Effect of Different Agroecosystem on Earthworm Diversity in Azaguié Locality (Côte d'Ivoire)

In most terrestrial ecosystems, earthworms are considered to be excellent bioindicators of biodiversity and soil quality. However, their diversity and abundance encountered depend on the systems considered and on the anthropic pressure exerted. The objective of this study was to assess the impact of a land use on the earthworm community. Earthworms were collected in TSBF (Tropical Soil Biology and Fertility) type monoliths by the direct manual sorting method in three types of farms (fallow, rubber plantation and mangosteen plantation) in the locality of Azaguié. Measurements of the physicochemical parameters of the soil, in particular, the total organic carbon level, the organic matter rate, the nitrogen rate, the conductivity, the hydrogen potential (pH) and the total phosphorus were carried out at the Laboratory Central of Agrochemistry and Ecotoxicology (LCAE) from 200 g of clod of soil sampled in different monoliths. These studies showed that in all the plots analyzed, the endogeic polyhumic worms presented the highest densities while the epigeic detritivorous worms presented the lowest densities. The distribution of earthworms in the mangosteen plot was homogeneous, unlike that of the other plots. However, the physico-chemical parameters of the soils were higher in the fallow. The canonical correspondence analysis carried out to assess the relationship between the variability of the earthworm density of the plots and the physicochemical parameters measured showed that the physicochemical parameters measured explained only a substantial part of this distribution of earthworms while the other part is possibly linked to other environmental factors.


Introduction
Environmental degradation due to inappropriate land use is a global problem that is becoming a growing concern in sustainable agricultural production systems (Ayoubi et al., 2011). Dynamic imbalances between the physical, chemical and biological properties of soils, which are continuously influenced by land use, lead to declines in productivity and sustainability (Somasundaram et al., 2013). In addition, many agricultural tillage activities have been identified as having irreversible consequences on the structure and dynamics of soil organisms such as earthworms, ants and termites (Montgomery, 2007;Evans et al., 2011). Yet these organisms play an important role in transforming the litter available for ecosystem functioning (Wardle, 2002). Among these organisms, earthworms are an important link in the functioning of the ecosystem (Lavelle and Spain, 2001). They are qualified as ecosystem engineers because of their great contribution to the recycling of biogeochemicals, which are the source of many ecosystem services (Lavelle et al., 2006). Indeed, these organisms participate in the decomposition of organic matter and the bioavailability of nutrients for plants and soil microorganisms. As well as contributing to the creation and maintenance of soil structure. Thus, their activities contribute to the maintenance of soil quality. Numerous studies have even shown that in ecosystems, their mode of distribution is conditioned by their relationship with plants, environmental parameters and human activities (Lavelle et al., 2006;Pelosi, 2008, Ehouman et al., 2014Butt and Briones, 2017). Other studies carried out on farms have shown that earthworm densities are generally between 50 and 400 worms per square meter, sometimes exceeding 1000 worms per square meter (Lee, 1985); the living biomass of earthworms is between 30 and 100 g per square meter (Lavelle and Spain, 2001) but can exceed 300 g per square meter (Lee, 1985) In cultivated environments, the density and biomass of earthworms are generally very variable and the size of the populations is smaller than in natural environments, which generally harbour a very large number of earthworms (Edwards and Lofty, 2009). Given the importance of earthworms in agricultural practices, this study aims to assess the impact of some land-use patterns on the earthworm community in the Azaguié region.

Study Site
The present study was carried out in the locality of Azaguié , precisely in Abbè Begnini, a village located between 5°35 and 6°15 North latitude and 3°55 and 4°40 West longitude. The site exploited for this purpose had a surface area of about 30 hectares with a vegetation cover marked by an alternation of fallow land (more than 25 years), Mangosteen and rubber trees.
The soil in the study area is clay-sandy with a fairly constant texture. But most of the soils in this locality belong to the class of ferrallitic soils (ferralsols) developed on Birmian schist of the arkosic type or sometimes variegated schist rich in silica or green schist.

