Introduction
Sheep farming is of global importance, given that these animals are raised on every continent (Skapetas & Kalaitzidou, Reference Skapetas and Kalaitzidou2017). Moreover, sheep offer an alternative to cattle, which are more severely affected by the frequent droughts attributed to climate change. Thus, sheep present an attractive option for the optimization of grazing areas, their easy management and for product diversification (Lucena et al., Reference Lucena, Martins, Souza and Magalhães2017).
Today, the main obstacle to the productive efficiency of sheep grazing is gastrointestinal nematodes (GIN), which cause significant productivity and financial losses and impair the well-being of these animals (Charlier et al., Reference Charlier, Morgan, Rinaldi and Van Dijk2014; Raineri et al., Reference Raineri, Nunes and Gameiro2015; Oliveira et al., Reference Oliveira, Ruas, Riet-Correa and Coelho2017). The inappropriate use of anthelmintics as the main strategy to reduce parasite burdens has led to the emergence of parasite populations resistant to most of the active ingredients used in chemical control, rendering them ineffective (Neves et al., Reference Neves, Carvalho, Rinaldi and Cringoli2014; Bichuette et al., Reference Bichuette, Lopes, Gomes and Felippelli2015). Environmental contamination and chemical residues of synthetic anthelmintics pose a threat to ecological dynamics and a serious risk to public health (Beynon, Reference Beynon2012; Beynon et al., Reference Beynon, Mann, Slade and Lewis2012; Silva et al., Reference Silva, Moreira and Peres2012; Siroka & Svobodova, Reference Siroka and Svobodova2013; Köksal et al., Reference Köksal, Kalin, Gülçin and Özdemir2016; Goodenough et al., Reference Goodenough, Webb and Yardley2019).
To address this problem, it is necessary to implement comprehensive endoparasite control plans aimed at reaching both GIN life stages, that is, their free and parasitic life stages. The purpose of controlling both stages is to reduce parasitic contamination of the environment and contact of infective larvae (L3) with animals, thereby reducing the parasite load in hosts (Santos et al., Reference Santos, Silva and Amarante2012; Martins Costa & Amarante, Reference Martins Costa and Amarante2015).
The L3 development, migration and survival are strongly influenced by the climate and microclimate within the forage canopy of each plant (Levine & Todd, Reference Levine and Todd1975; Wang et al., Reference Wang, Vineer, Morrison and Van Wyk2018). The differences in the characteristics of plants that make up the pasture provided to animals can influence larval dynamics both directly, through morphological structures such as pilosity, which acts as a physical barrier to L3, or indirectly, through canopy characteristics that alter the microclimate, affecting larval dynamics (Knapp, Reference Knapp1964; Silangwa & Todd, Reference Silangwa and Todd1964; Oliveira et al., Reference Oliveira, Costa, Rodella and Silva2009; Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2014). In addition, metabolizable proteins, vitamins and minerals in the diet are essential for strengthening the immune system and the body's resilience against parasitic infections, being more critical in the peripartum period of ewes and in growing lambs (Rocha et al., Reference Rocha, Amarante and Bricarello2004; Oliveira et al., Reference Oliveira, Miguel and Alves2019).
In this context, one of the strategies for the comprehensive control of parasites in sheep raising is forage planning, using forage species that hinder the migration and survival of L3 in the pasture while simultaneously providing high quality nutrition. This strategy can be extremely relevant for the categories most susceptible to parasitic infections, such as growing lambs and ewes in the peripartum period. Temperate forage legumes exhibit promising potential for controlling GIN. Their high protein content, digestibility and acceptance by ruminants makes this group an excellent option for protein supplementation in the diet of grazing sheep (Glienke et al., Reference Glienke, Gomes, Camargo and Pötter2010; Ates et al., Reference Ates, Keles, Yigezu and Demirci2017; Olubunmi et al., Reference Olubunmi, Oyaniran, Ogunsakin and Aderinboye2019). Moreover, some of these legume forage plants present bioactive properties in controlling GIN, especially due to their content of condensed tannins and flavonoids (Niezen et al., Reference Niezen, Waghorn, Charleston and Waghorn1995; Oliveira et al., Reference Oliveira, Bevilaqua, Morais and Camurça-Vasconcelos2011; Wang et al., Reference Wang, Mcallister and Acharya2015). Temperate forage legumes are also able to fix 100 to 200 kg ha−1/year of nitrogen, improving pasture quality and contributing to the implementation of multiple ecological relationships in the soil–plant system. These relationships, in turn, can enhance biota diversity and the physical and chemical properties of soil (Schultze-Kraft et al., Reference Schultze-Kraft, Rao, Peters and Clements2018). Lastly, due to their highly diverse morphology, they can generate variable microclimates and constitute physical obstacles to L3, affecting the survival and migration of GIN in the pasture (Oliveira et al., Reference Oliveira, Costa, Rodella and Silva2009).
