Impact of seasonal changes on the vegetation composition of some selected areas in Lagos state , Nigeria

Rapid degradation of wetlands and habitat fragmentation are major threats to species diversity and ecological paradigm in Lagos state, Nigeria. Best to our knowledge, no empirical study has explored the effects of seasonal changes on covering three major cities in the statewithin the major cities in Lagos state. For this, we carried out a study covering the three major cities in Lagos: Badagry, Epe and Ikorodu. Permanent plots were randomly established for monthly data collection (quadrat method; size=10 x 10 m), that span across one-year period covering the wet and dry season). We observed two major vegetation types (secondary forest and mangrove) in the studied areas. Additionally, we documented 107 species belonging to 90 genera and 40 families. Species diversity analysis indicated higher diversity, abundance, species richness and evenness in Ikorodu and Epe during the wet has a greater while high species diversity was observed in Badagry during the dry season. The Dahomey gap phenomenon have greater implication on the Badagry vegetation while the coastal nature of Epe and Badagry negatively influenced their vegetation structure. Analyses inferred that species diversity pattern in the studied area seem to be determined by seasonal changes alongside anthropogenic activities and as a environmental factors (rainfall, geomorphology and geology). Further, our result suggests that Ikorodu could be regarded potential biodiversity hotspot area for conservation purposes because of its high floral composition and diversity. Therefore, improved management and conservation of the vegetation of the Ikorodu area is hereby recommended. The findings from this study will be useful in guiding the Physical Planning and Urban Development and Environment ministries in Lagos state for appropriate management and monitoring of developmental projects for sustainable development.


Introduction
The Convention on Wetlands of International Importance (also known as RAMSAR Convention; Ramsar, 1994), defines wetlands as areas temporarily or permanently, naturally or artificially dominated by marsh, peatland, fen or water that may either be brackish or salty, fresh, free flowing or static (Cowan, 1995).Wetlands play vital roles in controlling soil erosion, purifying water, attenuating flood, regulating stream flow and serving as water reservoir, recreational and tourist centre among many other uses (Williams, 1990).Thus, the importance of wetland conservation and management cannot be over emphasized.
Biodiversity conservation scheme at any scale (local or region) is imperative to protect the ecological component in the environment and sustaining the benefits derived from them (Kotze, 1996).Over the years, vegetation/ecosystem conservation has become a global concern with special attention on wetlands (including Africa) because it serves as hotspot for biodiversity and have been experiencing degradation through anthropogenic activities.
Nigeria, the largest in West Africa, has an array of different vegetation types with an extensive wetland distribution in the South-Western zones particularly Lagos State.Different from the natural wetlands, are artificial wetland consequent to creation of ponds and dams with limited species diversity studies (Zakari, 1992;Eni et al., 2011;Olubode et al., 2011).Lagos is characterized by coastal environment (sandy barrier beaches, low-lying tidal flats, and estuaries) with three vegetation types (Mangrove, Secondary Rain Forest and Freshwater Swamp).Hitherto, the vegetation is still under constant threat with major causes of biodiversity loss being indiscriminate land reclamation for development (urbanization) and anthropogenic functions.(Dixon &Sherman, 1991, Filmer andPritchett, 2002;Oramah, 2006;Pascual & Barbier, 2006).Abegunde (1988) reported that subcomponents of urbanization such as sand filling of lagoon shores, excessive dredging, extensive reclamation of wetlands, encroachment on natural drainage channel and unrestrained deforestation, have become frequent activities that characterize Lagos State.Unfortunately, this has been the case in the last three decades and its impact on Lagos vegetation cannot be overemphasized.Just as in many coastal cities around the globe, a full grasp of the effect of anthropogenic factors on coastal vegetation is yet to be properly assessed and quantified.
Ecologists and environmental scientists have expressed more concerns on the post effect(s) of rapidly disappearing vegetation rather than vegetation analysis which may provide a clue to the causes of plant diversity loss.However, past limited studies available include Adekanmbi & Ogundipe (2009); Obiefuna et al. (2013), Oladele (2013), Adeniyi et al., (2016), Akinsoji et al. (2016) and Adeleye et al., (2017).Intriguingly, they all agreed that the vegetation was highly diverse but is under serious threat (Obiefuna et al., 2013;Oladele, 2013;Adeniyi et al., 2016).Nonetheless, we assumed that lack of empirical studies to investigate seasonal fluctuations on species diversity pattern in Lagos state could be due to (i) conflicting representation of seasonal changes in aquatic and dry land ecosystems and (ii) development of effective evaluation methods.As much as we know, previous studies on Lagos vegetation have focused on either the floral diversity of selected study sites or changes in the vegetation size, with no record on the effect of seasonal changes.In support of this study, Rosenzweig (1995) had suggested the need to investigate seasonal patterns of diversity in other to understand the dynamics of plant communities.Asynchrony in species diversity fluctuations is also key for ecosystem stability.Furthermore, a detailed assessment of wetland biotopes including their location, ecological characteristics, and distribution are essential factors suggested for proper conservation and implementation strategies (Cilliers et al., 1998).
Thus, this study aims at evaluating the changes in floristic compositions of different ecosystems (mangrove, fresh water swamp forest, coastline vegetation, and secondary forest) across Lagos state vegetation zones (Badagry, Epe and Ikorodu) over a seasonality gradient.Accordingly, we provided information on the conservation status and habits of all encountered species.Finally, we offered recommendations for improved conservation of plant community under sustainable development in the region.

