2.3 The Land Unit Map and its legend

2.3.1 Phases of work

The three main phases that led to the land unit map of the study area were the following:

• Bibliographical study, preliminary legend, and map;
• Survey field activity;
• Final land unit map and legend.

Bibliographical study, preliminary legend, and map

A first bibliographical study of the study area was performed to obtain a general overview of geography, climate, geology, geomorphology, agriculture, vegetation, and soil characteristics. The geographic limits of the study area were traced using the topographic map at 1:50,000 scale and the 1994 "Carte de l’Inventaire Forestal National" at 1:100,000 scale from the synthetic report by Ministère chargée des Eaux et Forets (1999).

Later on the analysis of the geomorphologic aspects of the area was carried out; the information from geological and topographical maps and from satellite images processing led to characterise two main landscape systems: one located in the southern part of the area and the other in the northern one. These systems were named as: the "upper part of Ben Slimane plateau" and the "lower part of Ben Slimane plateau". This last one was additionally subdivided in the "summit area" and the "scarp area".

During the preliminary photo-interpretation nine different landform types were identified (Table 5) as well as eight land covers (Table 6): four in natural vegetation and four in human influence areas. In particular, in both landscape systems, different gentle slope classes were identified and only in the lower part of the Ben Slimane plateau the natural managed vegetation was present.

According to the different characteristics in terms of landscape, landform, and land use, the preliminary map was realised by interpretation both aerial photo-interpretation, the different land units and assigning an identification code to each of them. These codes, with their description and frequency, integrated the descriptive information in terms of landform, land cover, and land use therefore completing the preliminary legend.

In the southern part of the study area, for which it was not possible to have aerial photos, we proceeded directly with the interpretation of satellite images using several band combinations and fusion processed as subsequently described (paragraph 2.4). For the south-eastern zone, information coming from NDVI analysis of January, April, and August 2000 and from the 1994 "Carte de l’Inventaire Forestal National" on the scale of 1:100,000 were used.

The same colour composites and fusions have been finally used as background for video digitalisation of the areas delimited with the aerial photo-interpretation, in this way obtaining more information specially for further unit subdivision based on soil colour. The preliminary map, with the data of roads network and paths, has helped to plan the field activity, localising one or more relevés for each preliminary legend code taking into consideration their frequency and surface.

Field survey activity

During the period from the 25^th of March to the 25^th of April 2001 the field activity was carried out in the study area.

The total number of realised relevés was 93; for each of them geographical co-ordinates through GPS, slope steepness, and slope length were measured. Through the indications derived from the topographical map, the relative information for location, elevation, elevation range, aspect, and slope position (Schoeneberger et al., 1998) were recorded. In addition, through direct observations land cover (Di Gregorio and Jansen, 2000), land use, coarse fragments (area and size), and human influence data for the most representative facet in the unit in terms of morphology, vegetation, and soil were collected with a brief description. In the Annex 4 some examples of filled forms, correspondents to the different characteristics of the relevés, are visible.

Geology and geomorphology

Data for type, area, activity, and degree of geomorphological processes when present were collected. Where feasible, indications on lithology, the nature of the degradation processes, and characteristics of parent material were reported (modified from McDonald et al., 1998).

Vegetation

The sample areas for data collection of natural vegetation were randomly located inside the land unit with a 10x10m or 20x20m transect. The total percentages of woody, shrubs, and herbaceous species cover were indicated together with a floristic list with species names, height, phenology, vigour, wood collection, regeneration, and grazing evidence. For the areas with prevailing agricultural activity, data relating farm were collected such as: total area, technical knowledge, land improvement, crop distribution and, rotation, watering points, species and, number of animals, crop information such as type, variety, stage, seeding and harvesting date, conditions, cultivation practices, average yield, and other information from the owners.

Soil

In the units most representative points, soil profiles were realised and described according to the guidelines of F.A.O (1990).

Fields relating the description of the site surface and the features of each horizon were present in the form, in order to collect data about: surface coarse fragments, erosion, sealing, cracks, salts, micro-topography, hydrological characteristics, depth, moisture, colour, rock fragments, texture, structure, consistency, pH, roots, cutans, mottles, efflorescence, nodules, voids, and biological features.

Soils were also sampled for the subsequent laboratory analyses.