According to Carvalho et al. (Reference Carvalho, Santos, Gonçalves, Moraes, D and J2010), due to their adaptability, the temperate forage legumes most commonly cultivated in Brazil are: white clover (Trifolium repens L.); bird's-foot-trefoil (Lotus corniculatus); and red clover (Trifolium pratense L.). These forage legumes present different morphophysiological characteristics that can affect the survival and migration of GIN in the pasture, and hence, the intensity of parasitic infections.
The purpose of this study was to determine the effect of cultivation of different temperate forage legumes on the survival and migration of GIN of sheep in the pasture. Thus, the idea is to establish forage plants that favour pasture decontamination while simultaneously providing a high-quality diet for grazing sheep. The findings of this study are expected to broaden the body of knowledge about the ecology of the most important GIN in sheep farming in the coastal region of southern Brazil.
Materials and methods
Study site and climate data
The study was carried out at the Center for Agroecology Research and Extension of Fazenda Ressacada, which is part of the School of Agricultural Sciences of the Federal University of Santa Catarina, located in the Tapera neighbourhood of Florianópolis, in the state of Santa Catarina (SC), Brazil (27°41′06.28″ S 48°32′38.81″ W), at an altitude of 2–4 m above sea level. This region has a homogeneous rainfall, and a temperate tropical humid mesothermal climate with high levels of humidity and rainfall distributed throughout the year. The soil is classified as a typical hydromorphic quartz arenite neosol, consisting mainly of dark sand with a high content of organic matter and the presence of a high-water table (Gonçalves dos Santos et al., Reference Gonçalves dos Santos, Klinger, Cunha Anjos and Oliveira2013).
Experimental design
Forage legumes were sown in separate experimental plots in August 2018. After the establishment of the forage legumes, they were artificially contaminated with faecal pellets from donor sheep naturally infected with GIN. The legumes were evaluated over an 8-week period to assess the survival and horizontal and vertical migration of larvae. The horizontal migration was assessed at two distances from the faecal pellets, 10 cm and 30 cm, and the vertical migration in two plant strata, lower (lower stratum) and upper (upper stratum) half of the plant (Carneiro & Amarante, Reference Carneiro and Amarante2008; Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2008). The evaluations were carried out from October to December 2018 based on four samplings: in weeks 1, 2, 4 and 8 after the deposition of faecal material on the plots. In addition, in week 6, the forage in all the experimental units was cut in order to simulate real grazing situations (table 1). The cut pasture was collected, as well as the remaining faeces and a soil sample from under the faeces. L3 were recovered from each sample for further identification.
Table 1. Experiment timeline.
Grass cutting at 10 cm from soil to simulate grazing.
The experiment was conducted with a four-factor factorial design in blocks with subdivided plots. Four factors with different numbers of levels were evaluated:
a) Legume species with three levels:
-
Species 1: T. repens L. (white clover)
-
Species 2: T. pratense L. (red clover)
-
Species 3: L. corniculatus cv. São Gabriel (bird's-foot-trefoil)
-
b) Time factor with four levels: weeks 1, 2, 4 and 8.
c) Distance from faeces with two levels: ten and 30 cm in diameter.
d) Vertical stratum with two levels: upper stratum and lower stratum.
The plot was a specific area 30 cm in diameter where there was one legume species. The sub-plot was represented by the distance from the faeces, and the sub-sub-plot was represented by the vertical stratum (fig. 1). Each treatment contained the four factors and had four replicates grouped into four blocks, blocking for ‘sample collection day’, so each sample collection was distributed over two to three consecutive days in order to ensure the logistical feasibility of each sampling.