Study Area
Lagos is geographically located in South Western Nigeria between longitudes 2º 42' E to 3º 42' E and latitudes 6º 22' N to 6º 42' N, with two distinct seasons (Ogundele, 2012); wet (April to July, October and November) and dry (August and September; December to March).Wind directions depends on the Inter Tropical Convergence Zone (ITCZ) seasonal positions.Additional environmental parameters include: Mean annual rainfall (1381.7 mm-2733.4 mm), temperature (maximum [29ºC and 34ºC] and minimum [24ºC and 28ºC]) and high relative humidity (> 70%) throughout the year (Ogundele, 2012).Sequel to the reconnaissance survey conducted prior to this study, we divided Lagos state into three zones (Badagry, Epe and Ikorodu) based on vegetation richness and ecosystem diversity (Figure 1).These areas are representatives of Lagos state vegetation zones which are experiencing rapid degradation due to development and anthropogenic activities e.g.farming, constructions, infrastructure installations, habitat fragmentation and land fillings.

Vegetation analysis
Georeferenced conserved permanent plots in each location were randomly located based on accessibility and less anthropogenic interference (Table 1, Figure 1).These sampled points were analysed using quadrat method (size=10 x 10 m; Song & Zhang, 2013;Vinchesi &Walsh, 2014).Our data gathering lasted over a a period of one year (September, 2016 to August, 2017) which spanned dry season (DS) and wet season (WS).These sampling points were visited monthly to observe the changes in the vegetation composition and distribution.Plant species were identified to species level using the following literatures: Angiosperm Phylogeny Group ( 2009), Hutchinson & Dalziel, (1954;1958, 1963, 1968and 1972) and Akobundu & Agyakwa (1998).Specimens with uncertain taxonomy were collected and preserved following the method of Blanco et al. (2006).Subsequently, they were taken to the Herbarium, Department of Botany, University of Lagos for identification.Further, we assessed the conservation status of the encountered species using International Union for Conservation of Nature red list web-interface (IUCN ver. Olowokudejo & Oyebanji 2016;Oyebanji et al., 2017).

