Increasing human pressure has presumably led to a decrease in the cover and herbage yield of Kuwaiti desert vegetation, but, to date, there has been little detailed study on such human impacts. A study of Rhanterium epapposum (local name arfaj) and Haloxylon salicornicum (local name remeth) steppe was therefore effected to determine the seasonal variation in above-ground phytomass and percentage cover, and to investigate differences between protected and adjacent grazed areas.

An average seasonal precipitation of 90 mm supported a mean of 223 kg ha-1 in arfaj steppe in 1979–1989, whereas an average mean seasonal precipitation of 73 mm during 1983–1989 maintained a mean phytomass of 102 kg ha-1 in the remeth steppe. Annual forbs and perennial shrubs were the greatest producers of dry matter per kg of phytomass in the arfaj and remeth steppes, respectively. The seasonal production of dry matter was related directly to the seasonal precipitation in the arfaj steppe, whereas the remeth steppe did not show an obvious relationship to the precipitation. The plant cover was 83% and 70% less, and herbage production was 76% and 91% less in grazed areas than in protected areas in the arfaj and remeth steppes, respectively.

Towards the end of the Gulf War in 1991, the retreating Iraqi forces destroyed numerous oil installations in Kuwait, causing widespread oil pollution to extended areas of the desert ecosystem. Vegetation development in an oil-contaminated area of northern Kuwait, where the natural vegetation is dominated by the dwarf shrub Haloxylon salicornicum, was studied seven years after the release of the oil. Some sites of the study area were largely unaffected, whereas others were contaminated to varying degrees by oil. Tar-like oil tracks accounted for the largest proportion of contaminated ground, and these remained largely unvegetated. However, a number of Haloxylon shrubs survived the oil contamination mainly due to the presence of phytogenic hillocks around their bases. These phytogenic hillocks provided ‘safe sites’ for a number of plant species. This also applied to blow-outs, former phytogenic hillocks on the oil tracks that had been subject to severe sand deflation in recent years. Species composition on both the phytogenic hillocks and in the blow-outs was very similar to that of the control area. Laboratory studies showed that the seed bank under the oil tracks had been annihilated. The number of seedlings emerging from soil samples was lower on the phytogenic hillocks and blow-outs than in the control areas. We conclude that recolonization of oil tracks will gradually take place as the hard surface of the tracks begins to disintegrate, or in part becomes covered by sand. The break-up of the track surface has already begun to a limited extent, mainly due to factors such as off-road driving, large herds of grazing animals, burrowing animals (lizards, rodents) and colonies of ants. It is suggested that a specific programme aimed at breaking up the hard surface and allowing it to become mixed with uncontaminated sand would probably greatly enhance recolonization.