Sampling Device
For each type of farm (fallow, rubber tree and mangosteen tree), square sub-plots of 50 m each side were first delimited using a GPS of the Gamin brand. Each sub-plot was subdivided into 100 squares of 5 m each numbered from 1 to 100. Then, within each subplot, a random draw without discount was carried out in order to choose 30 squares, each of which had a TSBF (Tropical Soil Biology and Fertility) type monolith 50 cm x 50 cm by 30 cm deep in its centre to collect earthworms and clods of soil ( Figure 1). ISSN 2157-6076 2020

Collecting Earthworms and Soil Clods
The earthworms were collected manually in the 0-10 cm, 10-20 cm and 20-30 cm strata using the method of (Lavelle, 1978). The worms were first stored in labelled jars containing 4% diluted formaldehyde. The jars were then transported to the laboratory of the Centre de Recherche en Ecologie (CRE) where species identification was carried out using the determination keys in (Csuzdi and Tondoh, 2007), based on morphometrics (shape, size, length of segments) and pigmentation of the individuals. Soil samples were also taken at the level of each stratum using a trowel and then packed in carefully labelled stomacher bags. These clods of soil were kept in coolers equipped with carboglace, then sent to the Central Laboratory of Agrochemistry and Ecotoxicology (LCAE) for the determination of the physico-chemical parameters of the soil.

Physico-Chemical Parameters
The physico-chemical parameters of the soils were determined from 200 g of each clod of soil sampled, parameters such as total organic carbon (% CO), organic matter (% OM), nitrogen (% N), conductivity (Cond), hydrogen potential (pH) and total phosphorus (Pt).
Nitrogen was determined by the Kjeldahl method described in the French standard NFV04-407 (AFNOR, 2004). The mineraliser used was the Speedi-gesteur type (Buchi, Switzerland) and the distiller was type FB15025 (Fisher Scientific, USA). The conductivity was determined using the diluted extract method (Montoroi, 2017). The pH (water) of the soil was determined electrometrically using an HQ40d multimeter (Hach, USA), according to the pH determination protocol developed by the Centre d'Expertise en Analyse Environnementale du Qué bec . Phosphorus was determined using the AOAC-958.01 method (AOAC, 1990). Phosphorus concentrations were determined using a UV/VIS spectrophotometer type UV Probe 1700 (Shimadzu, Japan).
Organic carbon and soil organic matter were determined by the fire loss method using a muffle furnace (Nabertherm, Germany) Hasine, 2008). The percentages of organic matter and organic carbon were obtained gravimetrically according to the following formula: Where: M0: 5 g of soil sample completely dried in an incubator at 105°C for 48 hours; M1: Platinum crucible mass (g); M2: Crucible mass + the ash of the soil sample after incineration in a muffle furnace (Nabertherm, Germany) at 550°C for 4 hours (g);

MO: Organic matter;
Journal of Biology and Life Science ISSN 2157-6076 2020 TOC: Total organic carbon; 1.724: conversion factor of organic matter to organic carbon.

Statistical Analysis
The distribution of the earthworm community across different land uses was assessed using the species richness and diversity index of Shannon and Weaver (Barbault ,1992) according to the following formula. Equitability (E), which provides information on the distribution of numbers between the different species (Barbault,1992), was deduced. This is the ratio of actual diversity to maximum theoretical diversity and is expressed by the following equation: Where: E: Equitability, H'max: maximum value that diversity can take, S: total number of species.
The higher H' is, the higher H'max is and the greater the diversity. The higher E is, the more homogeneous the environment is with respect to the distribution of species. The value of this index varies from 0 to 1; it is maximum when the species have identical abundances in the stand, and minimum when a single species dominates the whole stand.
The one-factor analysis of variance (ANOVA) with repetition, followed by the Tukey test, was used to make pairwise comparisons between the physico-chemical parameters of the different plots. These tests were carried out using Xlstat Pro version 7.5 software.
Canonical Correspondence Analysis (CCA) was used to measure the links between the dependent variables (species density) and the explanatory variables (physico-chemical parameters) of each plot (fallow, mangosteen and rubber trees). The data were analyzing with Xlstat 2015.4.01 software. For all the statistical tests carried out, the significance threshold was set at 5%.
In all the plots, the highest worm densities were obtained in the 0-10 cm stratum (11.56 ± 0.88 ind/m 2 (fallow); 37.00 ± 0.90 ind/m 2 (mangosteen) and 24.0 ± 0.80 ind/m 2 (rubber)) and the lowest densities were obtained in the 20-30 cm stratum (0.05 ± 0.00 ind/m 2 (fallow); 0.01 A B ISSN 2157-6076 2020 ± 0.00 ind/m 2 (mangosteen) and 0.03 ± 0.00 ind/m 2 (rubber)). However, for all strata, the abundance of earthworms was highest in Mangosteen soils, followed by Hevea soils and then fallow soils (Table 1). The Tuckey test used for the pairwise comparison of the strata indicated that at the 0-10 cm stratum, there was a significant difference between the worm densities in the mangosteen plot and those in the fallow and rubber tree plots (P < 0.05). On the other hand, in the 10-20 cm and 20-30 cm strata, no significant difference was observed (P > 0.05). Moreover, the Shannon-Weaver diversity index was greater for the mangosteen tree plot (2.81) than for the fallow (2.78) and rubber tree plots (2.61). Their respective equitabilities were 0.76, 0.68 and 0.64 (Table 2). This implies that the distribution of earthworm species was more homogeneous in the mangosteen plot than in the fallow and rubber tree plots. In general, the values of the different physico-chemical parameters determined in the sampled plots (fallow, mangosteen and rubber trees) were higher in the 0-10 cm stratum and lower in the 20-30 cm stratum, Thus:

.3 Interaction Between Earthworms and Physico-Chemical Parameters
Canonical correspondence analysis (CCA) showed that the physico-chemical parameters (organic matter, pH, conductivity, total phosphorus and C/N ratio) explain only a small part of the variability in earthworm density in each of the plots (23.90%, 16.62% and 42.47% of the total inertia respectively in the fallow, mangosteen and rubber tree plots) (Table 4a; 4b  and 4c). Thus, the variability in earthworm density in the fallow, mangosteen and rubber tree plots is mainly linked to environmental parameters other than the physico-chemical parameters determined (76.10%, 83.38% and 57.53% of the total inertia respectively at the level of the fallow, mangosteen and rubber tree plots) (Tables 4a, 4b and 4c). The Monte Carlo permutation test indicated that for each plot the canonical axes were not a linear combination of environmental parameters (p = 0.751 > 5%; p = 0.964 > 5% and p = 0.832 > 5% for fallow, mangosteen and rubber trees plots respectively).

Discussion
For all the plots analysed, polyhumic endogenous worms had the highest densities. while detritus-feeding epigeic worms had the lowest densities. The difference in density observed between these categories of earthworms is due, on the one hand, to the fact that they exploit different horizons for collecting their food and, on the other hand, to the constraints specific to their survival environment. Indeed, polyhumic endogenous worms generally live in the first few centimeters of soil. They thus build a network of sub-horizontal galleries from which they consume soil from superficial horizons rich in decomposing organic matter (Brown et al., 2000). This shelters them from the bad weather linked to the constraints of the environment. However, detritivorous epigeic worms live in the surface litter and feed on decomposing organic matter (Brown et al., 2000;Pelosi et al., 2008). As a result, they are the most exposed to climatic hazards, predation and cultural operations such as tillage and pesticide use. These factors adversely affect their diversity, density and biomass in the cultivated environment (Smeaton et al., 2003). The distribution of earthworms in the mangosteen plot was homogeneous, unlike that of the other plots. This reflects an accelerated mineralization of the organic matter produced by the leaves and twigs of the plant. This favours an increase in worm density (Cluzeau, 2005). On the other hand, the low diversity and abundance of earthworms in the fallow plot could be due to the importance of the diversity of fauna and flora, which could favour the presence of numerous predators and competitors (termites, ants...). Indeed, these predators have a negative impact on the density and distribution of earthworms during the search for food (Dominguez et al., 2001). Also, the presence of large roots due to the non-exploitation of this environment would prevent the creation of galleries (Houseman et al., 2000). Furthermore Lavelle et al. (2006), describing the dynamics of a worm population, concluded that in cultivated systems, rainfall and soil moisture are the elements that would most significantly influence the dynamics of the worm population. Temperature and light are therefore not excluded from the factors that could explain the difference in earthworm diversity observed between the three plots.
However, the average values of the physico-chemical parameters were higher in the fallow land. This could be explained by the stability of this ecosystem. However, in this environment, the stock of organic matter was poorly used by the plants. This is contrary to the cultivated environment, where organisms exploit this matter. The result is an increase in the density of earthworms, which could modify the physical properties of the soil (Cluzeau et al., 1987). Agricultural practices that could influence earthworms are ploughing and the use of pesticides (Chan et al., 2001, Thomas, 2004, Pelosi et al., 2014. But in this study, tillage was excluded because it was not used in the establishment and maintenance of mangosteen and rubber plantations. Nevertheless, a C/N ratio of less than 15 in the soils of these three plots would mean accelerated decomposition of organic matter. This rapid mineralisation of organic matter could be a corollary of intense biological activities in the soil (Soltner, 2000). These