Study Area Description

The study was performed at Rancho "Salinas", located in the municipality of Satevó, Chihuahua, 27°57' N and 106°07' W. The altitude is 1540 masl, the mean annual temperature is 18.1 °C, and the mean annual precipitation is 450 mm (Graph 1; ^{Medina-García et al., 2006}). The soils in the area are shallow (less than 30 cm) of alluvial origin with a sandy loam texture, regular internal drainage, moderate surface runoff, and a pH of 5.3 to 6.6 (^{COTECOCA, 1978}).

Graph 1 Monthly average precipitation (mm) in Chihuahua state. 

Evaluated Variables

The botanical composition of the area was determined in July (vegetative growth stage) using the dotted line method (^{Herrick et al., 2005}). For species identification of the species present in the area, reference samples were taken (three per species), stems, leaves, and fruit were collected. Then, samples were identified with the support of the herbarium of the Facultad de Zootecnia y Ecología (FZyE-UACH). Obtained weights were used to estimate the percentage of presence of each species during the phenological stage.

Field sampling

Four sampling periods were established during August to February to estimate forage yield, in the different phenological stages of the grasses present in the area. The first sampling was carried out from August 3 to 7 during the growth stage after the first rains, and 15 to 20 days after the first rains in the area. The second sampling corresponding to the development stage was from October 1 to 5. The period for the maturity stage was from December 3 to 7 and for the latency stage, it was from February 1 to 5.

Natal grass morphology

The morphological distribution (leaf, stem, inflorescence, and dead material) of natal grass was determined in 10 plants on each sampling date. Each plant was transported to the laboratory and sorted according to its morphology. The material was then weighed and the percentage for each plant component was determined.

Forage yield (kg DM ha ^-1 ).

In the study area, aerial biomass was measured at 25 randomly distributed points. To estimate the available forage, a 0.25 m2 quadrat was used, the grasses remaining within the quadrat were counted and cut, and samples of each species were placed in individual bags, previously labeled. The collected samples were taken to the Laboratorio de Nutrición Animal of the FZyE and placed in a forced air oven at 65 °C for a period of 72 h for drying and subsequent weighing. A Tor Rey L-EQ scale was used to weigh the samples. The dry weight obtained was extrapolated to DM production ha-1, using the following formula:

DM= (w (g)*4)/1000 10000

Where:

DM = dry matter, w = weight, g = grams, w = weight, g = grams.

Crude protein (CP)

The CP content was determined by the Kjeldahl method (^{AOAC, 2012}).

Acid Detergent Fiber (ADF) and Neutral Detergent Fiber (NDF). These variables were estimated using the method of ^{Van Soest (1963}).

in vitro Digestibility coefficient of Organic Matter (IVDOM) To obtain the data for this variable, the Daysi II analyzer was used, following the methodology of the Ankom® guide (^{Ankom Technology, 2014}).

Standard germination (%)

Petri dishes of 90 mm diameter with cotton and filter paper were used. The experimental unit was the Petri dish with 50 seeds in florets. Four replicates were used and each Petri dish was irrigated with 25 ml of distilled water. The Petri dish were placed in a growth chamber (Precision Scientific, model 6M) at a temperature of 27 ± 2 °C. The germinated seed was considered as that which reached 0.5 cm of aerial part or radicle.

Radicle length (cm)

It was measured on the seventh day after seed germination with the help of a ruler graduated to 1 mm.

Length of aerial part (cm)

It was measured on the seventh day after seed germination with the help of a ruler graduated to 1 mm.

Root:Shoot ratio

It was evaluated on the seventh day with the measurement of the aerial part and radicle. It was observed which quantity corresponded to the radicle in proportion to the aerial part.

Osmotic potential

Three osmotic levels were used to simulate drought and determine the effect of osmotic stress. Before this evaluation, a germination test was carried out with and without a seed coat. For this purpose, 90 mm diameter Petri dishes were used, provided with cotton and filter paper with four replicates for each osmotic level, and 50 seeds (caryopses) were deposited in each one of them. Each Petri dish was irrigated with 25 ml of the corresponding solution. The Petri dishes were placed in a growth chamber at a temperature of 25±2 ºC. The seed was considered germinated when the radicle or aerial part reached approximately 0.5 cm. Water stress treatments were 0.0, -1.0, and -2.0 MPa, and mannitol was used as an osmotic agent. Mannitol concentrations were calculated according to its molecular weight (182.17 g/mol), from the equation proposed by Vant'Hoff (^{Ruiz & Terenti, 2012}):

Ψπ = -CiRT.

Where: Ψπ = osmotic potential, C = concentration of the solution expressed as molarity (moles of solute per kg of water), i = constant for solute ionization (0.00545), R = gas constant (0.00831 kg MPa mol-1 K-1), T = temperature in degrees Kelvin (298.15 °K).

