The Nigerian riparian forest ecosystems had declined in extent and distribution and this had been attributed mainly to land use change. This study intended to provide an understanding of the links between plant diversity, composition, structures, and disturbances both anthropogenic and natural processes inducing the vegetation dynamics. Nine study sites were used for this study, within each site, five (5) plots (0.25 ha in size) were marked out and placed systematically at an interval of 10 m along the transect. A complete enumeration of plant species was carried out and identified at the species level. Diversity indices and structural parameters were determined and anthropogenic activities were ranked. A total number of 233 plant species were identified, belonging to 80 families; out of which, Euphorbiaceae and Apocynaceae were dominant families The density and basal area ranged from 2,200-6,000 ha^-1 and 2.59-17.58 m² ha^-1 respectively across the study sites. Pterocarpus santalinoides, Alchornea cordiflora, Chassalia kolly, Tetracera spp, Fimbristylis, Bambusa vulgaris and Cyrtosperma senegalense were the dominant species. The Shannon diversity index ranged from (1.38-3.49), Simpson (0.66-0.97), and Evenness diversity (0.43-0.84). Fisher alpha (10.03-30.21) and Whittaker beta diversity (0.36-0.89) values were highest in Ipetumodu (site VIII) and lowest in Ilesha (site II). Seventy-three (73%) of the species in this study had a low important value index (IVI). The dominance of some lianas and herbaceous species in the riparian forest sites showed disturbances, stages of ecological succession, and regeneration of the vegetation. Conservation is inevitable towards maintaining and protecting species diversity, ecosystem roles, and services of these forests in Nigeria.

Description of the study area: The study was carried out in Osun State located in Southwestern Nigeria. The state lies within latitude 7° 30′ N and longitudes 4° 30′ E. Ilesha West, Atakumosa West, Ife North, Ife North East, Ife South, and Ayedade were the Local Government Areas, where the sites are situated within Osun State, Southwestern Nigeria. Nine riparian forest sites (Ifetedo, Ilesha, Osu, Famia, Ibodi, Tonkere Gbongan, Edunabon, and Ipetumodu) were selected from these Local Government areas and designated as site I-IX respectively. These areas represented riparian zones with flooding patterns of Southwestern Nigeria and have high plant species diversity, despite varying ongoing anthropogenic activities (Figure 1, Table 1). The detailed climate, soil, and vegetation of the study area in Southwestern, Nigeria have been described (Borisade 2020, Borisade and Odiwe2021).

Data Collection

Vegetation Sampling and Data Collection for Plant Species

Woody species assessment: In each community, forest stands within the selected sites were sampled on a transect, 150 m long extended perpendicular to and across the streams/river. Nine sites were used for this study, within each site, five plots (0.25 ha in size) were marked out and each plot was placed systematically at an interval of 10 m along the transect, thereby giving a total of 45 plots across the study sites. Each plot was demarcated by narrow-cut thin lines to mark the boundary of the plot. Stem diameters were measured using a diameter tape at breast height (1.3 m). For individuals with buttresses or other stem irregularities at breast height, the diameter was measured above the buttresses, following usual forestry procedures. For shrub and lianas encountered, stems were counted, the diameter of three “average” stems from individual species was measured and then the composite (stem diameter (dbh)) of the shrub and lianas were calculated to enable the computation of its basal area in the manner used for the trees (Burton et al. 2005). For density calculations, each individual was considered as one, independent of its number of branches.

Herbaceous species assessment: To assess the species composition, density, and cover of herbaceous species mainly forbs, and other understory forms such as grasses, sedges, and ferns. In each established sampling plot (0.25 ha), five 1m x 1m quadrats were randomly laid. In each quadrat, all rooted plant species were identified and counted. Furthermore, five 10 m line transects were randomly laid using a measuring tape. A single cover pin was perpendicularly dropped at every meter point along the transect. Any plant species part or base touched was noted to estimate the aerial and basal cover of the species in the plot. The percentage cover was calculated as the number of ‘hits’ per species divided by the total number of pins dropped multiplied by 100.

