The biological material consisted of animal and plant biological material.

· The animal biological material consisted of adult snails of the species Achatina fulica whose average snail mass, length and average shell diameter were 138 ± 10 g, 14 ± 2.4 cm and 11 ± 1.5 cm respectively.

· The plant material was mainly used for feeding snails. It consisted of leaves and fruits of papaya (Carica papaya).

Three herbicides were used in this study. They are:

· Glycel 410 SL from the firm Topex Agro-élevage Développement SARL. The active ingredient was glyphosate at 410 g (41% isopropilamine salt) per liter of commercial solution. To achieve the desired effect, the manufacturer recommended diluting 200 mL of product in 15 L of water per 1/2 hectare of land. This rate corresponded to a solution of 5.47 g/L.

· Extra 720 SL from Crop Doctor. The active ingredient was 2,4-D formulated as a soluble liquid (SL) at 720 g (amine salt) per liter of commercial solution. The manufacturer recommended the use of 40 mL to 50 mL for a 16 L water sprayer. These dosages corresponded to solutions of 1.8 g/L to 2.5 g/L respectively.

· Sofa 40 SC distributed by the firm AF-CHEM SOFACO. The active ingredient was nicosulfuron. The formulation used was a suspension concentrate (SC) with 40 g of nicosulfuron per liter of commercial solution. The manufacturer recommended diluting 1 L of the product in 60 L to 200 L of water to treat 1 ha. These rates were 0.66 g/L to 0.2 g/L.

The purpose of this operation was to standardize the age of the snails and to have a large number of individuals available for ecotoxicological tests.

1) Collection of broodstock

Adult snails of the species Achatina fulica intended for breeding were obtained from the local market. They were weighed and then transferred to tanks for breeding. The criteria for their selection were essentially morphological. The average weight mass of the snails, the length and the average diameter of the shell were respectively 138 ± 10 g, 14 ± 2.4 cm and 11 ± 1.5 cm. The breeding was carried out in the laboratory.

2) Management of the breeding

The breeders were distributed in five breeding tanks at a rate of 12 animals per tank. Each tank was labelled according to the weight mass of the individuals. The snails were fed every two days with leaves and fruits of papaya at a rate of 10 g for the leaves and 50 g for the fruits, that is to say 60 g of food per tank. The decomposed wood, previously dried and crushed, was used as litter. The thickness of the litter was 10 cm. To humidify the environment, 500 mL of fountain water was added. The bedding was changed twice a week. The bins were inspected daily between 9 and 10 am to remove waste from the the environment and to recover eggs in case of egg laying.

The eggs obtained in the bins were then transferred to plastic boxes. The boxes were 30 in number. Each of these boxes contained 200 g of dried and ground decaying wood and 50 broodstock eggs. The whole was moistened with 50 mL of fountain water. A total of 30 environments were prepared. After about two weeks of incubation, the eggs hatched. Two (2) g of papaya leaves were placed in the environment after the eggs hatched to serve as food for the spat. The food was renewed every two days. The four-week-old spat were placed on environment of the same composition as the rearing environment at a rate of 20 individuals per tank. Ecotoxicological tests were carried out on this second generation of Achatina fulica individuals. The culture environment was prepared in plastic tanks of dimensions 42 cm × 32 cm × 22 cm.

The test protocol was inspired by the ISO 15952 standard on the effects of pollutants on juvenile snails (Helicidae) and the determination of effects on growth by soil contamination and then by the guideline No. 243 for testing chemicals, developed by the Organization for Economic Cooperation and Development [17] [18].

1) Preparation of the microcosms

The experimental plot was subdivided into four elementary plots 5 m apart. The area of each elementary plot was 1350 m². These plots were not previously treated with herbicides. 24 h before the beginning of the test, 5 microcosms were prepared and placed on each elementary plot.

They were each made up of circular frames of 22 cm × 10 cm sunk to a depth of 5 cm in the soil so that they were firmly attached. Three juveniles of the species Achatina fulica were placed in each microcosm. Before the start of the test, measurements of the snails were taken. The length, shell diameter and mass of the snails were 3 ± 0.5 cm; 1.8 ± 0.4 cm; 5 ± 1 g respectively. A quantity of 50 g of food was deposited every three days in each environment, i.e. 10 g of leaves and 40 g of papaya fruits. The experimental units were then covered with a mosquito net to prevent the escape of snails and the intrusion of other organisms. Wooden stakes were used to hold the nets to the ground. A total of 20 microcosms were used throughout the experimental plot. The test was conducted in undergrowth, a preferred environment for snail development.

2) Herbicide treatments of the experimental plot

Each elemental plot underwent a single herbicide treatment of 2,4-D, glyphosate or nicosulfuron at concentrations recommended by the manufacturer. The solutions were prepared with fountain water. Another elementary plot close to the previous ones is used as a negative control.

Some pawpaw trees at the study site were treated with 2,4-D, glyphosate, or nicosulfuron at the beginning of the test, at a rate of one pawpaw per herbicide. The leaves and fruit of these pawpaws were used as food for the snails during the test. The first foods deposited in the microcosms were not pre-treated. They were treated at the same time as the experimental plot. Another untreated papaya tree was used as food for the control snails.

The snails stayed in the environment for 28 days. Snail mass, length and shell diameter were measured weekly. Each snail was weighed individually. The length and diameter of the shell were measured with a caliper. The growth of exposed and control individuals was calculated from the increase in length, diameter and weight mass of the snail from the beginning to the end of the test. During the exposure to the different treatments, the more or less disturbed behaviors of the snails were described.

