limited control or ownership of resources and revenues. Percentage of land area owned by women ranges from none in Oman, about 5% in Syria and Lebanon, 11% in Jordan and 24% in Egypt. Statistics about asset entitlement and access are scanty and rarely separated by gender. Generally, women own plots smaller than men's. In Syria, 7% of women own animals and 1% own agricultural machinery. According to the report of the Convention on the Elimination of All Forms of Discrimination Against Women in Yemen, female farmers do not control land, water, agricultural equipment, credit or capital (CEDAW, 2001).

Some critics have highlighted the reasons for the underrepresentation of women's involvement in agricultural work, particularly in Muslim countries (El-Fattal, 1996). These are the association of women with domestic spaces to the exclusion of outdoor activities, such as work in fields; association of agricultural work with wage labor, while women are mainly unpaid; and association of farmers with plot holders-women working in the fields are rarely landowners. Many land and agrarian reforms have increased concentration of property entitlements, access and control in the hands of the male heads of households and assigned access to men for basic agricultural resources, such as water, seed and fertilizer, distributed by government agencies. Lacking control over, and entitlement to, production often implies restricted access to loans and social security, limited autonomy and decision-making power, and, eventually, curtailed ability to achieve food security. Women's limited access to markets also affects their control of revenue and decisions.

The increasing number of female-headed households, visible in many countries of CWANA, corresponds to an increase in women's workload and a decrease in their independence. In Pakistan and Sudan, 25% of households are headed by women, 16% in Egypt and Morocco, 13.6% in Yemen, 11% in Lebanon, Oman and Tunisia and 6% in Iran, Jordan, Syria and Turkey (FAO, 1995; Hartl, 2003).

In some cases, women become empowered because men are absent. They participate in decision making by managing small budgets and their mobility is increased because they sometimes go to the market to sell their produce, even if they still must let male relatives make major decisions, such as selling a cow (CNEA, 1996). The feminization of agricultural labor increased the rates of women's work in the

unpaid and informal systems. Their employment in wage labor is still characterized by gender wage differentials, precariousness, lack of social services and vulnerability. The increase in household work involves children, who contribute their share of work, to the detriment of their school attendance, free time, health and other children's rights.

Finally, despite women's increasing involvement in the fields, agricultural machinery is still usually designed for men, limiting, together with social biases, women's and children's access to technological improvements. In Syria, the introduction of agricultural machinery from the 1960s often increased women's and children's drudgery by strengthening the gender division of labor. Men were assigned mechanized work, leaving manual work to women and children.

1.3.2 Agricultural land use

A detailed examination of the aridity zones shows that over 4 million km2 of land in CWANA is available for good cropping and animal husbandry (Table 1-5). The greatest land use is permanent pasture, 550 million ha. Cropland and forests are 141 million ha and woodlands 124 million ha. Most of the permanent pastures and forests and woodlands are in the Nile Valley, Sudan and Somalia. Central Asia and North Africa are also rich in croplands and permanent pastures. Among the low-income countries, per capita accessibility to arable land is about 1 ha, of which 40% is irrigated and 60% rain-fed. The estimated daily income of an average farming family is about US$2.82 (Rodriguez and Thomas, 1998). By 2025, the lowest per capita cropland-less than one-tenth of a hectare-will be in the Arabian Peninsula and the Nile Valley. There is great disparity in cropland per capita among countries within a subregion. For example, cropland per capita in Sudan, 0.448 ha, is eight times higher than in Egypt, 0.053 ha. These are expected to be reduced to 0.0313 ha for Egypt and to 0.277 ha for Sudan by 2025, if current trends of population growth and land degradation continue.

There has been no appreciable overall increase in cropland in the last three decades. In 2005, the area of arable and permanent crops is 5.6% in North Africa, 10% in Central Asia, 1.9% in the Arabian Peninsula and 16.3% in the Middle East (FAO, 2006b). Cereal production increased slowly in North Africa, about 50% between 1975, 18.5 million tonnes, and 2005, 28 million tonnes. Central Asia had an