|Expedition to the Foret de Day, Goda Masif, Djibouti, 2006: Baseline survey of the Cloud forest of Juniperus procera
In February 2004 I acted as plant ecologist on a three man expedition to the Juniperus procera forest of the Goda Massif in Djibouti. The primary aim of the expedition was to investigate the current status of the critically endangered Djibouti Francolin, a small partridge like bird, with a view to encouraging its conservation. My job was to report on the vegetation. A subsidiary aim of the expedition was to report on the Bankouale Palm, an endangered palm found in the area. The project was funded by the World Pheasant Association, and was organised in conjunction with the Government of Djibouti. This project is being organised as a direct response to the conditions we found there and which we have reported on in a paper in Palm, the journal of the Palm Society, in three papers to the BES meeting in September 2004, and one which we have submitted to the New Scientist.
Djibouti is situated at the southern end of the Red Sea, and forms the northernmost land of the Rift Valley. The Goda Massif is a range of hills of horizontally layered volcanic rocks, 1700 m high on the north side of the Gulf of Tadjourah, and is home to 70% of the land based plant biodiversity (Audru et al., 1998). The massif is the eastern end of a plateau which slopes gently to the dry interior to the west, and is dissected by a number of steep sided wadis to form a very complex landscape of ridges and valleys, with small areas of high plateau. The valley sides are typically covered with a canopy of Terminalia forest, with areas of Acacia woodland and grassland. Throughout are found Jujubes, Zizyphus spp., the African Olive, Olea africana, Bankoua, Lannea triphylla in some areas. Various Ficus species line the wadi bottoms,and strangling figs can be found scattered throughout.
On the upper slopes and covering the plateaus on the summit is an area of cloud forest of Juniperus procera forming a single species canopy covering a total of some 20 square kilometers. It has an understorey of Buxus hildebrandtii and Tarchonanthus camphoratus. This juniper can be up to 20m high, and some of the trees we measured had a diameter at breast height of over 80 cm, making them probably over 400 years old.
This area collects most of Djibouti's rainfall. There are few data for Djibouti, but 10 years of data were collected in the 1970's in nearby Randa showing that in some years 25 cm. of rain fell. elsewhere in Djibouti the rainfall can be as low as 18 mm. It is clear however that the Juniper acts as a typical cloud forest by trapping the cloud which falls as rain from the canopy. The forest has no doubt acted as a refuge for the nomads during the hot summer months for many years, but it has taken on a new significance in the last 50 years since the border with Ethiopia was closed by the then Emperor, the transhumance routes closed, and the nomadic lifestyle changed to a more sedentary one. This has had a twofold effect of increasing grazing pressures and of increasing the utilisation of the scarce water resources for permanent terraced gardens. Nevertheless the Forest is advertised as a haven of greenery in the the desert of Djibouti. A report by Welch and Welch (1998) states that the forest was in adequate health, though commenting on the lack of regeneration.We were therefore very shocked to find on our arrival that the forest was dead.
When we arrived at the forest we were faced with a view of skeletal trees stretching as far as we could see. Whilst some cutting may have taken place in the last 100 years, none was evidenced on this trip, and grazing will only affect regeneration. We looked for evidence of fungal attack, and found very little. We saw no evidence of insect attack. Our counterpart. Houssein Rayaleh informed us that this had all taken place during the last 4 years. As we explored the forest thoroughly we discovered that perhaps 60% of the forest has died, with a further 30% of the trees being moribund,i.e. at least 75 % of the tree being dead. Of the remaining 10 % a grove of young trees exists at Mandah in the far east of the Massif whilst the rest are individual trees scattered throught the upper slopes of the wadis
We plan a four week expedition to the same area to survey the forest in sufficient detail to provide a base line for any work in the future, and to investigate the causes of the decline of the forest.
The basis of the survey will be a series of stratified random 100 sq.m. quadrats ensuring that the three main dead areas are covered, Godah, Barra Barre, and Mandah, the moribund area of South Godah, and the slopes and Eastern area of Mandah, where there are still some living trees. With 3 weeks in the field we should be able to collect data from 80 quadrats. Whilst we can manage 8 quadrats a day in the UK woods, this terrain is quite different, all equipment will have to be carried by ourselves or on camel. We hope to visit Randa for 2 days to complete the count of the Bankouale Palm initiated earlier this year.
The Quadrat data:
The quadrat data will conform to English Nature standards: a 10x10 m quadrat, with the inner 4 sq m being used for a count of the ground flora, the inner 6 for a count of the shrub layer and the whole for a count of the canopy species. The quadrat axes will be aligned on a north south axis and the ne and sw quarters will be used to record diameters at breast height of the canopy species Each will be accompanied by digital photographs of the ground flora and the quadrat from the northern corner, a gps measurement from the quadrat centre, a 500 gm soil sample and pH reading taken in the field, together with basic physical data to include altitude, slope, orientation.
On our return The soil samples will be analysed to record nutrient levels and these will be combined with the quadrat data and analysed using a standard principle components analysis and a reciprocal averaging program to reveal underlying trends in the vegetation.
We will take cross sections of at least 10 recently dead junipers to try to relate the diameter of the trees with age. This is difficult with Juniper as not only do they grow slowly, they are capable of not growing at all under drought conditions, thus only an approximate ageing can be attempted. However this will be combined with the DBH data to analyse the population structure of the forest at the time of its death.
Seed bank :
The soil samples will be investigated to discover if there is any seed bank for the juniper. This will be crucial information if there is to be any hope of regenerating the forest.
Professor Colin Prentice, Bristol University Earth Science Department has agreed to help in preparing climate models for the area as the available weather data is very scant.
Apart from the reports required of us, we will be publishing the data appropriately in peer reviewed journals and providing the Government of Djibouti with a copy of all our results, including a GIS based database of the results linked to a taxonomic database of the species in the locality. There is no university in Djibouti but our counterpart, Houssein Rayaleh is a keen naturalist and the importance of the forest is well understood by him and those in Bankouale. It is hoped to follow the project up
Houssein Rayaleh, of the Department of Tourism and Development, facilitated the previous expedition and is keen to have us return to finish the work. He would take care of visas and entry requirements and would participate in some of the work with us as he did before. We would be based at the Bankouale Camp, run by Houmet Ali, where the previous expedition was based , and Houmet would look after the logistics of the expedition. I hope to be accompaned by Harriett Gillett of the Plant Division of the World Conservation Monitoring Centre, Cambridge.