The effect of 2,4-D, glyphosate and nicosulfuron herbicides on reproduction, is assessed by the number of eggs laid and the hatching rate, using the guideline No. 243 for chemical testing, developed by the Organization for Economic Cooperation and Development [18].

1) Effect on egg production of Achatina fulica snails

The test to evaluate the effects of 2,4-D, glyphosate, and nicosulfuron on snail reproduction follows the same procedure as the previous test on growth. The only differences are in the number and age of individuals used per microcosm. Two individuals of 5 months of age from the farm and weighing 130 ± 3 g were placed in each experimental unit. Eggs were counted per week in each experimental chamber.

2) Effect on the hatching rate of Achatina fulica snails

The experimental unit was a 500 mL transparent plastic box. These boxes contained 200 g of decomposed wood litter previously crushed and dried, to which were added 50 eggs of Achatina fulica from the breeding environments. 24 h after oviposition, the microcosms were buried in holes whose bottom was lined with stones to ensure efficient water drainage and so that the level of the boxes was 2 cm above the surrounding soil. The microcosms were humidified with 50 mL of fountain water to bring the temperature of the environment to 26˚C ± 1˚C which is the optimal temperature for egg incubation. The lid of the boxes was also perforated to allow good aeration of the environment. Note that this study was conducted at the same time as the study of the effects of 2,4-D, glyphosate, and nicosulfuron on the growth and reproduction of juvenile Achatina fulica snails. Microcosms for both tests were arranged horizontally at a distance of 1 m from each other. 10 microcosms were used for this test and placed on each elementary plot. These plots were treated with a single type of herbicide, either 2,4-D, glyphosate or nicosulfuron. Another plot close to the previous ones served as a negative control. The environment was visited every five days. The hatching rate was determined by the following equation:

HR ( % ) = NEH × 100 TNE (1)

where:

HR (%): hatching rate;

NEH: number of eggs hatched;

TNE: total number of eggs put in culture.

Some behaviors due to the exposure of snails to herbicides could be observed during this study. These are: the ability of snails to feed normally or avoid food, the activity of the animals and their spatial position in the test chamber: for example, active or inactive in the high position or on the supports, or active or inactive on the food [13]. These behaviors are observed, on days of food change. The snails were inactive at the time of observation. They hid at the bottom of the microcosm in a hole of about 3 cm that they dug in the ground. After one minute, the snails started to be active. In the microcosms in the control and nicosulfuron-treated plots, no food remains were observed (Figure 2(a)). In contrast to the above, a significant amount of decaying feed residue was observed in the microcosms on the glyphosate and 2,4-D treated plots (Figure 2(b)). The greatest amount of feed was observed in the microcosms treated with 2,4-D.

Statistical analyses at this level showed that growth parameters varied with the different types of 2,4-D, glyphosate and nicosulfuron treatments. Snail weight mass, shell length, and shell diameter showed significant differences at the 5% threshold (Table 1). Compared to controls, 2,4-D, glyphosate and nicosulfuron inhibited snail growth. In order of toxicity, our results show that nicosulfuron is less toxic than glyphosate, which is almost equal in toxicity to 2,4-D.

1) Effect on egg production

Egg laying took place 7 days after the establishment of the microcosms on the test and control plots. Moreover, the average number of eggs laid per microcosm and per plot was the same.From day 7 to day 14, the number of eggs laid increased significantly, reaching a peak around 200 eggs laid in the control snails. For snails treated with the different herbicides, a significant drop was observed. Oviposition was zero from day 14 to day 42 for snails in plots treated with 2,4-D and glyphosate. In the nicosulfuron-treated plot, egg laying dropped to zero on day 21. Egg laying resumed on day 28 with an average of 150 eggs per microcosm and decreased slightly on day 35 before dropping sharply on day 42 with an average of 20 eggs. In the control snail plot, from day 14 onwards, egg laying began to decline progressively on days 21, 28, 35 and 42. This egg laying which was around 200 eggs on day 14 went down to around 80 eggs on day 42. Looking at the control curve, we can see that the egg laying starts on day 7, reaches the peak on day 14 and decreases from day 21 to day 42 (Figure 3).

The mean number of eggs ± standard deviation laid per pair of snails in the contaminated environment was 41.66 ± 54.37; 48.83 ± 48.55 and 62.5 ± 57.81 for 2,4-D, glyphosate and nicosulfuron, respectively, compared to 123.16 ± 31.01 in the control environment. Statistical analyses revealed a significant difference at

the 5% threshold between the different treatments and the number of eggs laid. From the averages obtained, it was found that all herbicide treatments of 2,4-D, glyphosate and nicosulfuron affected egg laying compared to the controls. In terms of toxic effect on oviposition, Tukey’s HSD test revealed that 2,4-D, glyphosate and nicosulfuron were equal (Table 2).

2) Effect on hatch rate

The first hatching of eggs deposited on the 2,4-D and glyphosate-treated plots was observed at 19 days after oviposition while those on the nicosulfuron-treated plot were observed at 17 days after oviposition. The incubation time of eggs deposited on the control plot was 15 days. The average hatching rate ± standard deviation of eggs from the 2,4-D, glyphosate and nicosulfuron contaminated plots were 5% ± 2%; 20.8% ± 4.2%; and 49.6% ± 11.86%, respectively. In the microcosms on the control plot 82.4% ± 5.17% of the deposited eggs hatched. The Kruskall Wallis ANOVA test revealed a very highly significant difference between hatching rate and herbicide treatment (˂0.05). This result reflects a reduction in hatching rate due to 2,4-D, glyphosate, and nicosulfuron. The incubation time of the eggs was different depending on the herbicide product used. Based on the results obtained, 2,4-D was more toxic than glyphosate. Glyphosate was more toxic than nicosulfuron.