The pastoral area of Niassa, created in 2000 under Joint Decree No. 2000-37/MRA/AGRI/MEE/MEF/MATS/MEM/MIHU of 21/7/2000, is located east of the Department of Gogo, province of Zoundwéogo [Ministry of Animal Resources (MRA), 2006]. It lies between latitude 11°30 35′ and latitude 11°39′13″ N and longitude 0°47′25″ and longitude 0°52′54″ W. It is bounded on the east and north by Lake Bagré, on the south by the Koulwoko river and on the west by the territories of Samtenga, Yirpala, Kopelin and Nagrigré.

It covers an area of 6,386 ha divided into two (02) separate blocks separated by an agricultural front approximately 7 km wide. The first block covers an area of 4,374 ha and the second block, located to the north, covers an area of 2,012 ha (Figure 1). It is made up, on the one hand, of areas known as “smallholding,” which are made up of housing, food crops and fodder crops. Each smallholding should have a surface area of around 5 hectares for a pastoralist based in the pastoral zone. There is also a pasture area for livestock. The occupancy rate of the total area within the pastoral zone should be no more than 15% for the housing and food crop zone, and at least 85% for the pasture area. To set up in the zone, you need to be a livestock farmer with at least 20 head of cattle and a license.

This zone constitutes a privileged ecosystem to support pastoralists in their quest for quality and quantity of pastoral resources. By creating this pastoral zone, the Country aimed to rationalise access to pastoral resources, to promote the adaptation of pastoralists to the effects of climate change while promoting peaceful cohabitation of agricultural and livestock activities. However, these areas have been depleted over the years due to the lack of fodder supply in terms of quantity and quality, colonization by croplands and landowners’ questioning. This undermines the original objectives assigned to them.

For a study of the dynamics of land use, satellite imagery is an important element for describing and monitoring environmental changes over time and space. This study used Landsat satellite imagery from three (03) years: 2000, 2010 and 2020, i.e., a depth of 20 years. This gives a clear picture of changes in land use and, consequently, changes in vegetation cover.

The following materials and tools have been used:

− Six (06) wet-season satellite images were used at a rate of two images per year. These are Landsat images of October and November selected based on their availability on the site and their quality in the study area. The spatial resolution of the images is 30 m. The October and November periods were targeted due to low cloud cover, good vegetation expression and low absence of bushfires. These images come from Landsat ETM+ (Enhanced Thematic Mapper Plus) sensors for the years 2000 and 2010 and OLI_TIRS for the year 2020 (Table 1). They were acquired free of charge from the National Observatory for Sustainable Development (NOSD) through its Division for Capacity Development, Information and Environmental Monitoring (CDIEM). Vector data were also mobilized from the General Directorate of Pasturelands Management (DGEAP) for mappings.

This component was conducted through a survey in the three (03) villages bordering the pastoral zone as well as in the farming hamlets established in the pastoral zone. These were the villages of Kopelin, Nagrigré and Samtenga. The farming hamlets are Tansablego, Kopelin pastoral zone, Wapassi, Zomnongo, Mbouta, Sao, Ragordgande and Ragordguibi. These villages were chosen based on the following criteria: proximity to the pastoral zone, local people’s support for the creation of the zone and their use of pastoral resources. Here, a household is made up of one or more people sharing one or more fields of crops or herds of cattle. Households were selected using a purposive sampling method based on the following criteria: 1) ownership of cattle and small ruminants for livestock farming, 2) consent of the household to be surveyed, 3) at least 30 years of residence in the study area to provide a history of phenomena and changes for pastoral resources since the creation of the pastoral zone in 2000. The survey was conducted in two (02) phases, the first phase took place from 03 April to 01 May 2022 and the second phase from 27 March to 15 April 2023. In some 196 households were surveyed on the basis of consent (Table 2).