Data analyses
Primary data (observational records) collected analysis during our survey were used for the diversity analyses.Frequency distribution and relative frequency of occurrence location were calculated for each species based on locations and season (Appendix 1-3).Our diversity index analysis (Dominance, Evenness, Equitability, Shannon-wiener and Simpson's indices) were conducted using Palaeontological Statistics (PAST) software (v 2.17c, Hammer et al., 2001).Also, our study presented percentage plant family (APG, 2009) and habit (Life-form) distribution (Adeniyi et al., 2016;Walker et al., 2016).We evaluated similarities and differences among the selected studied sites using Jaccard correlation (statistical function in PAST software) based on species occurrence during the study.
Figure 1.Map showing the selected study areas and respective sampling points within Lagos, Nigeria.
We observed seasonal differences in species composition and abundance across all study areas, moreover some plants (e.g.Alchornea cordifolia, Aspilia africana, Chromolaena odorata, Cyrtospermum senegalense, Dryopteris, Ipomoea involucrata and Raphia hookeri) were observed to be present in both seasons ( This finding suggests that seasonality plays little or no role in the distribution of 'cosmopolitan' plant species (i.e.species present in both DS and WS).Since geographical space did not limit their occurrence, we infer that shared environmental and climatic characteristics could be predominant factors that necessitated the co-occurrence of these species.Thus, these species may be described as Lagos vegetation cosmopolitan species.Moreover, these plants may not be abundantly recovered, they will always be present in any of the vegetation types in any location in Lagos state.Edaphic and topographic as well as climatic factors may have been responsible for the uneven distribution of these species.Although our findings showed that these species could be space controlled and not evenly distributed, yet, further studies are required to justify this fact across the entire geographic range of these species.

Badagry location
We encountered two vegetation types (secondary swamp and dry secondary forest).A total of 55 species belonging to 48 genera and 27 families were documented (Table 1).Our study documented the dominant family (Cyperaceae,7 species) and species (T.domingensis) while most species were forbs (36%; Figure 3a, 3b).Furthermore, frequencies and relative frequency of species occurrence varied with the locations (Appendix 1).Our study showed that A. cordifolia occurred in all the sampling points within Badagry while A. gangetica, C. diffusa and F. trichopoda occurred in at least three of the sampling points in Badagry (Appendix 1).
Our finding correlated seasonal changes with species diversity in Badagry.Although differences in the percentage frequency of most species present in both seasons was little or rather negligible, however, several others have clear differences (e.g. A. cordifolia (WS [3.25%] and DS [12.65%]) and L. abbysinica WS [11.80%] and DS [3.02%]; Appendix 1).Further, these factors are burning might have affected species distribution (abundance and diversity) in Badagry include burning, clearing, exploitation for medicinal purposes, extraction of non-forest timber products, farming, housing projects, fuel wood, transportation and sand filling.The observations are similar to results from past studies (Filmer & Pritchett, 2002;Oramah, 2006;Pascual & Barbier, 2006;Adeniyi et al., 2016;Olowokudejo & Oyebanji, 2016) on the influence of human factors on species diversity.

A B Epe location
Similarly, we identified two vegetation types (freshwater swamp secondary and dry secondary forest), and a total of 52 species, belonging to 46 genera and 29 families (Table 1).The dominant family was Poaceae (six spp.), Desmodium, the most common species (three species) while most of the species were forbs (n=21 spp. [40%]; Table 3, Figure 4A, 4B).Only one species (A.cordifolia) occurred in all the sampling points while three species (C.afer, D. tortuosum, and E. coccinea) were restricted to one sampling point (Appendix 2).
Freshwater aquatic species recorded in this area were N. lotus, S. natans, C. dermasum, E. crassipes and P.stratiotes, although they were found outside the studied quadrats.Previous studies of Okorie (2012) have described mosaic vegetation types in Epe to include freshwater Swamp Forest.Adeniyi et al. (2016) also documented these species but they did not capture C. dermasum (submerged species) which is recorded in this study.
Clearing and sand filling of swamp dominated by Raphia spp., farming, road construction, extraction of non-forest timber products, indiscriminate burning, and fuel wood are some of the anthropogenic factors observed in Epe during our field survey.These activities have been reported to profoundly affect species distribution, abundance and diversity.Previous study (Okorie, 2012)