Data analysis Data were analyzed using a completely randomized analysis with the statistical program SAS 9.1.3 (2006).

Botanical Composition

The botanical composition of the area consisted mainly of grasses (Table 1). The natal grass was the species with the highest botanical composition (95.7 %). The fairyduster legume (Calliandra eriophylla) had the highest composition (57 %). Among the herbaceous plants present in the area, the one with the highest botanical composition (61.3 %) was slender dwarf morning-glory (Evolvulus alsinoides). According to the previous results, ^{Fuentes-Ramírez et al. (2010}) report that when there is the presence of an invasive species in ecosystems, the key species of that type of vegetation tend to decrease and there are changes in the structure of the communities and a reduction in the regeneration rate of native species (^{Mata-Balderas et al., 2020}). Obtained data suggested by other authors, it is likely that native species are being displaced, since the natal grass in these areas contributes from 72 to 89 % of aboveground biomass production (^{Lavandera-Barreras et al., 2019}).

The number of grasses present in the study area increased over time. During the first sampling date, only three species of grasses were observed, where natal grass represented more than 51% of the production per species. The natal grass, having a persistent seed bank, could have an advantage over the other grasses present. For the last sampling date, ten species of grasses were observed, however, the natal grass maintained a botanical composition of 45%, which is superior to the other grasses (Graph 2). Most of the grasses present in the study area are native and of livestock importance, important aspects to conserve biodiversity in the grasslands of the Chihuahua state (^{Jurado-Guerra et al., 2021}; ^{Leis et al., 2012}; ^{McGranahan et al., 2013}). The botanical composition that can occupy the natal grass, ^{Miranda (2012}) reported similar values (82 %) in terms of its botanical composition.

Graph 2 Botanic composition in an area invaded by natal grass (Melinis repens) through the phenological stages. 

Forage Yield

When characterizing the natal grass in the study area based on its forage yield, a production of 244 kg DM ha^-1 was achieved during the initial sampling date (Growth phase), with natal grass accounting for 78.6% of the total production (Table 2). On the second and third sampling dates, yields were 1310 and 1444 kg DM ha^-1, respectively. On the second sampling date (Flowering), natal grass accounted for 72 % of the total production and by the third sampling date (Maturity) it increased to 87 %. By the fourth sampling date (Latency) forage yield reached a production of 2118 kg DM ha^-1. This can be attributed to the presence of winter rains, which could have favored the growth of the natal grass since it contributed 88.5 % of the available biomass. ^{Díaz-Romo et al. (2012}) reported biomass production in Melinis repens from 1736 to 2913 kg DM ha^-1 in rainy years, while in dry years it ranged from 707 to 1488 kg DM ha^-1. The above determines that forage production of natal grass can vary depending on rainfall (^{Stokes et al., 2011}). However, it can be assumed that natal grass, despite being an invasive species, is slow-growing (^{Gutiérrez-Gutiérrez et al., 2022}; ^{Melgoza-Castillo et al., 2014}; ^{Morales et al., 2014}) and is related to unfavorable environments (water and nutrient limitations), which favors its establishment in arid areas (^{Mata-González & Meléndez-González, 2005}). In comparison with other native species such as sideoats grama (Bouteloua curtipendula), ^{Beltrán-López et al. (2013}) reported values between 1835 and 2000 kg DM ha^-1 similar to those obtained during the latency stage in this study. However, studies where introduced species were evaluated report yields from 2000 to 2500 kg DM ha^-1 (^{Esqueda & Carrillo, 2001}).

Natal Grass Morphology

When analyzing the structural components of the natal grass, it was observed that the amount of leaves decreased throughout the sampling periods. It was found that leaves represented about 50 % on the first sampling date, however, by latency it decreased to 23 % (Graph 3). Likewise, the number of stems increased considerably during the sampling stages. Likewise, in the growth stage, only 17 % was observed and for the flowering stage, it increased up to 43 %, however, the highest quantity of stems occurred during the latency stage (56 %). During the latency stage, the lowest digestibility is presented due to the Lignin content in the plant, making its palatability lower (^{Gutiérrez-Gutiérrez et al., 2019}). Regarding the amount of inflorescences, a considerable increase was not observed throughout the sampling stages, since it only fluctuated from 10 to 14 % in flowering and latency, respectively. As for dead material, 23 % was obtained in the growing season, which is attributed to the high percentage of material from the previous cycle; however, for the flowering and latency stages, it fluctuated from 3 to 6 %, respectively. Similar studies have been reported by ^{Pérez-Amaro et al. (2004}) with mulatto grass, which found a decrease in the percentage of leaves and an increase in stems and inflorescence during the growth stage. In elephant grass, something similar happened, since there was a decrease in leaf weight and an increase in the number of stems (^{Calzada-Marín et al., 2014}). Regarding the above, ^{Melgoza-Castillo et al. (2014}) mentioned that in the latency stage, the crude protein content of natal grass decreases to 4%, which is perhaps due to the increase in the number of stems and inflorescences during this stage (Calzada-Marín et al., 2014). Another aspect to consider is that when there is a higher stem content, palatability and digestibility by cattle decrease, which is why is possible its permanence in cattle farms (^{Chávez & González, 2008}). However, although the biomass distribution has not been quantified, it is observed that natal grass has a high proportion of stems to leaves, which is why it is considered a poor-quality grass during the last phenological stages (Melgoza-Castillo et al., 2014).