Fig. 1. Description of the experimental plot.
Sowing
A 441 m2 experimental area was divided into four blocks. Each block contained three seedbeds, one for each forage legume species. Each bed was divided into four plots of equal size, 1.20 m by 1.20 m, corresponding to the number of sample collections scheduled over the study period.
The experimental area was initially ploughed to prepare the seedbeds. Beds measuring 1.20 m in length and 7.20 m in width were prepared for each forage species. Oversowing was carried out by hand, after which the area was trampled to simulate animal activity.
The species sown were white clover (8 kg of seed per hectare), and red clover and bird's-foot-trefoil (both with 10 to 12 kg of seed per hectare). As this procedure consisted of oversowing, the quantity of seeds corresponded to twice the amount recommended for sowing (Carvalho et al., Reference Carvalho, Santos, Gonçalves, Moraes, D and J2010).
An area of 30 cm in diameter with the best establishment of the legume was selected and demarcated in each bed. Plants other than the corresponding forage legume were weeded out weekly to ensure that the experimental plots remained monophyte. Monoculture was used to ensure the best establishment of the forage legumes in the beds and to evaluate the specific effect of each legume on the dynamics of larvae in the pasture.
Faecal donor animals and contamination of the experimental plots
The faeces used for artificial contamination of the experimental beds were obtained from four rams of the Crioula Lanada breed, naturally infected with GIN, exhibiting mild infections. Faecal pellets to contaminate the beds were collected five days prior to being deposited on the experimental beds. The faeces were collected using sheep faecal collection bags, which are normally used for metabolic assays.
Faecal bags were placed on two animals at a time, 6 h per day, for five days. A total of 2 kg of faecal pellets were collected, enough to contaminate all the experimental beds and for the control samples (1060 g). The collected faeces were stored at 10°C until the moment when they were spread on the experimental beds (Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2007).
On the day of contamination of the experimental units, the faeces were mixed carefully to ensure that the pellets were not broken up, with the objective of depositing the pellets in their integral form in the experimental units. They were then weighed and separated into 20 g samples, which were placed in plastic bags and refrigerated until the moment of deposition.
Before spreading the faecal pellets, the beds were mown down to 10 cm from the ground to simulate grazed pasture, conserving the height after grazing recommended for the legumes used (Fontaneli et al., Reference Fontaneli, Santos and Fontaneli2012), and standardizing the height of the forage plants, as described by Silva et al. (Reference Silva, Amarante and Kadri2008). Immediately after mowing, 20 g of faeces were deposited in the centre of each of the 48 plots. Five additional faecal samples of 20 g were used as control samples to count the eggs per gram of faeces (EPG) (Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2014) and for a faecal culture control, with the purpose of determining the average number of eggs with which the experimental units were contaminated, the rate of development of the eggs and the genera of GIN present. The faecal cultures were performed using Roberts & O'Sullivan's (1950) technique and the L3 were identified as described by Keith (Reference Keith1952).
Based on an analysis of the faeces control samples the experimental plots were contaminated with an average of 150 EPG, which equates to an average of 3000 eggs deposited on each experimental plot, with an average development rate of 14.50% (435.2 L3 on average). After faeces culture, the following nematode genera were identified: Trichostrongylus spp., Cooperia spp., Haemonchus spp., Oesophagostomum spp. and Ostertagia spp.
Data collection
In each collection (fig. 1), one plot per species and per block was used, making a total of 12 plots at a time. The plots were chosen according to the week of data collection, and collection always started at 7 am.
Height
The minimum and maximum heights of the forage were measured to determine ‘Distance 10,’. The height was measured at four points, forming a cross, to determine ‘Distance 30’. These heights were averaged to generate the variable ‘average height.’
Collection of pasture, faeces and soil
The total of the remaining faeces and a sample of soil taken from approximately 2 cm below the faeces were collected from each plot. The quantity of the soil sample collected was what fits in a full soup spoon (Santos et al., Reference Santos, Silva and Amarante2012). Each of the pasture, faeces and soil samples was placed in a properly tagged plastic bag.
Four samples of pasture were collected per plot. All the sward of the upper half (upper stratum) and lower half (lower stratum) of the forage (close to the ground) was collected, separately, at each distance (10 cm and 30 cm) (Almeida et al., Reference Almeida, Castro, Silva and Fonseca2005).