A Ikorodu location
The sampled vegetation types comprised secondary (dry and swamp forest) and mangrove forest (Table 1).Ikorodu recorded the highest diversity (66 species, 59 genera in 34 families) compared to Epe and Badagry regions.The dominant family was Cyperaceae, (seven spp.) while most species were forbs (n=20 species [30%], Figure 5A, 5B).Our study recorded four mangrove species (A.aureum, D. lunatus R. racemosa and P. reclinata) and long stretch of R. hookeri along the wetland/swamp that extends across Ikorodu sampling point 3 (Table 1).
Species diversity was higher in WS (n=61 species [62%]) than in the DS (n=37 species [38%]).Panicum maximum is the most dominant species while A. cordifolia and C. odorata are dominant in WS and DS respectively (Appendix 3).

Seasonal difference in the occurrence of taxa
Results from this study showed that the occurrence of most taxa varied with seasons in all the studied locations, as evidenced by: (a) difference in species diversity and evenness in dry and wet seasons (b) varying species number, individuals and equitability (c) differential values of species abundance and frequencies on both seasons.Observations made regarding the differences in species abundance, diversity, evenness and equitability with respect to season is consistent with the study of Cruz-García (2015).Comparing the seasonal occurrence of taxa in the three sampling locations, the highest species diversity (n=61) was recorded in Ikorodu during the DS while least species number (n=36) was recorded in Epe during the DS.In contrast, Badagry, had high species diversity and the abundance during the WS (Table 3).Generally, species diversity pattern observed fluctuates with seasonal changes (WS, high diversity and DS, low diversity).Previous study of Shimadzu et al. (2013) showed that rainfall is a strong environmental factor that determines the response of species and therefore dictates the seasonal changes in species' abundance.This effect was evident in our study although with exceptions e.g. A. cordifolia and S. africana which increased during the DS but reduced drastically during the WS in Badagry.Similar observation was made in Epe (A. cordifolia, P. pentandrus, and P. foetida) and Ikorodu (S. acuta, P. maximum, and B. coccineus).Despite these evidential changes in species abundance over a seasonal gradient, we observed that some species were available and abundant all-round the study period suggesting that their abundance were not affected by the temporal changes.For instance, in Badagry, C. odorata, T. domingensis, and R. hookeri were analysis present in both seasons (Figure 6).
Our species diversity index analyses (Table 3) revealed high diversity in two locations (Epe and Ikorodu) during the WS and one location (Badagry) in DS.The highest species abundance was recorded A B during the WS in Ikorodu compared to other localities (Table 3).Williams et al. (2008) demonstrated that rainfall is the most significant variable among within-year climatic variability that determines vegetation spatial pattern.This shows that the abundance, distribution and plant assemblages/diversities in tropics is predominantly determined by rainfall.This result somewhat supported the general expectation of higher species occurrence in the WS (Epe and Ikorodu).This therefore suggests that anthropogenic factors, more than the physical environment, played a vital role in determining seasonal taxa occurrences and species diversity in these locations (Cruz-García, 2015).The species evenness was observed to be high across all the studied localities, but highest in Epe (0.9365) than Badagry (0.9017) and Ikorodu (0.925).There was a direct relationship between taxa occurrence, species diversity and species evenness across the studied areas (Table 4).The area with relatively high number of species (Ikorodu) was observed to be more diverse and its representative species were more evenly distributed.This may be attributed to the size of the species pool (Silva et al., 2010).
A reduction in population size usually results from the effect of environmental constraints such as waterlogging and drought which leads to decrease in species abundance, infer diversity and availability (Tokeshi, 1999).Thus, we inferred that the proximity of Epe and Badagry to the Lagoon coupled with occasional incidence of waterlogging might have resulted in the decrease in the number, abundance and diversity of species encountered.Notwithstanding, our results may indicate that most individuals in these locations are flood resistant (Table 1).

