The study was carried out within the mangroves of Tiko, located between latitude 4˚28'30"N and 3˚54'26"N, and longitude 8˚57'10"E and 9˚30'49"E (Figure 1). Tiko has a land area of 48,400 km² [19] and a population of 134,649 inhabitants as of 2010. This mangrove zone, at the foot of Mount Cameroon, is located on the Cameroon Atlantic Coast of the West African Gulf of Guinea [20]. It has a maritime to equatorial climate with a short dry season (December to February) and a long rainy season (March to November). The area receives excessive rainfall with mean annual rainfall of 3000 to 4000 mm. Relative humidity ranges from 80% to 95% depending on the season while mean annual temperature ranges from 28˚C to 35˚C. Originally evergreen lowland forest and mangrove swamps, 80% of the forest land of Tiko municipality has been converted to oil palm, rubber and banana plantations by Cameroon Development Corporation (CDC) and only few patches of secondary forests exist. The creeks harbour large areas of mangrove forests which are very highly exploited for wood. The soils are ferralitic, volcanic and very fertile [21]. The principal activity employing over 75% of the cosmopolitan population is agriculture. In addition to the CDC plantations, there are also small scale farms producing cocoa and food crops. Other economic activities of the town include trading, fishing and wood extraction. Unfortunately the supply of energy to this dense population is erratic and expensive. Regular power cuts and irregular gas supply together with poverty have pushed the population to depend on biomass. The biomass comes mainly from old rubber trees and mangroves. The high density mangrove wood is preferred for construction poles, fish smoking, charcoal production and firewood for home use.

Sampling sites were identified (Figure 1, Table 1) after a reconnaissance survey in 2012 based on the duration of perturbation in the locality within the mangroves through interviews of locals.

Semi-structured questionnaires and open ended interviews were used to assess the different activities in the area as drivers of change that degrade the mangrove wetlands in Tiko. Purposive sampling techniques were used to select the respondents with the questionnaires having been divided into two main groups. The first group of respondents was made up of the general public living in and around the creeks. Here 100 randomly selected respondents answered questions related to their use of products derived from the mangrove ecosystem. The second set of questionnaires targeted professionals such as loggers, wood and pole collectors, wood splitters, charcoal producers, fishermen and fish smokers. Here, 60 questionnaires/interviews were administered. The first part of all questionnaires/interviews collected demographic data while the second part focused on their activities in the mangroves.

Two sets of composite sediment samples in triplicates were collected at low tide from each site. Samples for physico-chemical characterization were collected using a soil corer. At each site, a total of 10 randomly collected core samples were bulked and subsampled and put in labeled plastic bags for transportation. Samples were air dried to constant weight according to Reddy [22], crushed in a mortar and sieved through a 2 mm sieve. Each sample was subsampled, packaged, coded and sent to the Soil and Environmental Chemistry Laboratory of the University of Dschang for analyses. Particle size was determined using Bouyoucos hydrometer method, while the pH was measured using a pH meter with a glass electrode. Available phosphorus was analysed by the Bray II method. Total nitrogen and ammonium were determined by the Kjeldahl’s method. The levels of the exchangeable bases (Na⁺, K⁺, Mg²⁺ and Ca²⁺) were determined using flame spectrophotometry. Organic carbon was analyzed by the Black [23] method. The amount of organic matter was determined by the formula:

% OM = % OC × 1.724 [24].

Samples for benthic algal analysis were collected and processed according to Danielson [25]. Samples were collected using petri dishes and trowel. Composite core samples were transferred into beaker and distilled water added to the mark of 150 mL. Samples were then poured from the beaker into wide-mouth, nalgene bottles and 3 drops of 10% Lugol’s iodine added to each bottle to fix and preserve cell structure. They were then transported to the Life Sciences laboratory of the University of Buea in ice-filled cooler [26] for analysis. A drop of each centrifuged (2500 rpm for 20 minutes) sediment sample was mounted on a clean slide. Counting and identification of species was done using an Olympus BH – 2 light microscope equipped with Normaski optics at a magnification of 1000 X. Slides for quantitative and qualitative analyses were prepared in triplicates, and whole count method was employed. Following counting, extrapolation was done by multiplying the count from three drops (3 ml) by a factor of 333.33 to get the number of cells per litre. The Sedgwick Rafter counting chamber was used to determine algal density [10]. Identification of algae followed relevant text books and publications, including [27] - [38] and were classified according to algaebase.org. [39].

Data obtained from the questionnaires were analyzed using Microsoft Excel 2010 to produce simple charts and other descriptive statistics for each of the parameters in all the questionnaires. Kolmogorov-Smirnov test for normality at α = 0.05 showed that sediment nutrient data were normally distributed and suitable for parametric analyses (ANOVA) while species distribution and abundance data was not. As a result, distribution and abundance of species across sites was compared using the Kruskal-Wallis non-parametric ANOVA. For distribution, presence/absence data was used, with presence coded as = 1 and absence of species = 0. Simple Correspondence Analyses were then done, to investigate spatial associations of the species distribution and abundance within the different sites and nutrient distribution across sites. To assess diversity and similarity of the different sites, the following indices were calculated:

Seven different activities or drivers of change were identified namely agriculture, human settlement, logging, fuel wood collection, fishing, fish smoking and petty trading (Figure 2).

