Study area
The Republic of Benin is located entirely in the intertropical zone, between the parallels 6°30' and 12°30' of latitude North and the meridians 1° and 3°40' of longitude East (Neuenschwander & Toko, 2011). It is bordered to the North by the Republics of Niger and Burkina Faso, to the South by the Atlantic Ocean, to the West by the Republic of Togo and to the East by the Republic of Nigeria. In addition to the North-West zone in the Atacora ranges, the center in the hills department, Benin generally presents a slightly uneven relief. By its extension between the coast of the Gulf of Benin and the Niger valley (6°30 to 12°4 N), Benin presents a varied range of climates. In effect, Benin climates are characterized by a relatively low annual rainfall which shows a variation of 900 to 1300 mm per year. Figure 1 shows the location map of Benin with the major urban centers as well as the main road network of the country.
Benin has a varied geography, ranging from coastal plains in the south to central plateaus and higher hills in the north (Vissin et al., 2007). The “Oueme” River crosses the country from north to south and is the largest river in Benin. Benin’s biodiversity is remarkable and presents a variety of its ecosystems, which include coastal areas, savannahs, rainforests, mangroves, lagoons and wetlands (Tchibozo et al., 2008). This diversity of landscapes promotes a great richness in terms of flora and fauna. Indeed, Benin presents a variety of animal species including mammals, birds, reptiles, amphibians and fish (Amoussou et al., 2012). There are elephants, lions, leopards, cheetahs, hippos, crocodiles, antelopes, monkeys, pangolins and many other species of mammals. The birds are also numerous and varied, with migratory and resident species. In the coastal areas, you can observe different species of seabirds and of pelicans (Adomou et al., 2009). Benin's flora is diverse, with a wide variety of trees, of plants and flowers (Djossa et al., 2008). Tropical forests are home to species such as cheese, teak, iroko, mahogany and baobab, which is an emblematic symbol of Africa (FAO, 2001). Benin also has several protected areas and national parks (Bakarr et al., 2004).
Botanical description and origin of Chromolaena odorata and Mesosphaerum suaveolens
This description relates to Akoègninou et al., (2006) and Aboh et al., (2008). Chromolaena odorata belongs to the Asteraceae family. Chromolaena odorata is a colonizing species of open environments, flowering from November to May (Akobundu & Agykwa, 1989). It is propagated either by cuttings, by regeneration from stumps or by seed dispersal. Its dispersion is mainly anemochorous. The species is a semi-lignified perennial herbaceous, very fragrant, diffuse and fast-growing perennial that can reach 3 m to 3.5 m tall and spread by seed or basal shoots. The species is native to the West Indies and tropical America (Gautier, 1992). The first appearance would have taken place in Nigeria between 1936 and 1942 (Delabarre, 1977). On the other hand, in Ivory Coast and the Central African Republic, coffee planters and pepper planters would have voluntarily introduced the species to use it as a cover (Huguenin, 1993). The plant appeared in Benin in the early 1970s from Nigeria (Avlessi et al., 2012).
Mesosphaerum suaveolens is called “Gros baume’’ in French, “azongbi’’ in Fongbé (Akoègninou et al., 2004). Mesosphaerum suaveolens belongs to the Lamiaceae family. It should be noted that it is an upright and very aromatic annual species. It can reach 1.5 m or even 2 m in height and spreads by seed. The stem is woody, polygonal, much branched, leafy, greyish, pubescent and marked with glandular dots. Originally from tropical America, this species is now widespread in tropical Africa and Asia (Hutchinson & Daziel, 1963). The species prefers fallow land, coastal sand and wooded savannah and colonizes quickly.
Analysis of local people's knowledge of the effects of land use on the proliferation of Chromolaena odorata and Mesosphaerum suaveolens
The causes of invasion by invasive alien plants are largely linked to human activities human activities that disrupt native ecosystems (IUCN 2004). Their seeds are transported voluntarily or accidentally by human activities or animal movements (Radji et al., 2010). Human activities include: road construction, land development, grazing, sowing fodder plants, erosion control work and the use of vegetation fires (Pysek P et al., 2010). Global climatic and geological changes are also important factors contributing to the establishment of invasive alien species (Hobbs & Huenneke, 1992). It is important to analyze and understand local populations' knowledge on the spread of invasive species in relation to land use. To achieve this objective, a survey was carried out among the local population. The target groups were mainly the farming populations of the communes surveyed. The sample size was determined using the Dagnelie formula (Dagnelie, 1998).
$$\varvec{n}=\frac{{\varvec{U}}^{2}\varvec{*}\left(\left(1-\varvec{p}\right)\varvec{*}\varvec{p}\right)}{{\varvec{d}}^{2}}$$
with n: the number of respondents; U: the value of the confidence level of a normal law (1.96); p: the proportion of people who know the species; d: margin of error set at 7% for this study (Singh et al., 2009).
Mapping and description of the structure of the different land cover classes: Data analysis and processing
The land cover map was produced from a 2022 Landsat image of 30-meter resolution downloaded from the Usgs website (www.usgs.org, on November 2022). Considering the complexity of the heterogeneity of our environment, the object-oriented approach was preferred, using Envi 4.7 software, unlike the classic approach based on a classification-oriented pixel from the spectrum (Adjonou et al., 2019). The object-oriented approach does not deal with the pixel in isolation. Indeed, the oriented approach treats the pixel in its spatial and spectral environment, by grouping pixels within objects interpreted based on spectral values, their size, shape, context (Benz et al., 2004).
Different steps were necessary for the processing of satellite images. Indeed, we have: (1) the Image Pre-processing which is a step that consisted of a radiometric correction of the Landsat image in order to correct any atmospheric biases and to go from the pixel value to the digital count to the reflectance value; (2) the colored composition: following several combinations, a colored composition presenting the best discriminations of the types of ground occupation (Sarr, 2009) was chosen; (3) classification and evaluation consisted of transforming the image into a thematic map. The classification was based on the spectral properties to group the different objects of the image into thematic class. To do this, the supervised classification was used with the maximum likelihood algorithm from the colored composition chosen for this purpose; (4) ranking validation: two ranking validation indices were calculated. This includes in particular the global precision (character the proportion of well-classified pixels, evaluated as a percentage) and the kappa index (character the ratio between the well-classified pixels and the total of the pixels probed) (Sparfel et al., 2008). In addition to these indices, field data were also used for validation.
Study of an association or non-association relationship between land use units and the abundance of Chromolaena odorata and Mesosphaerum suaveolens
In order to study the relationships of association or dependence between the land-use units and Chromolaena odorata and Mesosphaerum suaveolens, a remote sensing analysis of vegetation was carried out. Land use was mapped. The results were then cross-referenced with data on the presence of Chromolaena odorata and Mesosphaerum suaveolens, to highlight the habitats most frequently colonized. The land-use units considered refer to the classification of the different types of land and land uses present in the country.
Then, statistical tests were carried out to analyse the results. The Pearson's Chi-squared test was performed. A mosaic plot was used to represent the association relationships between species abundance and ecosystem type. Thus, depending on the value of the standardized residuals, the degree of association is assessed. Residual values between − 2 and 2 show an independent relationship and are represented graphically by a transparent color. Association relationships are observed when values are less than − 2 or greater than 2. Values below − 2 indicate an association with a low representativeness of the sub-group considered (red color), while values greater than 2 indicate a high representativeness of the sub-group (blue color).
The significance threshold is 5%, and the various tests were carried out using R 4. 2. 2 software (R Core Team, 2022).