It was conducted using an interview guide and a questionnaire through Kobocollect application. They consisted in seeking herders’ perceptions of the evolution of pastoral resources in the Niassa pastoral zone over the past 20 years. The aim was to get their opinion on pastoral resource trends. The various data collected from pastoralists and agropastoralists covered 1) the socio-economic and professional characteristics of households, such as level of education, main activity, secondary activity and sources of income; 2) the perception of pastoralists and agropastoralists of changes in natural resources, such as plant species that are becoming extinct or have appeared in the area, and changes over time in fodder and surface water resources. In addition, data on how the area is managed have been collected.

The Excel spreadsheet and SPSS version 21 software were used to process and analyse the collected data. As such, the proportions of respondents by question were calculated using SPSS and the tables and figures were produced using Excel.

The herders interviewed belong to two (02) ethnic groups, namely, Fulani (67.27%) and Mossi (32.72%), with 2.3% women and 97.7% men. Natives account for 20% and non-natives 80%, who come from localities in the Centre Nord, Plateau Centrale and Centre regions. Depending on the type of main or secondary activity carried out, there are two main actors: pastoralists who account for 69% of the respondents and agropastoralists 31%.

The age of those interviewed is between 32 and 80 years. The most important age group is between 41 and 50 years. It represents 35.5% of respondents and is followed by the 30–40 age group, which represents 25.5% of respondents (Table 3).

In terms of educational attainment, 50% of respondents attended Koranic school; 27% were able to take literacy courses in Fulfulde or Mooré; 5% reached primary school and very few reached middle school (1%). However, 17% did not have access to any of these types of training (Table 3).

Analysis of Table 4 shows that pastoralists and agropastoralists have several sources of income. Indeed, 100% of pastoralists and agropastoralists sell cattle as a source of income; 100% of agropastoralists and 80% of pastoralists sell small ruminants as a source of income. But the difference is that the sale of milk is much more reserved for pastoralist households (over 80%), whereas agropastoralist households engage in the sale of agricultural products and the small-scale trade.

All the respondents (100%) have cattle (Table 5). Ownership varies with the type of farmer. The analysis of Table 3 shows that agropastoralists have fewer number of animals than pastoralists whatever the type of species. On average, pastoralists have 54 ± 26 cattle, 32 ± 12 sheep and 36 ± 16 goats. Agropastoralists have an average of 10 ± 4 cattle, 25 ± 8 sheep and 15 ± 8 goats.

As far as breeds of livestock are concerned, Fulani zebu is the only breed of cattle owned by farmers. For the other species, Figure 2 shows that several breeds are bred. Mossi goats, which account for 87.2% of goats, are the most popular breed, while Mossi sheep are the most popular breed, accounting for 78% of sheep.

Analysis of Table 6 showing changes in livestock numbers in the zone from 2012 to 2021 shows that the number of animals has increased significantly

Since the creation of this zone, it has been noted that the stakeholders in charge of technical management have not implemented the project to plan the distribution of the occupation of the zone, as there was no management tool. It was only six (06) years after its creation in 2006 that the management plan and specifications were drawn up. And a management committee has been set up to ensure compliance with the rules laid down in the specifications and the management plan. This observation reveals a shortcoming since its creation, because farmers have settled there without any basic rules being established, without any management plan.

Although management tools have been developed, the herders do not scrupulously respect the management rules governing access to the area and pastoral resources. According to the results of the survey, only the rules laid down for the management of water wells and vaccination pens are respected by the livestock farmers. Respondents were unanimous in stating that use of the area’s resources is free and open to all. In fact, 98.8% of those questioned said that they had access to pasture areas, compared with 1.2% who were sceptical, blaming the uncontrolled installation of housing and farms that reduce the pasture area. Cattle tracks are opened to facilitate internal and external livestock movements. However, 57.7% of the farmers interviewed said that these access corridors are not functional, especially in the rainy season when they are obstructed by crop fields.