Ecosystem Similarity
Our ecosystem correlation analysis (Jaccard correlation, dendogram), classified the sampling points into two main ecosystem groups (A [secondary forest] and B [mangrove vegetation]; Figure 7).Group A depicting a secondary forest species comprised of individuals from Ikorodu (1-3); Badagry (2-3), and Epe (1-3).Furthermore, group A was subdivided into three subgroups: subgroup1 included populations from dry secondary forest ecosystem (Badagry 3, Epe 2, Ikorodu 2 & 3), subgroup 2 consisted of individuals from fresh water swamp forest ecosystem (Badagry 1 & 2, Epe 1and Ikorodu 1) while subgroup 3 comprised of individuals from mosaic of dry and swamp ecosystem (Epe 3).Moreover, most similar locations were (Badagry 2 and Epe 1 =0.34) followed by Epe 2 and Ikorodu 2 having Jaccard similarity index value of and 0.26.Group B which included individuals from mangrove vegetation were collected from Badagry 4 and Ikorodu 4 (similarity index value=0.16).Although, Badagry 4 is not entirely mangrove ecosystem, the presence of mangrove species (A.aureum) necessitated its classification as mangrove ecosystem.This occurrence has been attributed to be vestiges of a once flourishing mangrove ecosystem in the Holocene (Sowunmi, 2004) which has now been replaced by freshwater elements.
Finally, we infer that the effect of seasonality on species abundance and diversity as seen in all locations could be due to conflicting processes such as primary and secondary succession (Zhang et al., 2012), climaxing (White, 1979) and invasion (Grice, 2006) within the plant community which ultimately contributed to species variability in terms of abundance and diversity.Other probable reasons include availability of resources, environmental conditions, species interaction (interspecific and intraspecific competition; Oladele, 2013), demographic stochasticity (Shimadzu et al., 2013) and anthropogenic activities (Shaheen et al., 2011).

Conservation status
Our study classified the species encountered into three conservation categories (according to IUCN, Table 1): NE (92 spp.), LC (14 spp.) and VU (one spp.).Documentation of H. stipulosa (vulnerable species, IUCN version 2.3) is novelty in this study.However, we recommend further ecological studies to ascertain current conservation status of these species especially H. stipulosa.The occurrence of conservation concern species such as the above-mentioned H. stipulosa and the increasing of anthropogenic factors justified the current rating (potentially threatened) proposed in this study.

Conclusion and recommendation
The impact of climate change on floral diversity is presumed to be largely due to changes in rainfall and temperature pattern, which consequently, resulted in variations in nutrient cycles, microbial and physiological activities in plants.We observed different species diversity within the locations even though they activities shared similar climatic conditions and anthropogenic although at different intensity.Our study revealed that recent development and urbanization project would have devastating effects on the plant community and may probably reduce wetland cover in the region.On this premise, we recommend that, (i) Ministry of Environments should map out the Ikorodu -Epe axis for conservation purposes, (ii) adopt strict monitoring policy and wetland conservation programmes (iii) conduct regular vegetation assessment and (iv) integrate biodiversity conservation into state development plans (sustainable development

Figure 2 .BFigure 3 .
Figure 2. Distribution of species across all studied areas during the sampling period.

BFigure 4 .
Figure 4. Distribution of species in Epe, Lagos state, Nigeria.A: Family distribution; B: Percentage distribution of Plant life-forms.

Figure 5 .
Figure 5. Distribution of species in Ikorodu, Lagos state, Nigeria.A: Family distribution; B: Percentage distribution of Plant life-forms.

Figure 6 .
Figure 6.Taxonomic classification of species encountered in Badagry, Epe and Ikorodu areas of Lagos.

Table 1 .
Locality information showing the geographic coordinates and habitats

Table 3 .
Diversity analyses of floral species encountered in each sampling locations during dry and wet seasons.
farming.Despite this, we recorded impressive floristic composition.Our result therefore indicate that Ikorodu is an area of conservation importance.

Table 4 .
Species diversity indices of Ikorodu, Epe and Badagry in Lagos State.