observations are contrary to recent studies which have shown a C/N ratio higher than 50 in the Azaguié locality, which would indicate a slow mineralisation of organic matter at soil level (Kouamé et al., 20014;N'guessan, 20014). Among the physico-chemical parameters, the pH values showed that the three plots had acidic soils in the light of the pH scale commonly used by laboratories (Zro Bi, 2012). These pH values were relatively higher than the pH values (4.7-5) obtained by Kabrah et al. (2000) for the soils under palm groves in the Mé region (Côte d'Ivoire). Moreover, the pH values of the present study are very close to those for Kouamé (2014) where the pH values were between 5.22 and 5.39 for the soils of Yamoussoukro, and between 5.90 and 6.00 for the soil of Azaguié . The results of these authors also confirm the near constancy of the pH up to 30 cm depth. Finally, the acidic character of the Azaguié soils was also reported by (Bi Voko et al., 2013) who recorded pH values < 5.17. Studies carried out by (Singh and Schluze, 2015) showed that mineralization of organic matter would release mineral elements in the soil in ionic form in solution. In addition, the mangosteen, rubber and fallow plots had soils with low total phosphorus content according to the classification scale of the "Comité d'Etude et de Development de la Fertilisation" (COMIFER) (Vedie, 2008). These total phosphorus contents were lower than those obtained by Kouamé (2014), who found total phosphorus concentrations of 229.08 mg/kg and 214.75 mg/kg respectively for the 0-20 cm and 20-40 cm strata of the Azaguié soil. This difference could be due to plant cover and agronomic exploitation. Indeed, these values were obtained on a dense forest soil whereas our study site presented soils in exploitation (mangosteen and rubber tree plots) and a 25 years old fallow.
The variability in earthworm density in the plots analysed (fallow, mangosteen and rubber trees) was explained in part by the physico-chemical parameters (organic matter, pH, conductivity, total phosphorus and C/N ratio). This would be due to the low percentages of CCA inertia that provide information on this relationship (26.67% and 32.87% respectively for the mangosteen and fallow plots). This suggests that environmental parameters, such as the physico-chemical parameters used in the CCA, are involved in the variability of worm density in the different plots. At the fallow level, parameters such as humidity, temperature and light could be impacted by the variation in worm density. According to Edwards and Bohlen (1996) and Faurie et al. (2012), these factors greatly influence the life of these organisms. Concerning mangosteen and rubber tree plots, in addition to humidity, temperature, and light, agricultural practices, specifically the use of pesticides such as herbicides, could probably impact on the survival of earthworms because they are much more sensitive to them (Chan, 2001;Thomas, 2004). Indeed, according to observations, the use of glyphosate and 2,4-D herbicides has been the preferred method of weed control on these farms for more than 10 years. Thus, the contribution of these herbicides to influencing the survival of earthworms cannot be excluded because of the notorious toxicity of herbicides formulated on earthworms. To this end, the work of Zarea and karimi (2011); Kpan kpan (2017) and Kpan Kpan et al. (2018) has shown an increased toxicity of glyphosate and 2,4-D herbicides in earthworms, particularly in the species Eudrilus eugeniae. These herbicides significantly reduce egg-laying, the rate of cocoon hatching and the number of juveniles produced from each.

Conclusion
The present study has shown that polyhumic endogenous worms seem to be the most adapted to the different modes of land use in the locality of Azaguié . For the farms analysed (fallow, mangosteen and rubber trees), the mangosteen plot was favourable to the development and distribution of worm communities. The physico-chemical parameters of the soil, particularly organic matter, pH, conductivity, total phosphorus and C/N ratio, were well illustrated in the fallow plot. Moreover, in all the plots, the survival and distribution of earthworms seemed to be affected by other environmental factors. The survival of earthworms on a farm would therefore depend on the plants that are cultivated there and the farming practices.