More than half of the population (52.6%) use mangroves as fuel wood, 27.6% use it for smoking fish, 13.2% use it for building, while 2.6% use the mangrove plants to make charcoal and 5.3% as props during the construction of storey buildings (Figure 3). Rhizophora mangle (red mangrove) was the most used species.

The perception of respondents on ecological changes in the Tiko mangrove swamps showed that 68.4% noticed changes in tree size, with the consequence that loggers have to move further into the creek to meet reasonable size of trees to cut. Some respondents reported an increase in water temperature (2.6%) and observed soil erosion in the mangrove swamps (18.5), while others noticed a reduction in water volume (5.3%). They reported a general drop in the quantity and quality of fish (22.4%) caught, as fisher men have to move further into the

creeks. Some respondents (31.6%) did not notice any changes in the mangrove swamps.

The properties of the sediments from the different sites varied as indicated in Table 2.

Sediment texture was dominated by clay followed by silt in all the sites. Sediment pH in KCL and water were generally low, ranging from 2.9 to 5.0, an indication of acidic conditions.

As expected, both organic carbon and organic matter levels were very high across all sites. In the current study, only site 3 (settlements and small scale agriculture) had organic carbon value of 4.82. While nutrient levels were generally moderate to high across sites, Mg values were particularly high for most of

*Effective Cation Exchange Capacity.

the sites. Mg concentration at Site 4 was 57.32 meq/100g while at Site 3 it was 4.72 meq/100g. In this study ECEC was high in all the sites with values ≥ 100.62 cmol/Kg. The values of the major cations (Ca, Mg, Na and K) showed a lot of variations in the soils in this study across sites. Ca levels ranged from 2.48 to 8.08 meq/100g.

A total of 141 algal species belonging to 53 families and 10 divisions were identified in all study sites. The division with the highest number of species was the Bacillariophyta (60 species) followed by Chlorophyta (38 species), Cyanophyta (16 species) and Euglenophyta (8 species) (Table 3).

The most abundant family was the Fragilariaceae with 14 species, while 32 families were represented by only one species each (Table 4). The most abundant species were Microcytis sp (24.09% at Site 7, 18.5% at Site 4, and 15.6% at Site 6), Chlorobotrys regularis (24.5% at Site 1, 17.1% at Site 5 and 14.8% at Site 3), Microceleus sp (12.9% at Site 5), Synedraacus var radians (18.5% at Site 2) and Tribonema sp (18.5%).

The mean abundance of benthic algae ranged from 0.59 ± 0.17 to 2.08 ± 0.58. Site 4 had the highest mean benthic algal abundance while Site 6 had the least (Figure 4).

The Kruskal-Wallis test comparing abundance across sites showed there exist significant differences between different sites (H = 42.52, DF = 6, P < 0.01) and the following groups showed significant differences; Sites 3 and 7 (Z = 4.93, P < 0.001), Sites 5 and 7 (Z = 4.54, P < 0.001), Sites 4 and 7 (Z = 4.35, P < 0.001), Sites 3 and 6 (Z = 3.93, P < 0.0001), Sites 5 and 6 (Z = 3.55, P = 0.0004), and Sites 4 and 6 (Z = 3.35, P = 0.0008).

Observed patterns of species abundance across sites were explained by two main components. The first two components explained 29.62% and 20.72% (Table 5) of the total abundance of species. These first two components cumulatively contributed 50.34% of the total inertia. The two components effectively explain species abundance across all sites (Qual = 0.977 and 0.026 respectively). Component 1 explained abundance of species at Site 2 (Corr = 0.896). Synedra acus var radians,Trachelomonas conica f punctata,Resultoe mikron,Fragilaria capucina,Epithemia sp and Closterium parvulum have highest abundance at Site 2, and occur at very low numbers in other sites. Sites 5 and 7 were best explained by Component 2 (corr = 0.533 and 0.283 respectively). Coscinodiscus alborani (10), Coscinodiscus radiatus (11), Microceleus sp (38), Palmella sp (18), Tabellaria sp,Gyrosigma spencerii,Goniochloris sp, andBiddulphia mobiliensis, have their highest abundance at Site 5. Similarly, Phaeocystis globosa, and Rhopalodia acuminate (10) have the highest abundance at Site 7, occurring in very low numbers across the other sites (Table 5, Figure 5).

Qual = the proportion of the row inertia represented by the two components. Coord = the principal coordinates of the rows. Corr r = the contribution of the component to the inertia of the row. Contr = the contribution of each row to the axis inertia, Inert = the proportion of the total inertia contributed by each row.