The L3 were recovered from the pasture, soil and faeces using the Baermann technique, with adaptations by Niezen et al. (Reference Niezen, Miller and Robertson1998b) and Rocha et al. (Reference Rocha, Bricarello, Rocha and Amarante2008).
Grass cutting to simulate grazing
In week 6, the pasture was cut in all the experimental units where it had not yet been collected, corresponding to week 8. A 10 cm height was maintained, separated only by distance. Height data at the time of cutting were collected using the same method as that of the regular collections.
Climate variables
Daily weather data, maximum, average, and minimum temperature, average relative humidity, rainfall and solar radiation, were obtained from the database of INMET, Brazil's National Institute of Meteorology.
During the experimental period, the highest and lowest average temperatures were 32.8 and 15°C, respectively. The lowest average temperatures were recorded during week zero (0) and week five (5), with intervals of 19.68 to 22.36°C, and 20.26 to 24.56°C, respectively. The highest average temperatures were recorded in weeks 3 and 7, with intervals of 21.34 to 27.72°C and 25.3 to 27.96°C, respectively. Solar radiation varied from 8.22 to 54.81 cal/cm². The lowest mean radiation, 23.63 cal/cm², occurred in week 1, and the highest, 46 cal/cm², in week 7 of the experimental period (fig. 2).
Fig. 2. Daily values of rainfall (mm), air relative humidity (%), solar radiation (cal/cm²), minimum temperature(°C) and maximum temperature (°C) of the experimental period.
Total rainfall in this period amounted to 256.4 mm, and there was no rainfall only in week 6. The lowest rainfall, 4.6 mm, occurred in week 8 and the highest, 63.4 mm, in week 0, distributed over the first four days. Relative humidity varied from 59 to 95%. The lowest average relative humidity, 68.7%, was recorded in week 6, and the highest, 81.5%, in week 0 (fig. 2).
Statistical analysis
The data were inspected and described, and then separate multilevel models were fitted for the analysis of the response variables, that is, third-stage larval count in pasture, soil and faeces collected from the plots. All statistical analyses were performed in R software (R Core Team, 2019). The experimental design was developed to evaluate a total of 48 plots, making a total of 192 samples of pasture, 48 samples of soil and 48 samples of faeces. However, 14 samples were lost: eight of pasture; four of faeces; and two of soil. Additionally, at week 8, a complete plot was lost, which corresponded to six observations, and, due to the high frequency of 0 (zeros) in the recovery of larvae (97.27%), this week was excluded from the larvae recovery models to better fit the data.
Multilevel analyses
Multilevel analyses were fitted using the lme4 package (Bates et al., Reference Bates, Mächler, Bolker and Walker2015). Explanatory variables were evaluated in univariate models, and then significant variables (P < 0.2) were included in multivariate models. The models were manually reduced. Variables not associated (P > 0.05) with the response variable were removed from the models. The final multivariate models were simulated 2000 times using the arm package (Gelman, Reference Gelman2018). The coefficients of the fixed effects were extracted from the simulations and then exponentiated to report the results as rate ratios and 95% credible intervals. P-values for fixed effects were obtained using the Wald type 2 Chi-square test. Goodness-of-fit of the models was assessed using the DHARMa package (Hartig, Reference Hartig2017). All models fit the data, showed no overdispersion, and random effects were normally distributed.
a) Larval count in pasture
Multilevel negative binomial regression was fitted using stage-3 larval count as the response variable and the following variables as explanatory variables:
Fixed effects:
-
Variable 1: Interaction between legume forage species (bird's-foot-trefoil, white clover and red clover) with the plant's vertical stratum (upper, lower)
-
Variable 2: Time (weeks 1, 2 and 4 of assessment of larvae recovery)
Random effects:
-
Block (n = 4) and plot (n = 48) nested within each block
b) Larval count in soil
Multilevel Poisson regression was fitted using stage-3 larval count as the response variable and the following variables as explanatory variables:
Fixed effect:
-
Variable 1: Time (weeks 1, 2 and 4 of assessment of larvae recovery)
Random effects:
-
Block (n = 4) and plot (n = 35) nested within each block.
