The impact of Alternanthera brasiliana on vegetation and soil seed bank was assessed in Ile-Ife, Nigeria. Ten sample plots, 10 m x 10 m each, were established in invaded plant communities with high density of Alternanthera brasiliana and adjacent uninvaded plant communities where the weed species has low density. In each sample plot, twenty 1 m x 1 m quadrats were randomly laid and all rooted plant species were identified and counted. Post-dispersal soil seed bank was collected by randomly taking five core samples of top soil per sample plot to estimate the soil seed bank density and floristics of the sites. The species composition of soil seed bank was compared with that of the above-ground vegetation so as to assess the invader’s impact on the vegetation using Sorensen’s index of similarity. The results showed that Alternanthera brasiliana invasion significantly impacted on the species diversity (t = 5.27; df = 18; p = 0.0003) and evenness of species distribution (t = 4.50; df = 18; p = 0.00005) in the aboveground vegetation, and the species diversity (t = 5.37; df = 18; p = 0.00004) and evenness of species distribution (t = 6.19; df = 18; p < 0.0001) in the soil seed bank. This study concluded that Alternanthera brasiliana has significantly caused alterations in key parameters of the aboveground vegetation and those of the soil seed bank. It is likely that with increasing resident time, these alterations might increase more significantly to enhance the spread of Alternanthera brasiliana.

Study area

The study was carried out in some abandoned sites interspersed within the Obafemi Awolowo University estate in the City of Ile-Ife, in the southwestern Nigeria (Fig. 1). The university estate covers approximately 56.09 square kilometers (Makinde et al. 2021). Ile-Ife is found in the lowland rainforest type (Keay 1959) (latitudes 7° 30′ to 7° 35′ N and longitude 4° 30′ to 4° 35′ E, 213–457 msl) (Hall 1969). Ile-Ife has a short dry season that persists for approximately four months (November to March) with the rainy season occupying the remaining months with an average annual temperature of 25.5 °C, average annual rainfall of 1302 mm, average relative humidity of 82%, annual solar radiation of 164.30 Wm^-2, and average annual wind speed of 2.06 kmhr^-1 (Atmospheric Physics Research Group 2013). The land use system of the university estate is apportioned as follow: 18% of the area constitutes built infrastructure, 22% serves as agricultural and research farmland, with 60% remaining undeveloped (Makinde et al. 2021). Alternanthera. brasiliana is widely distributed through the university estate. For this study, invaded sites with close proximity to uninvaded sites (A. brasiliana absent or at very low density) were selected for sampling. The invaded site were located between latitudes 7°31.391′ to 7°31.531′ N and longitudes 4°31.676′ to 4°31.872′ E while the uninvaded sites ranged between latitudes 7°31.198′ to 7°31.207′ N and longitudes 4°31.737′ to 4°31.751′ E. The study area is underlain by rocks of the Basement Complex of Precambrian age (De Swardt 1953), with soils belonging to the Ultisols class (USDA 1975, Uwalaka and Muoghalu 2021).

Impact on aboveground vegetation parameters

To determine ecological impact of the plant invader on the invaded plant communities, ten 10 m × 10 m sample plots (termed large plots) were established in invaded plant communities with high cover or density of A. brasiliana and adjacent uninvaded plant communities where A. brasiliana is not present or had very low cover or density in the Obafemi Awolowo University estate. Within the ten large plots, twenty 1 m × 1 m quadrats were randomly placed and sampled to estimate the density and frequency of species in each large plot. In each quadrat, all rooted plant species were identified and counted. To determine the percentage cover of species, five 10 m line transects were randomly laid using a measuring tape in each large plot. A single cover pin was then perpendicularly dropped at every metre point along each transect according to the method of Evans and Love (1957), modified by Uwalaka and Muoghalu (2021). The “Lucky draw randomisation method” was used by generating random numbers to produce points where quadrats were laid, and transects were sampled. All plant species parts that contact the pin was recorded to estimate the cover of the species in each large plot (Uwalaka and Muoghalu 2021). The percent cover was then calculated as the number of ‘hits’ per species divided by the total number of pins dropped multiplied by 100 (Uwalaka and Muoghalu 2021). Voucher specimens of the plant species which could not be identified in the field were collected and identified in the IFE Herbarium of the Department of Botany, Obafemi Awolowo University, Ile-Ife, Nigeria. Species nomenclature follows that of POWO (http://www.plantsoftheworldonline.org/).

Impact on soil seed bank parameters

From the same plots where aboveground vegetation parameters were measured, soil seed bank samples were collected after seed rain by randomly taking five core samples of the top soil (0–20 cm depth) per large plot between December 2017 and February 2018 after the resident plant species (invader and other plant species) had completed their growth cycles and their seeds had been dispersed. The soil samples were air-dried, and the debris, stone and gravel were removed by passing the soil samples through a 4 mm mesh sieve. The soil was thinly spread in perforated 45 cm-diameter plastic trays and placed in a screenhouse at the Department of Botany, Obafemi Awolowo University, Ile-Ife, at an average temperature of 39 °C, light intensity of 3517 lux and humidity level of 54% (Uwalaka and Muoghalu 2021). To ensure that seeds were not accidentally introduced into the trays, three trays containing post-dispersed soil seed bank samples that had been sterilised in an oven at 105 °C for 48 hours to ensure deactivation of the seeds present in the soil, were used as control (Uwalaka and Muoghalu 2021). Germinated seedlings were counted, identified and removed every week and the soil stirred to bring seeds buried in the soil to the surface. Unidentified germinated seedlings were transplanted into planting bags, watered daily until they flowered and were then identified. The seed bank study lasted 12 months.

Data and statistical analysis

Species density in the aboveground vegetation was estimated as mean density ± 95% confidence interval per hectare for each species. Species richness was estimated as the number of species in each plot per site. The sum of relative density, relative frequency and relative cover was used to estimate the species importance value of each species in the invaded and uninvaded communities (Cottam and Curtis 1956) which was in turn used to assess the dominance of each species in the communities. To determine the relationship between the aboveground vegetation and soil seed bank, the species composition of the site types was compared using Sorensen's index of similarity with the formula: ISS =  x 100. Where; C is the number of species common to both seed bank and standing vegetation, A is the number of species in seed bank, B is the number of species in standing vegetation. Indicator Species Analysis (ISA) was performed on the dataset using the ‘Indicspecies’ package in R, after grouping the plots into two, viz. invaded and uninvaded. The plots served as indicators (row) while the species were the vector (column). The importance value index of the species, which was the measure of abundance, was used as factor in the dataset. Principal Component Analysis (PCA) was further used to identify the impact of A. brasiliana on the invaded site using the importance value index of each species as standard of abundance. The Chi-square test of independence was carried out to test for the association between plant species forms in aboveground vegetation and their corresponding soil seed banks in both invaded and uninvaded sites. An independent samples t-test was used to compare the means of the total species density, diversity and evenness between the invaded and uninvaded plots of the aboveground vegetation as well as those of the invaded and uninvaded soil seed banks. Plants were divided into guilds (native, invader of interest and other alien species), and the contributions of species to the aboveground vegetation and soil seed bank were expressed as a percentage. The percentage contributions were then used to test for the influence of the presence of other alien species in the aboveground vegetation and seed bank of both invaded and uninvaded sites using a contingency test. The analysis was done using the PAleontological STatistics (PAST) version 3.17 software (Hammer et al. 2001).

The files are in text and excel formats.