Site 3 was found to be the most diverse (H' = 3.81) while Site 7 was the least diverse (H' = 2.943). The most abundant species were recorded at Site 5 (4.246) and the least at Site 6 (2.045). Seventy two (72) species were recorded at Site 3, and the lowest number of species (33) was recorded at Site 7 (Table 6). Species evenness varied across sites, with site 3 being the most even (0.8909) while site 5 was the least (0.7822), an indication that site 5 had a few dominant species. The highest value of the Euglenophycean index was recorded in site 2 (0.21), while the lowest was in site 6 (0) as none of the species used in the calculation was present.

Several sites recorded a number of common species (Table 7), an indication of similar biophysical conditions. Sites 3 and 5 shared the highest number of common species (41) while Sites 6 and 7 had the least number (3).

Similarities between the sites are presented in Table 8. It was seen here that Sites 1 and 2 were very similar with their high value of 0.606. Sites 6 and 7 were the least similar (0.338). Despite their close proximity and similar activities, these sites vary in physicochemical properties as site 7 receives both industrial and urban effluents.

*Each value in the cell represents species which are common between two sites.

*Each value in the cell represents similarity between two sites. Sorenson range: 0 - 1, and sites are more similar as index approaches 1.

The mean benthic algal distribution had a ranged from 0.23 ± 0.04 to 0.51 ± 0.04. Benthic algae were more widely distributed at Site 3 and least at Site 7 (Figure 6).

The Kruskal-Wallis test showed significant differences in the distribution of benthic species between sites (H= 38.95, DF= 6 and P < 0.001). Site 1 was only significantly different from Site 7 (Z = 4.73, P < 0.0001), Site 5 was significantly different from Sites 6 and 7 (Z = 3.03, P = 0.0024 and Z = 4.37, P = 0.0007 respectively). The following pairs of Sites also showed significant differences. Site 3 and Site 7, Site 4 and Site 7, Site 3 and Site 6 (P = 0.0007), Site 4 and Site 6 (P = 0.0011), Site 2 and Site 7 (P = 0.0076).

Simple Correspondence analyses showed that the observed benthic species distribution can be explained by two main components. The contingency table indicated that Components 1 and 2 contributed 22.01% and 20.08% respectively of the total inertia (Table 9). The cumulative contribution of these two components was 42.09%. The two components effectively explain the pattern of distribution of species across all sites (Qual = 0.036 and 0.792). Component 2 best explains distribution of species at Site 4 (Corr = 0.79) and Site 6 (Corr = 0.341). Site 4 was highly associated with Achnanthes hungarica,Cenchridium globosum,Chroodactylon sp,Cosmarium sp,Fragilaria virescens,Dinophysis acuminoua,Euglena sp,Gomphonema sp,Mastogloia baltica,Phacus sp and Trachelomonas olonga which were not found in the other sites. Pleurotaenium cylindrium,Nitzschia intermedia,Nitzschia ignorata,Navicula sp2 andEuastropsis sp were found solely at Site 6. The distribution of species at Site 2 and Site 5 are best explained by Component 1 (Corr = 0.499 and 0.696 respectively) (Table 9). Palmella sp,Surirella spiralis,Stenopterobia ancepts,Nitzschia periata,Gyrosigma sp,Coscinodiscus alborani,Closterium kuetzingii f minus,Closterium kuetzingii f minus,Closterium arcuarium,Chlamydomonas penium, and Biddulphia mobiliensis were only found at Site 5. Similarly, Closterium parvulum,Epithemia sp,Fragilaria capucina,Resultoe micron andTrachelomonas conica f

Qual = the proportion of the row inertia represented by the two components. Coord = the principal coordinates of the rows. Corr = the contribution of the component to the inertia of the row. Contr = the contribution of each row to the axis inertia, Inert = the proportion of the total inertia contributed by each row.

punctata were present only in Site 2 and the rest of species observed were cosmopolitan across sites. Figure 7 shows how the species were associated with the different sites.

Pearson correlations of the different sediment (soil) parameters and benthic community diversity descriptors showed a strong positive correlation between

benthic algal diversity and Richness (r = 0.82 and P = 0.024). There was a negative correlation between benthic diversity and Organic Carbon (r = −0.844 and P < 0.05) benthic diversity and Organic Matter (r = −0.846, P < 0.05), benthic richness and available phosphorus (r = −0.785, P < 0.05), Base saturation and K (r = −0.916, P < 0.05).

Table 10 shows the correspondence analyses between nutrient and benthic algal diversity parameters across sites. The first two components explain 93.84% of the total inertia (Qual = 0.976 and 0.951 respectively). Analysis of the contingency

Qual = the proportion of the row inertia represented by the two components. Coord = the principal coordinates of the rows. Corr = the contribution of the component to the inertia of the row. Contr = the contribution of each row to the axis inertia, Inert = the proportion of the total inertia contributed by each row.

table indicated that Components 1 and 2 contributed 87.73% and 6.11% respectively of the total inertia. Component 1 best explains the patterns in Site 1, Site 3, Site 5 and Site 7 in terms of physicochemical and benthic parameters, while the patterns in the rest of the sites are best explained by Component 2.