c) Larval count in faeces
Multilevel Poisson regression was fitted using stage-3 larval count as the response variable and the following variables as explanatory variables:
Fixed effect:
-
Variable 1: Time (weeks 1, 2 and 4 of assessment of larvae recovery)
Random effects:
-
Block (n = 4) and plot (n = 34)
Results
L3 count as a function of the time factor
The time factor significantly affected the recovery of L3 from pasture, as described in detail in table 2, and the pattern of larvae recovery over time remained independent of the distance from the faecal bolus (ten and 30 cm in diameter), the plant stratum (upper stratum and lower stratum) and the forage species. In week 2, L3 recovery from pasture was 2.3 times higher than in week 1. However, in week 4, the L3 recovery decreased to levels similar to those found in week 1.
Table 2. Results of a negative binomial regression for pasture and a multilevel Poisson regression for faeces and soil data, comparing the survival of infective larvae of gastrointestinal nematodes in weeks one, two and four after contamination with faeces.
The faeces were deposited on 24 October 2018 on plots composed of monocultures of white clover, red clover and bird's-foot-trefoil. For pasture plot: block: 36, block: 4; for faeces plot: 34, block: 4; and for soil plot: block: 35, block: 4. For pastures, n is the number of sub-plots (height and strata) evaluated in each treatment, while for faeces and soil, n is the number of plots evaluated. The experiment was designed to evaluate 48 sub-plots per week in the case of pastures, and 12 plots per week in the case of faeces and soil. Lower numbers of plots or sub-plots are the result of missed observations. L3: total count of infective larvae in the samples; ICC: interclass correlation coefficient; and CrI: credibility interval of 95%. In tables 2 and 3, the indices for the item pasture belong to the same negative binomial regression.
Table 3. Results of a negative binomial regression comparing the migration of infective larvae of gastrointestinal nematodes in the upper stratum of bird's-foot-trefoil, white clover and red clover four weeks after the deposition of sheep faeces contaminated with gastrointestinal nematode eggs.
The faeces were deposited on 24 October 2018 on plots composed of monocultures of white clover, red clover and bird's-foot-trefoil. Plot: block: 36, block: 4. n is the number of sub-plots evaluated in the upper stratum treatment. In the case of pastures, the experiment was designed to evaluate 24 sub-plots; lower numbers of plots or sub-plots are the result of missed observations. L3: total count of infective larvae in the samples; ICC: interclass correlation coefficient; and CrI: credibility interval of 95%. The indices in Table 3 and the data on pasture in table 2 are part of the same negative binomial regression.
In soil, the time factor also influenced the L3 recovery rate (table 2). The highest L3 recovery rate was recorded in week 1. In week 2, the recovery rate was 0.2 that of week 1, and in week 4 the recovery rate remained 0.17 that of week 1.
Larvae recovery from faeces was also influenced by the time factor (table 2), and the number of L3 decreased from week 1 to week 4 by 0.05 times after faecal deposition.
L3 count as a function of the factors of forage Species, plant Stratum and distance from faecal pellets
The L3 recovery was significantly influenced by the forage species and the vertical stratum of the pasture, these two factors being interactive (table 3). L3 recovery was, on average, 9.8 times greater in the upper stratum of white clover than in the upper stratum of bird's-foot-trefoil, reaching counts of up to 27.4 times more larvae. For the upper stratum of red clover, L3 recovery was on average 5.7 times higher, reaching up to 18.1 times more than in the bird's-foot-trefoil in the upper stratum. L3 recovery from soil and faeces was unaffected by the factors of forage species and plant stratum. The factor of distance from the faecal pellets did not affect the response variables of pasture, soil and faeces.
Average height of forage species
The average height of bird's-foot-trefoil was 13.59 cm in week 1, 20.03 cm in week 2, 31.38 cm in week 4 and 15.25 cm in week 8. White clover reached average heights of 15.97 cm, 17.19 cm, 22.25 cm and 10.69 cm in weeks 1, 2, 4 and 8, respectively. The heights reached by red clover in the above-mentioned weeks were 20.22 cm, 27.19 cm, 31.59 cm and 10.54 cm. The corresponding heights of the upper and lower strata of each legume species over time are described in table 4.
Table 4. Total average height and average heights corresponding to the upper and lower strata of each forage legume in the weeks of collection.