Graph 3 Structural components of rosacea (Melinis repens) during phenological stages. 

Nutritional Value

The crude protein (CP) content of all grasses varies according to their phenological stage. During the growing season, grasses present their highest nutrient levels. In this case, the natal grass presented an average CP value of 11.53 % during the growing season. In a similar study where different grasses were analyzed, Melinis repens was reported to have a CP value of 11.2 % in the growing season (^{Ramírez et al., 2009}). This could be because, during the growing season, the natal grass has the highest leaf content. Other authors mention that the CP content of natal grass is lower than that reported in this study (3-6 %), however, these levels are observed for the maturity stage (^{González-García et al., 2017}; ^{Melgoza-Castillo et al., 2014}; ^{Sánchez-Maldonado et al., 2014}). In another study conducted in Aguascalientes, Mexico, values of 12.6 % CP were found in the growth stage (^{Flores-Ancira et al., 2016}) which are higher than those obtained in this study, unlike the data reported by Sánchez-Maldonado (2014) where he found CP values of 3.58 % in the growing season.

Neutral Detergent Fiber (NDF) contains the cell walls of the plant and is composed of cellulose, hemicellulose, and lignin (^{Holechek et al., 2011}). In this regard, it can be observed in Table 3 how Melinis repens was the species that presented the highest value (71.36 %) of NDF compared to species such as Bouteloua gracilis (69.07 %) and Bouteloua curtipendula (67.78 %). The data obtained by ^{Flores-Ancira et al. (2016}) were similar to those reported in this study, where values of 66.4 % NDF were obtained. ^{Ramírez et al. (2009}) found higher values (74.1 %) than those reported in this study. However, other research carried out in the 80s in the central grasslands of the state of Chihuahua, ^{Chávez et al. (1986}) found NDF values between 72.1 and 76.5 % in native species during the growing season. These values are within those reported for Melinis repens in this study. The Acid Detergent Fiber (ADF) data differ from those reported by Flores-Ancira et al. (2016) where values of 33.2 % are reported, which are lower than those obtained (44.07 %) in this study during the growing season. However, the NDF and ADF values reported by ^{Sánchez-Maldonado (2014}) were similar to those obtained in this research.

Sideoats grama grass had the highest CP content (13.74 %) compared to natal and Blue grama (11.53 and 11.94 %, respectively). In the case of Blue grama, ^{Morales et al. (2009}) reported values of between 12-15 % thus, it can be more consumed by cattle compared to natal grass. Similarly, other authors have mentioned that Bouteloua gracilis presents 8.6 % of CP during the flowering stage (^{Beltrán-López et al., 2013}). In addition, ^{Reyes-Estrada et al. (2014}) analyzed the diet of cattle in pastures in Durango state, where Melinis repens is present, and reported CP values of 10.49 ± 1.20 %.

Osmotic Potential

The results obtained from the germination test for natal-covered grass yielded maximum germination of 30 %, which is high compared to other reports where values between 2 % and 25 % are reported (^{Díaz-Romo, 2012}; ^{Gutiérrez-Gutiérrez et al., 2022}; ^{Stokes et al. 2011}). Only ^{Carrillo-Saucedo et al. (2009}) obtained similar results, 31 %. On the other hand, by eliminating the seed coat and only germinating the caryopses, the germination percentage increased to 76 %. Therefore, only the caryopses were used to evaluate the osmotic potential. Table 4 shows the germination behavior when caryopses were subjected to water stress up to 2.0 MPa. Stokes et al. (2011) reported similar results when studying the biology of natal grass seed; they did not obtain germination at - 2.0 MPa. Another fundamental aspect is high values in the Root:Shoot (R:S) ratio, which represents a greater effort by the species for water acquisition (^{Ervin et al., 2009}; ^{Mastalerczuk & Borawska-Jarmułowicz, 2021}). Due to the above, it can be observed in Table 4 how the natal grass increases its R:S ratio, which would be beneficial for obtaining water during low precipitation events.