Plants were identified to species level and unknown species were cross-checked using IFE herbarium, Flora of West Tropical Africa by Hutchinson and Dalziel (1954), Handbook of West African Weeds (Akobundu and Agyakwa1998), Trees of Nigeria (Keay1989), Plant lists and PROTA. Voucher specimens of the different species were collected, dried, and deposited in the IFE herbarium, Department of Botany, Obafemi Awolowo University, Ile-Ife, Nigeria.

Disturbance ranking: Indication of anthropogenic activities in the riparian forest was made within the study sites and disturbance scores were obtained for each site using the method described by Mani and Parthasarathy (2006) which was modified. Disturbance scores were given to each site by qualitatively evaluating various disturbances (tree logging, mining, grazing, flooding occurrence, agriculture, and settlement) ranked into low (1), occasional (2), and frequent (3) levels of disturbances. The summation of all the scores within each site was considered and sites with high values were ranked as high disturbance sites while those with low values were as low disturbance sites.

Data and Statistical Analysis: The data from the complete enumeration of woody species were used to determine stem densities per hectare for different life forms (trees, shrubs, and climbers) and this was calculated using the number of individuals divided by sample areas. The data were used to establish floristic composition in terms of species, genera, and families.

The important quantitative analysis such as density, frequency, and abundance of tree species, shrubs, climbers, and forbs species was determined as described by Curtis and McIntosh (1950). The basal area for woody species was also determined.

Density: The density of each species, D_i, was calculated as the number of individuals in a unit area: D_i = n_i / A where D_i is the density for species i; n is the total number of individuals counted for species i, and A is the total area sampled.

Frequency: The frequency for each species was calculated as: f_i = j_i / K where f_i is the frequency of species i, j_i is the number of samples taken and k is the number of occurrences.

Basal area: Basal area which indicated the importance (dominance) of tree and shrub cover (Carratti et al.2004), was determined as: Basal area (m²ha^-1) = C² / 4π where: C = girth size (diameter at breast height), π = 22/7 = 3.14.

The basal area of each species was determined by adding the basal area of individuals of the species and the plot basal area was calculated by adding basal areas of all species in each plot while the site basal area was calculated as the mean woody species basal of all the sample plots.

Importance Value Index: This index was used to determine the overall importance of each species in the community structure.

Relative Dominance = Basal area per family/ Total Basal area×100

Relative Frequency = Frequency of each species/ Sum of frequency values of all species×100

 Relative Density = Density of each species/ Density of all species ×100

The three relative values were added together to obtain Importance Values (IV) for each species.

Diversity Indices: The total number of species recorded in the sampling plots (species richness), rarefaction measure as well as Shannon-Wiener diversity Index (H’), dominance and beta diversity, and species evenness were employed to quantify and characterise species diversity and species-abundance distributions of the plant communities.

Alpha species diversity was calculated for each site as Shannon-Weiner diversity, described by (Magurran2004), Dominance was evaluated using Simpson’s index of dominance (1949); Beta diversity was determined using Whittaker’s (Whittaker 1972); Rarefaction diversity (E(Sn)) was computed from the floristic data in order to compare species numbers from samples of different sizes among the community types (Hsieh and Li 1998). The Species evenness was calculated using Pielou Index (J) (1977).

The confidence interval of structural parameters such as density and basal area was set at 95%. The disturbance scores, species density, and diversity indices (Shannon-weiner diversity, species evenness, and dominance) were subjected to Correspondence Analysis (CA) and the overall influence of these parameters on the riparian vegetation sites to Detrended Correspondence Analysis (DCA) using PAleontological STatistics (PAST) version 3.17 software (Hammer et al. 2001). 

PAleontological STatistics (PAST) version 3.17 software (Hammer et al. 2001).