Heights were collected at weeks 1, 2 and 4 after contamination of the plots, at which time all experimental units were cut to a height of 10 cm.
LSH: average lower stratum height, length from ground to mid-plant.
USH: average upper stratum height, length from mid-plant to top.
The height of the forage legumes was greatest in weeks 2 and 4 and lowest in week 8 (after cutting to simulate grazing). The time factor had a significant effect on the legume forage species (P < 0.01), and the bird's-foot-trefoil reached a greater height than white clover in weeks 2, 4 and 8, with mean differences of 5.21 cm, 11.5 cm and 7 cm, respectively. Red clover reached a greater height than white clover in week 2, with an average difference of 5.75 cm.
Dry weight of faeces
On the day the plots were contaminated, the average dry weight of faeces was 6.1 g (±0.1). During the experiment, the average dry weights of faeces in weeks 1, 2, 4 and 8 were 4.32 g (±0.85), 4.29 g (±1.0), 2.97 g (±0.96) and 3.02 g (±1.80), respectively.
The dry weight of faeces was influenced by the time factor (P < 0.01), and the weight in week 8 differed from week 1, decreasing by 1.65 g. The forage species did not show a detectable influence on this variable.
Discussion
Influence of different forage legumes and the time factor on the survival of larvae in pasture, faeces and soil
The forage species had no influence on the survival of L3 in the pasture, in the faeces and in the soil, throughout the evaluated weeks. Although the architecture, structure, leaf shape and type of growth of the studied forage species are different, these differences did not lead to variations in the microclimate under their canopy capable of differently affecting the survival of the larvae. Knapp-Lawitzke et al. (Reference Knapp-Lawitzke, Küchenmeister, Küchenmeister and Von Samson-Himmelstjerna2014) found that the recovery of L3 from pastures containing mixed forage legumes (L. corniculatus L., var. Bull; Medicago lupulina L., var. Ekola; and Trifolium repens L., var. Rivendel) was easier than from pastures composed solely of grasses (Dactylis glomerata L., var. Donata; and Lolium perenne L., var. Signum). The authors attributed this finding to the structure of forage legumes, which retain more moisture, diminish direct sunlight and reduce climatic oscillation, thereby providing a ‘refuge’ for free living parasites. In this sense, it is possible that the three legumes provided favourable conditions for the development of eggs into L3, since L3 were observed in the pasture one week after the deposition of faeces. On the other hand, some authors have found that larval survival is not influenced by the type of forage (Marley et al., Reference Marley, Cook, Barrett and Keatinge2006a; Oliveira et al., Reference Oliveira, Costa, Rodella and Silva2009), which strengthens the idea that macroclimatic conditions have a greater impact on L3 survival. In this sense, the spring weather conditions in a coastal area in Florianópolis/SC (high humidity and temperatures) could explain the results found in the present study.
Regarding the time factor, the number of L3 in faeces decreased progressively over the weeks, and in week 4 73% of the evaluated faeces no longer showed L3, demonstrating that they migrated to pasture or died by desiccation. In Lages/SC in southern Brazil, no eggs were found in sheep faeces 10 days after deposition and larvae were found in faeces up to 55 days after deposition during spring (Souza et al., Reference Souza, Bellato, Sartor and Ramos2000), demonstrating a much longer period of permanence in the faecal bolus. It is possible that in the coastal climate of Florianópolis/SC the eggs hatched and migrated quickly; however survival was short, probably as a consequence of greater energy expenditure and greater contact with solar radiation on top of the forages (Crofton, Reference Crofton1948; Santos et al., Reference Santos, Silva and Amarante2012; Leathwick, Reference Leathwick2013; Gasparina et al., Reference Gasparina, Baby, Fonseca and Bricarello2021).
