Study area

The study was carry out in sheep flocks in Conkal, Yucatán, México (21°04'30.1" N; 89°30'18.4" W). With an altitude of 8 m asl, the region has a warm subhumid climate (Awo), a 26.5 °C average annual temperature, and 900 mm annual average precipitation. Soils are calcareous and shallow, with high rockiness (Lithosols and Rendzinas)¹⁴.

The experimental animals were 22 Pelibuey sheep, divided into two groups of 11 animals (five males and six females) per treatment. Average animal age was 78 d and average weight was 19.2 ± 1.4 kg. The sheep were individually marked and identified. They were managed in accordance with the Official Mexican Standard (NOM-062-ZOO-1999), which follows technical specifications for the production, care and use of experimental animals. Prior to the experiment, the animals were vaccinated with 2.5 ml triple typhoid bacterin and deparasitized with Ivomec^® (0.2 mg per kilo live weight).

Two grazing systems were tested. The silvopastoral system (SSP) covered a 130 x 24 m area and was planted with a mixture of forage species: African star grass (Cynodon nlemfuensis) as a base forage, Leucaena leucocephala established in rows (0.5 m between plants, 3 m between rows); a living fence consisting of Hibiscus rosa-sinensis sown every 0.25 m and interspersed with Gliricidia sepium every 2 m; and a centerline of Guazuma ulmifolia planted at 3 m intervals. The African star grass (C. nlemfuensis) pasture (PE) covered a 130 x 24 m area and contained only this forage species.

The total area of each grazing system (3,120 m²) was divided into eleven paddocks measuring 9 x 22 m each. These were bounded by a mobile electric fence. Before it was grazed, each paddock was homogenized by pruning tree and shrub species to 50 cm height and star grass to 10 cm above ground level. The animals were rotated through the paddocks using a 3-d occupation to 30-d fallow ratio, with an animal load equivalent to 1 AU/ha. Over a five-month period (August-December) the animals were grazed daily from 0700 to 1400 h. When not grazing they were kept in individual pens, fed a commercial balanced feed (1% live weight) and provided free access to water.

Social hierarchy tests

Sheep behavior was evaluated using a list of behaviors with forms of dominance expression^15,16. Dominance tests were done by placing two sheep from the same lot in a test pen after 18 h food restriction. They were then offered 20 g of commercial balanced feed and conflicts allowed to occur between them. For five minutes, the frequency of each conduct in the catalogue was observed and recorded using focal-animal sampling for each sheep¹⁷, and dominance and subordination attitudes documented for each animal. This test was done once a month with all sheep in each group (SSP and PE), and results synthesized in a contingency table of paired tests¹⁸.

Forage selection

Using direct observation¹⁹, records were made of the first 100 bites taken by sheep of forage plants in the paddocks. Observations were made every 15 d over two consecutive days between 0700 and 1200 h in each paddock. The data collected also included time, day and animal identification number. In the SSP system, the bites were classified by the species consumed: C. nlemfuensis; L. leucocephala; G. sepium; G. ulmifolia; and H. rosa-sinensis. In the PE system, the bites were classified as star grass or weeds.

Eggs per gram feces (EPG)

Samples (10 g) of fresh feces were collected every 15 d directly from the rectum of each animal and placed in previously marked polyethylene bags. The manure from each animal was homogenized and processed individually to quantify gastrointestinal nematode egg counts per gram of feces using the McMaster technique²⁰.

Hematocrit quantification (HT)

A blood sample (3 ml) was taken directly from the jugular vein of each animal every fifteen days. Each sample was placed in a previously marked test tube containing disodium EDTA and processed with the capillary microhematocrit technique²¹.

Group social hierarchy analysis

Group hierarchy linearity was estimated with the Landau hierarchy¹⁸, which allows calculation of the degree of stratification in a lot using the linearity formula:

Where:

h= linearity index,

n= number of animals in group,

Va= number of animals dominating each individual.

Aggressiveness, dominance and movement efficiency were estimated using equations applied in grazing ruminants²². Values in these indicators range from 0 to 1, with 1 representing absolute linearity, maximum aggressiveness, absolute dominance and maximum movement efficiency.

Aggressiveness was estimated with the equation:

Where:

Ag= aggressiveness index,

Agi= aggressions initiated by individual,

Agt= total aggressions participated in.

Dominance was calculated with:

Where:

DI= dominance index,

D= intragroup dominance index,

Exr= relative movement success.

The intragroup dominance index (D) was calculated using:

Where:

Do= individuals dominated,

Don= individuals not dominated

Relative movement success (Exr) was calculated with:

Where:

Dz= number of times the individual moved,

Dzo= number of times the individual was moved.

Movement efficiency was calculated with the formula:

Where:

Efdz= individual movement efficiency,

Dz= number of times the individual moved,

Ndz= number of times the individual did not move.

Social rank level was estimated based on the dominance index (DI); high= dominated ≥50 % of adversaries; medium = dominated 10 to 49 % of adversaries; and low = dominated <10 % of adversaries.