Most of the farmers interviewed (87.95%) acknowledged the existence of a management committee for the pastoral zone but felt that it did not play its role properly. They pointed to a number of shortcomings of the management committee: 1) the lack of organisation and cohesion between a management committee members, 2) the lack of a framework for discussion with producers, 3) the lack of supervision of the installation of livestock-raising households in the zone, 4) the lack of a mechanism for receiving and installing transiting pastoralists, and 5) the lack of awareness-raising, monitoring and repressive measures. The interviewees felt that all these shortcomings were exacerbated by the lack of firmness and support from the authorities in charge of livestock farming. And secondly, with the advent of terrorism, where members of the management committee and surveillance units are regularly threatened.

The results of the satellite image processing enabled us to distinguish six (06) land use classes in the Niassa pastoral zone. These are woody savannahs, shrub savannahs, gallery forests, croplands, water bodies and bare land. From analysis of Table 7, the Kappa indices obtained for the three (03) treatments are respectively 0.92 for 2000, 0.97 for 2010 and 0.90 for 2020.

Table 7 shows that the overall accuracies of the confusion matrix are 93%, 97% and 93% respectively for the years 2000, 2010 and 2020. This proves that over 90% of the pixels in the three (03) images were correctly classified. In short, the values of the various supervised classification accuracy indicators obtained for the different images reflect the good quality of the classification results.

Figure 3 and Table 8 present the situation of the units of land use in the pastoral area. According to the reference period of 2000, the pastoral area is dominated by the woody savannah, the cultivation areas, and the gallery forest. These three (03) classes occupied more than 86% of the study area with 54.68% for woody savannahs; 15.54% for forest galleries and 15.35% for cropland. While in 2010, it was dominated by croplands with 36.01% followed by woody savannahs (28.44%) and forest galleries (17.39%). Likewise, in 2020, these units maintained the same ranks of occupancy in varying proportions.

The analysis in Table 9 shows that:

- Croplands classes increased significantly by 134% over the period 2000-2010. Similarly, over the period 2010–2020, they increased but less than in the first period (9%). Overall, they have more than doubled their initial surface area, with an increase of 155%.

Table 11 gives an overview of the plant species that have appeared in the pastoral zone according to the opinion of the farmers. Nearly all respondents (96.5%) deplored the invasion of the area by plant species that are not or are not eaten very well by livestock. Among these invasive species are Senna obtusifolia (L) (100%) and Hyptis suaveolens (L) (90%), which would not be very appetising to livestock according to respondents. In addition to these two species, they reported cases of Loudetia togoensis (95%), Microchloa indica (85%), Zornia glochidiata (95%), Sida acuta (45%), and Cymbopogon schoenanthus (45%).

As for the way in which they spread, they state that these species were spread by zoochory according to 87.9% of respondents, by soil impoverishment according to 61.6% of respondents, by their non-palatability by livestock (56%), by hydrochory (11.6%) and by anemochoria (9.3%). In fact, some of these species like Senna obtusifolia (L) and Hyptis suaveolens (L) have developed thanks livestock refusals. This action allowed them to have a capacity for rapid growth compared to other species. In addition, they have high germination power and abundant soil seed stock. These species appear as soon as it first rains and, because they are not grazed, they gain the upper hand over other species by inhibiting their growth.

As far as woody fodder species appearing in the area are concerned, pastoralists have noted the presence of Acacia seyal (Del) and of Acacia nilotica (L.)

Figure 5 shows the various factors causing the degradation of fodder resources. Thus, The causes of the disappearance of forage or utility species are many and vary from one species to another. As such, pastoralists and agropastoralists blame the following factors for the disappearance or drastic decline of the appetized herbaceous fodder species. According to 83.7% of respondents, this was overgrazing. The reduction in grazing areas and fodder supply leads to a rush into residual fodder, thus creating an overload of pasture. This does not allow rapid regeneration of appetized fodder species due to the intense rhythm of the pasture but favours the development of non-appetized species. As a result, there is a gradual decline of appetite species and sometimes degradation of soil and vegetation cover with the appearance of bare soil as a result of settlement. They also mentioned the fall in rainfall (66.3%) and pockets of drought (55%) that hinder the development and growth of fodder species during the rainy season, hence the decrease in fodder supply in the dry season. Also, land clearance (24.4%), poor agricultural practices (8.1%) and bush fires (4.7%) contribute to the scarcity of fodder species. For them, the lack of land is the cause of the decrease in the rainy season pastures, which leads to overcrowding in non-agricultural areas. As regards poor agricultural practices, it is mainly the use of herbicides that no longer favour the exploitation of post-harvest weeds. They report that bushfires have become rare, but as soon as they occur, the damage is enormous in terms of residual forage losses in the dry season.