In pasture, the peak of L3 occurred in week 2 of collection, but in week 1 the L3 were already found in the pasture and in the soil. Solar radiation has been shown to be determinant in the migration of free-living stages in the same region (Bricarello et al., Reference Bricarello, Costa, Longo, Seugling, Basseto and Amarante2022). It is probable that episodes of rainfall greater than 15 mm on the third, fourth, ninth and twelfth days after deposition of faeces favoured the exit of L3 from the faecal bolus, thus rapidly and progressively decreasing the number of larvae in the faeces. These results coincide with studies by Van Dijk & Morgan (Reference Van Dijk and Morgan2011) and Wang et al. (Reference Wang, Van Wyk, Morrison and Morgan2014) in which these authors found that faeces that remain moist and rainfall of 20 mm/day favour the rapid migration of L3 out of the faecal bolus. In week 4, the number of L3 recovered from pasture was very low, and from week 6 this number was even lower, indicating that after 30 days of contamination, L3 survival in the pasture was minimal. Possibly, the brief survival of larvae in the pasture was caused by the climatic conditions of spring on the southern coast of Brazil, where high humidity, average maximum temperature/day of 27.29°C ± 2.79 and average solar radiation of 1477 Kj/m2 ± 575.76 may have contributed to the rapid death of L3 by desiccation. Ramos et al. (Reference Ramos, Pfuetzenreiter, Costa and Dalagnol1993) also found that L3 survival in pasture was shorter in spring (less than 60 days) than in other seasons, when larvae normally survived for between 100 and 120 days. Under high temperature and high humidity, L3 mainly migrate vertically towards the tip of the canopy (Silva et al., Reference Silva, Amarante, Kadri and Carrijo-Mauad2010) and the greatest survival occurs in conditions of medium temperatures, low radiation, low precipitation and high humidity (Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2007).
In the soil, L3 peaked in week 1, gradually decreasing over time. However, the number of L3 found in the soil was very low, which may mean that their migration into the soil was hindered in some way. Ramos et al. (Reference Ramos, Pfuetzenreiter, Costa and Dalagnol1993) suggest that an increase in the number of L3 in soil can be ascribed to adverse weather conditions, with low moisture levels. These authors also reported that fewer L3 in soil was associated with increased rainfall. The results suggest that the climatic conditions of the present study were favourable for the L3, reducing their need for refuge in the soil. In the current work, only larvae in the top two centimetres of soil were counted. The number of L3 in the soil were probably higher than those recovered, since larvae can migrate deeper into the ground. Sturrock (Reference Sturrock1965) reported that even in the presence of heavy rain, 85% of L3 in soil were found at a 5 cm depth, while only 0.8% reached depths of 25 cm to 30 cm.
It is important to point out that the low contamination of faeces by nematode eggs with which the plots were contaminated, in combination with the low percentage of L3 recovery in the pasture due to the high mortality rate throughout the development stages (Silangwa & Todd, Reference Silangwa and Todd1964; Niezen et al., Reference Niezen, Charleston, Hodgson and Miller1998a), may have influenced the low total recovery and rapid decrease in L3 in the present study. The Agroecology Research and Extension Center maintains a flock certified as organic, with sheep from crosses with rustic breeds such as Crioula Lanada, where most individuals have low faecal oocyst counts (Pereira et al., Reference Pereira, Longo, Castilho and Leme2020) and the use of anthelmintics is very reduced and used selectively when needed in growing lambs and in peripartum ewes. It is possible that under conditions of high parasite burden, or greater shedding of eggs in the faeces, as in young or immunosuppressed animals, the reduction in L3 contamination in the pasture would have been slower (Silangwa & Todd, Reference Silangwa and Todd1964; Niezen et al., Reference Niezen, Charleston, Hodgson and Miller1998a).
Influence of forage species on larval migration in the pasture
Regarding horizontal migration, there was no difference in the number of L3 recovered in the diameter from 0 to 10 cm and in the diameter from 10 to 30 cm. This indicates that the three legumes provided favourable conditions for horizontal migration. Several studies have already provided evidence that the horizontal movement of the larvae is minimal, and 89% of the L3 are found in the diameter from 0 to 15 cm of the faeces (Crofton, Reference Crofton1948; Almeida et al., Reference Almeida, Castro, Silva and Fonseca2005).