The variables of dominance, feed preference, hematocrit level, EPG and the interactions between them were analyzed with a mixed linear model for repeated measurements over time, using the MIXED procedure²³. Spearman correlation tests were also run (P< 0.05) to identify the relationships between food preferences, parasite infection levels, hematocrit levels, dominance, movement efficiency and aggression. Data were analyzed with the SAS^® ver. 9.0 statistical package.

Dominance tests

Dominance test results identified a linear hierarchy (1 dominant and 10 subordinates) in the SSP and a bidirectional hierarchy (2 dominants and 9 subordinates) in the PE (Figure 1). In small groups containing animals of the same sex and size, social structure is often linear or nearly linear²⁴.

Figure 1: Social hierarchy in grazing flocks in a silvopastoral system (SSP) and a star grass pasture 

Hierarchy level exhibited only slight linear tendencies in the two systems: SSP, h= 0.75 and r² = 0.9198; PE, h = 0.50 and r² = 0.9822 (Figure 1). A hierarchy is considered linear when its Landau index value surpasses 0.9. This occurs in groups of male animals^17,18, stabled goats, which exhibit clear hierarchical gradation (h= 0.92 and 0.99)²⁵, and lambs, which have a significantly hierarchical social structure²⁶. Buffalo heifers have largely semi-linear hierarchies, with 55.24% unidirectional dominance when in large pastures, and 54 to 63% in small pastures²⁷. Worth noting is that no prior reports exist of hierarchical ranks in mixed groups of sheep under grazing conditions.

The most dominant sheep in both studied groups showed a greater amount of aggressive behaviors (F1, 21= 0.65256, P= 0.000154) involving attacks or threats²⁸. In combination with knowledge of the function of each animal, an understanding of the characteristics of social organization is essential to more efficient management of animal groups and development of optimum production systems²⁹. Social hierarchy in a group of animals is influenced by different factors and defined as inhibition of the behavior of a submissive animal by a dominant animal through threats, butting and other aggression³⁰. A notable effect of being dominant in the present study was a lower parasite load (Figure 2). This coincides with a report of higher EPG values in animals belonging to middle and low (i.e. subordinate) hierarchical categories³¹.

Figure 2: Dominance level () and gastrointestinal nematode egg counts per gram feces () in sheep in silvopastoral (SSP) and star grass monoculture pasture (PE) grazing systems  

Forage selection

Animal forage preference in the SSP was highest for C. nlemfuensis (68 %), followed by H. rosa-sinensis (22 %) and L. leucocephala (10 %)(F2, 11= 15.95349, P= 0.00034); neither G. sepium nor G. ulmifolia were consumed. This same trend in species preference has been reported in empty adult ewes in the same kind of silvopastoral system³². This coincides with reported sheep grazing behavior in that they are intermediate-selectivity consumers that prefer ground grasses, but will occasionally graze trees and bushes³³. Forage selection in sheep is also heavily influenced by social interactions; indeed, these can communicate aversion to certain plants that have caused unpleasant effects in the past³⁴. Of note is that animals which consumed the most H. rosa-sinensis had a lower number of bites of C. nlemfuensis (r_s= -0.763636, P= 0.05) (Figure 3), even though C. nlemfuensis had greater biomass availability in the system³². This was most probably due to the fact that H. rosa-sinensis foliage contains fewer antinutritional compounds^35,36,37, which strongly influence rejection behaviors³⁸.

Figure 3: Forage preference and hematocrit levels in sheep in a silvopastoral system  

The sheep in the SSP that consumed the most L. leucocephala exhibited higher amounts of hematocrit (r_s= 0.694269, P= 0.05). This coincides with a report of hematocrit values higher than 28 in Pelibuey sheep grazed in silvopastoral systems containing L. leucocephala, G. sepium, A. lebbeck and P. maximum grass. It is also similar to the higher hemoglobin and cell volume values reported for Pelibuey ewes and lambs that had consumed a diet supplemented with L. leucocephala or L. pallida foliage^39,40. These are favorable indicators for progeny growth and breeder health⁴¹, and therefore have a positive impact on system productivity and sustainability⁴². Part of this impact may be due to the iron (Fe) content of L. leucocephala (average= 381.30 mg Fe kg^-1 DM)⁴³. Consumption of 94.38 g DM L. leucocephala by sheep provides 200 ppm Fe⁴⁴, an amount higher than the 30 to 50 ppm required by sheep⁴⁵. Adequate Fe intake promotes accelerated growth; increased resistance to infection; absence of anemia (reflected in the hematocrit), lethargy and increased respiratory rate; and decreased mortality rates from Fe deficiency⁴⁶. An additional benefit of L. leucocephala consumption is the reduction in gastrointestinal nematode EPG values in response to its secondary metabolites content (total phenols and saponins)⁴⁷; in fact, it has an inhibitory effect >50% (at 100 mg/ml) on third stage larvae (L3)⁴⁸.

A consequent effect in the present results for the SSP was that the dominant sheep, which had more access to forage, also exhibited greater resistance to parasites and increased hematocrit levels. This agrees with the established knowledge that higher social status individuals tend to have higher productivity⁴⁹.