In terms of the disappearance or low presence of appetized or utilitarian woody fodder, farmers blamed the following factors: the decrease in rainfall, the abusive wood cutting, the phytopathology, overgrazing, and the pockets of drought respectively from 75.8%, 66.1%, 22.1%, 12.8% and 11.6% of respondents. Four (04) factors emerge from the analysis of the situation as being the most implicated: overgrazing, reduced rainfall, pockets of drought and abusive wood cutting.

The rates of change show a gradual trend for cropland units of bare lands and water bodies over both periods. The increase in the areas of croplands and bare lands in pastoral areas has been observed by many authors Kima et al. (2016) have observed a decline in grazable areas in favor of croplands and bare lands, in the province of Boulgou; Bambara (2019) has also observed the same trends in the progression of cultivable areas and bare lands in the pastoral area of Sidéradougou and Nouhao. The increase in croplands in pastoral areas is due to the failure to comply with the specification and the development and management plan for the pastoral zone. Indeed, the specification allows each farmer in the pastoral area to exploit up to 5 ha, i.e. 1 ha for his habitat, 1.5 ha for croplands, 1.5 ha for forage crops and 1 ha for fallow land. In general, the gradual trend in cropland units could be explained by the increased need for arable land to meet the high food demand due to the population explosion. In fact, the authorisation and lack of control over the installation of pastoral households has led to an increase in the number and size of households, either through the installation of new households or the creation of households from former households. The increase in the size of the population in the pastoral zone will lead to an increase in the area of housing and arable land for food. In addition, the failure to comply with the specifications has led to the area being coveted by landowners who have set up crop fields. This high demand for land leads to a reduction in the area of pasture by deforestation of the units of woody savannahs, shrub savannahs and galleries forests. Indeed, the work of Powell (2004) showed that the increase in the rate of population has influenced the area of croplands which are growing at the same rate as that of human population, i.e., about 3% per year, and with a slight intensification. As a result, this demographic pressure has led to a high demand for arable land to meet the food needs of the population at the expense of pastures (Kadéba et al., 2019; Dixon et al., 2019). This is consistent with the work of Thierry et al. (2018) in northwestern Benin which explains the high demand for cultivable land and buildings due to the increase in population (Tankoano et al., 2016). The deforestation of these areas allows new fertile land to be taken advantage of and thus increases agricultural production. For some authors, the lack of arable land on the outskirts of pastoral areas and the fertilization of this land by animal excretion could be one of the reasons for its invasion by farmers. As a result, parks or pastoral areas are under anthropogenic pressure to find better land for agricultural activities (Gomgnimbou et al., 2010; Amegnaglo et al., 2019; Bambara, 2019; Sieza et al., 2019). The increase in croplands in and around pastoral areas is one of the causes of conflict between farmers and herders during the rainy seasons because cropland damage by livestock is recurrent. The areas of pasture are fragmented and scattered between cropland, which are becoming difficult to access, and herd surveillance must be very active. Thus, Turner et al. (2016) pointed out that changes in grazing land in cropland often have implications for the community land tenure regime that limits grazing rights. The reverse in turn leaves less productive poor land with bare soils (Schlecht et al., 2006; Hiernaux and Assouma, 2020).

Although cropland can make a significant contribution to livestock feed during the dry season through the use of agricultural by-products (tops, straw, bran) by livestock (Hiernaux et al., 2009; Djohy et al., 2023). The occupation of pastures by fields in the rainy season can lead to overgrazing.

In addition to the high anthropogenic pressure on the vegetation cover, the effects of climatic hazards such as droughts and reduced rainfall would result in the mortality of certain woody and herbaceous species (Kosmowski et al., 2015; Kabore et al., 2019).