The legume forage species played a decisive role in the vertical migration of larvae. The largest number of L3 was observed in the upper portion of the red clover and white clover, indicating that the bird's-foot-trefoil hindered the vertical migration of L3 and/or prevented them from remaining in the plant's upper stratum. Although no studies were found comparing L3 migration in these three forage legumes, when compared to other forage species, a higher contamination rate has been reported in clover species (Niezen et al., Reference Niezen, Charleston, Hodgson and Miller1998a; Knapp-Lawitzke et al., Reference Knapp-Lawitzke, Küchenmeister, Küchenmeister and Von Samson-Himmelstjerna2014) and a lower contamination rate in bird's-foot-trefoil (Marley et al., Reference Marley, Cook, Barrett and Keatinge2006a), corroborating the current findings. In the current study, no differences were found between the numbers of L3 recovered in white clover and red clover, which agrees with the results reported by Marley et al. (Reference Marley, Fraser, Roberts and Fychan2006b). These authors found a smaller number of L3 in the upper stratum of the pasture compared to Lolium perenne (perennial ryegrass), although they detected no differences in strata between the two clover species, which agrees with the findings obtained in the present research.
Analysing the morphological differences between the studied forage legumes, the small leaves of the bird's-foot-trefoil, in comparison with the larger and umbrella-shaped leaves of the clovers (Carvalho et al., Reference Carvalho, Santos, Gonçalves, Moraes, D and J2010), may have caused greater exposure of the stem to solar radiation (Scheffer-Basso et al., Reference Scheffer-Basso, Vendruscolo and Cecchetti2005; Van Dijk et al., Reference Van Dijk, Louw, Kalis and Morgan2009), retaining less moisture and causing greater oscillation in the temperature below the canopy (Amaradasa et al., Reference Amaradasa, Lane and Manage2010; Knapp-Lawitzke et al., Reference Knapp-Lawitzke, Küchenmeister, Küchenmeister and Von Samson-Himmelstjerna2014), which limits larval migration to the upper stratum of the forage. In contrast to other authors (Niezen et al., Reference Niezen, Charleston, Hodgson and Miller1998a; Nieto et al., Reference Nieto, Martins, Macedo and Zundt2003; Marley et al., Reference Marley, Fraser, Roberts and Fychan2006b), the type of stem or the presence of trichomes does not seem to have been decisive, since the two legumes that favoured the migration of the L3 to the upper stratum have distinct stems, that is, red clover has a cespitose stem and white clover has a stoloniferous stem. In addition, there was no difference in the number of L3 in the upper stratum between white clover, which is glabrous, and red clover, which has trichomes.
Over time, legume species present different heights, which corresponds to their biology (Carvalho et al., Reference Carvalho, Santos, Gonçalves, Moraes, D and J2010); for this reason, the height variable was tested for the number of L3 found in the pasture, in the faeces and in the soil, although no influence of this variable on L3 recovery was found. The fact that all the species had the same height at the time of contamination might have been decisive for this result. Studies have shown that, regardless of the forage, the lower height at the time of deposition of faeces influences the development and migration capacity of larvae (Silangwa & Todd, Reference Silangwa and Todd1964; Rocha et al., Reference Rocha, Bricarello, Rocha and Amarante2008).
Finally, it is important to consider that plant morphology can be altered according to the type of management and animal stocking to which the pastures are subjected, that is, grazing on white clover can decrease leaf size and thicken stolons (Sharp et al., Reference Sharp, Edwards and Jeger2012). Additionally, Rocha et al. (Reference Rocha, Bricarello, Rocha and Amarante2007) suggest that the morphological characteristics of plants can both facilitate and/or limit the vertical migration of larvae when interacting with the climatic conditions in the different seasons of the year. Ultimately, these factors could modify the results of the present study, which means it is, therefore, important to study the dynamics of L3 in these forage species under different climatic conditions and types of management.
Final remarks and conclusions
The findings of this study indicate that when compared to white clover and red clover, bird's-foot-trefoil inhibits the migration of L3 to the plant's upper stratum, lowering the risk of infection by GIN in grazing sheep, as they are known to prioritize leaves from the apex of the plant. Pasture trials that study the dynamics of survival and migration of free-living stages of GIN in different forages are important to contribute to the development of auxiliary tools in the integrated control of parasites in grazing sheep.
Conflicts of interest
None.
Financial support
The authors gratefully acknowledge the financial support received from the Coordination for the Improvement of Higher Education Personnel, Brazil (CAPES), Finance Code 001 and for the master's scholarship (M. Garcia-Méndez).
Ethical statement
The project was approved by the Ethics Committee on Animal Use of the Federal University of Santa Catarina (CEUA/UFSC) under Protocol No. 7204130918.
