**1. Introduction**

The cocoa tree (*Theobroma cacao* L.) belongs to the Malvaceae family and is native to the tropical rain forests of Central and South America [1]. The Germans first introduced cocoa in Africa through Ghana in 1857 and in Cameroon precisely through Victoria (Limbe) in 1886 [2]. The nutritional and pharmaceutical importance of cocoa makes it one of the main export products for certain tropical countries. In Cameroon, and particularly in the center region, Cocoa is still grown in the traditional Agroforestry way with the shade of some forests, fruits trees and oil palms [3–5]. These systems have been widely described in relation to the environmental and ecosystem services they provide, but much less regarding productivity related to their structure [4, 6, 7].

Soil fertility under tropics is mainly influenced by biological interactions, and trees are the main driver of these, as they provide the rough material to achieve these [8]. Bellow and above grounds interactions significantly influence the status of soil health through their rooting habit, but also through the decomposition and mineralization of the litter fall. Several studies have demonstrated that the reduction of nutrient and organic matter content in the soil is a serious threat for agricultural production and food security in many tropical countries [9, 10]. Research has been involved in this theme to understand the mechanisms of conservation and improvement of soil fertility by trees. Numerous studies have identified links between traditional knowledge of trees associated with cocoa agroforest and soil health (fertility), although some processes are difficult to codify [8, 11, 12].

Although not always recognized by agronomists, trees in cocoa-based agroforests have more uses for local farmers than just providing a suitable microclimate for cocoa trees [3, 8]. Some indigenous species are maintained in the system by local farmers for their fertilizing capacities through nutrient recycling. Such tree species most mentioned as positively influencing fertility of soils and/or having other desirable attributes in traditional land used as described by [8, 11] include species such as *Ceiba pentandra* described to have floral and leaf litter fall that improves soil upon decomposition in rainy season, gathers dew in the dry season and prevents soil from drying, woody parts decompose rapidly and add to fertility, soil around always wet; *Milicia excelsa* described to provides good shade, improve microclimate, leaf litter improve soil conditions, deep rooting habit, abundant leaf litter and high leaf litter decomposition rate and soil fertility value; *Alstonia boonei* have *a* deep rooting habit, tall tree, wide and open crown, intermediate leaf area, abundant leaf litter with high decomposition rate, provides good shade and maintains soil moisture, gathers dew/exudes water, fertility value; *Ficus mucuso and Ficus exasperate* described to possess deep rooting habit, tall tree, wide and open crown, big leaf area with high leaf litter decomposition rate, high fertility value etc. Yet, little information is available on the contribution of these species leaf litter nutrient to the productivity of soils under cocoa fields. It is therefore hypothesized that trees species in cocoa-based agroforests play a major role in the improvement of soil conditions, hence the productivity of the entire system.

Appropriate tree species selection based on nutrient cycling is a vital issue in agroforestry practices [13]. So far, the screening or prioritization of commonly present indigenous trees species of cocoa-based agroforestry systems (CBAFS) is based on ethnoecological and ethnopedological studies [11, 12] but a scientific approach was carried base on the rate of mychorrizal colonization of the roots of some of these indigenous tree species. The results established a positive correlation between local farmer's classification of ten indigenous tree species with high fertility potential, and the colonization of the roots of those by mychorrizal [11]. However, no or little attempt has been taken to assess the fertility potential of these trees in term of their nutrient content present in their leaf litterfall, hence the need for this study.

This study, therefore, aims to investigates farmer's perceptions and knowledge on the fertilizing potential of the leaf litterfall of the non – cocoa tree associated to cocoa in CBAFS. If the pertinence of the perceptions is established, such measures

could be introduced in management plans for cocoa- based systems in order to enhance their ecological and economic performance.
