pumping (suitable semi-intensive and intensive systems).
It should also be noted here that in Southeast Asia and Taiwan where shrimp farming is most widespread, it is considered important to consider the adequacy of fresh-water or low-salinity water supplies in order to dilute the seawater to a level considered optimum for growth, particularly for P. monodon. This issue can be considered academic in Kenya due to the limited number (2) of permanent rivers in existence.
Furthermore the technology developed and practised at the Ngomeni site which served as the impetus for this survey uses undiluted seawater which even rises to the hypersaline level of 50 ppm or higher after some time in the pond due to evaporation. At this level the Ngomeni ponds have been able to realize yields mainly of P. indicus of as high as 335 kg per ha. However, yields of less than 100 kg per ha were not unusual. It is not possible to categorically quantify the Ngomeni pond performance due to the extremely variable stocking densities (from 10 000 to 72 000 per ha) and culture periods (77–119 days). The small proportion of P. monodon which occur (five percent or less) have been observed to grow to an average of 30 grammes.
Hydrological parameters such as pH and dissolved oxygen normally collected for most site studies were dispensed with since these factors when measured in clean, open waters will not at any rate reflect the level that would prevail once the water is confined in ponds and has interacted with the soil and with pond organisms. Furthermore in the case of the present survey, the water supply is invariably undiluted seawater which is known to be a buffer solution whose pH level is slightly alkaline.
All possible areas along the coast were examined regardless of present day accessibility, usage and ownership status. The objective was to obtain a comprehensive picture of the Kenya coast on a macro level enough to classify whole areas as either suitable or unsuitable. Should any shrimp farm be established in any of the areas identified however, it would still be necessary to subject it to a more detailed study for planning and designing purposes.
We were greatly guided by the excellent topographic and soil and surface geology maps produced by the Survey of Kenya and Kenya Soil Survey. Extensive use was made of these charts which are produced at a scale of 1:50 000.
Coastal areas which were shown by the charts to be high ground, sandy or rocky beaches were no longer included in the field survey. Oftentimes the nature of these coastlines were visually confirmed while travelling from one point to another.
The areas considered interesting are those indicated in the topographic maps as mangrove or open tidal flats within or behind such mangrove areas. These areas are described in the soil maps as “poorly drained, very deep, very dark greyish-brown, strongly saline, humic material” in the case of mangroves. The open flats are described in the same soil map as “poorly drained, very deep, yellowish-brown, mottled, loose, moderately saline, sand to sandy loam”. These two types of areas were investigated through actual site visits. Some areas indicated as seasonal swamps in the charts but adjacent to mangrove areas or low enough to be reached by spring tides were also examined.
Where possible, the target areas were approached both from the shoreline with the use of a boat as well as from the high ground. Actual penetration of the mangrove areas, of course, could be done only on foot at low water when such areas are exposed. In some places, the area to be surveyed could be accessed only from the high ground due to the lack of a watercraft or direct access to the shoreline. In many other places, the area can be approached only from the sea due to the lack of roadways close to the coast.
The field survey routine consisted of taking the ground elevation using a leveling staff and an Abney hand-level with the water level as the baseline. The time at which the reading was made was noted so that the level of tide can be reckoned from the tide table published by the Kenya Ports Authority. Where access to the water edge was circuitous or the line of sight obstructed by heavy foliage, the ground elevation was reckoned by measuring the level of the high water marks on the mangrove tree trunks. Such visible marks were interpreted to reflect the mean high-tide level of water.
Upon penetration of an area, open pits were dug with the use of a shovel in order to obtain the soil profile down to a depth of at least 60 cm. In some cases where the soil profile looked interesting the test pits were dug down to 100 cm. Excavations were made both in forested and clear open access. Soil samples were taken at about 5 cm below the surface. However, in areas with marked statificatoin, a second sample was taken from the next layer. Soil samples were no longer taken in areas which were obviously unsuitable for pond development after a cursory examination, such as for instance, areas where the substrate is loose sand with no evidence of clay or loam material whatsoever even 50 cm below ground level.
The samples were brought back to Malindi where they were dried and powdered. These were then subjected to manipulative tests in order to characterize each sample. Soil pH was taken using a portable digital pH meter (DKK, HPH-22) on a solution of one part powdered soil with two parts water.
The hectarage of the areas identified a suitable were later estimated by using the topographic map. At a scale of 1:50 000, one cm2 on such maps is equivalent to 25 ha.
In undertaking the survey we subdivided the coastline into five functional areas. We later found these areas to coincide very closely with the five administrative districts into which the coastal province has been subdivided, namely: Lamu, Tana River, Kilifi, Mombasa and Kwale (see Figure 1). We have therefore adopted the boundaries of the administrative districts in grouping the various locations surveyed for the purpose of this report. The location of sampling stations (Figure 2 and 3) were established by using clearly identifiable land marks such as points, creeks, jetties, etc.