The extension of bare land in the area could be explained by poor practices, overgrazing, logging, mining, and the practice of bushfires in a context of population pressure. Many studies point to poor farming practices. Indeed, shortening or absence of set-aside times, slash-and-burn agriculture, overgrazing, and selective logging would account for 70% of land degradation (MRA, 2006; Gomgnimbou et al., 2010; Bambara, 2019; Kadéba et al., 2019). In addition, the combination of reduced and fragmented pastureland results in low pasture production, including animal overload at pastoral interstices during the rainy season. This overgrazing is responsible for the degradation of the pastures through the appearance of low-production and unbred species as well as bare soil. Furthermore, any changes in land use inevitably led to a decline in biodiversity in recent decades (Foley et al., 2005), which may explain the reduction or disappearance of some high-value forage trees. For example, agriculture is responsible for about 80% of deforestation worldwide (Kissinger et al., 2012).

The increase in the units of the water bodies could be explained by the development of the Bagré reservoir and hydraulic infrastructure (02 boulis) in the pastoral zone.

Pastoralists said that pastoral resources have been degraded over the years. This degradation is reflected not only in the low availability of fodder resources but also in the poor quality of pastures. As a result, certain woody and herbaceous species of high importance for livestock and human activities, such as Pterocarpus erinaceus (PROTA), Afzelia africana (SM), Andropogon gayanus (Kunth), Andropogon ascinodis (C.B.Clarke) Mitragyna inermis (WILLD), Bombax costatum (PROTA), Strophantus sarmentosus (DC), Vitex doniana (PROTA) are on the verge of disappearing, leaving room for other less species importance such as Senna tora (L), Hyptis suaveolens (L), Microchloa indica. The same observation was made by Sarr et al. (2013), who found that species such as Pterocarpus erinaceus and Bombax costatum are threatened with extinction because they are overexploited. Djohy et al. (2022) have added that floristic diversity is threatened by the loss of several plant species. This is said to be due to the high anthropogenisation of the pastoral zone, pruning and climate factors. The pastoral zone, which is a government estate, has not been exploited in accordance with its development and management plan and its specifications. As a result, other non-pastoral activities have proliferated in the area, such as gold panning, charcoal production, timber exploitation and the expansion of fields. In addition, the increase in the number of animals and the lack of control over the settlement of pastoralists are contributing to the degradation of pastoral resources through over-exploitation. When the zone was created in 2000, the number of livestock expected to use the area was far exceeded, and now there are more than 10,000 livestock units. For pastoralists, several factors are behind this degradation. These include drought, logging and artisanal gold mining, overgrazing and poor farming practices. The decline in forage and woody species is a result of human activities and the decline in rainfall. These results are consistent with those of Abdou et al (2020) which reported the degradation of fodder availability and the proliferation of unpalatable species such as Sida cordifolia and Cassia mimosoides that hinder proper development of some grasses. They also blamed insufficient rain and expanding croplands. Also, the work carried out by Coulibaly et al. (2021) in the commune of Diéma in Mali, highlighted the disappearance of Andropogon gayanus, Dactyloctenium aegyptium and Brachiaria sp. on the routes as a result of climate change.

Beyond the anthropogenic factors that lead to the degradation of pastures, the effects of climate change also contribute to the loss of fodder species. Fodder supply of rangelands has declined drastically in recent years due to the effect of climate change, pressure from human activities and reduced rainfall (Kanembou et al., 2009; Amegnaglo et al., 2019; Abdou et al., 2020; Djohy et al., 2022).

In fact, overgrazing, poor farming practices and selective logging lead to a degradation of the pastures, and this is reflected in a reduction in forage phytomass, in the pastoral value of the pastures. When the pasture is selective, it does not allow the renewal of grazed species when it is intense. This leads to the proliferation of unbred species by livestock and the reduction or disappearance of palatable species (Botoni, 2003; Anthelme et al., 2006; Bambara, 2010).