Geographical location of the study area

The experiment was established in the experimental field of the Regional University Unit of Arid Zones of the Autonomous University of Chapingo, in Bermejillo, Durango, Mexico, located at the coordinates 25.8° N and 103.6° LW, at an altitude of 1,130 m asl. The region has a dry desert climate with summer rains and cool winters, an average annual rainfall of 258 mm, an average annual evaporation of 2,000 mm, and an average annual temperature of 21 °C, with maximum temperatures of 33.7 °C and minimum temperatures of 7.5 °C ¹⁶.

Experimental design and management

A randomized block design in a split plot arrangement with three replications was used. The large plots had two soil moisture contents -optimal OSMC (26 % ± 1.5) and suboptimal SSMC (22 % ± 1.5)- established on the basis of the moisture abatement curve¹⁷, according to the regression equation obtained:

Where: SM%= soil moisture content, and X is the negative energy stress in MPa, considering that the field capacity (FC) and the permanent wilting point (PWP) correspond to an energy stress of 0 and -1.5 MPa, respectively. The FC and PWP were calculated based on the above, which were 26.5 % and 17.5 %, respectively. The small plots containing four accessions and one variety of L. corniculatus from different regions (Table 1).

The experimental unit was one plant in each 20 kg rigid plastic pot 35 cm in diameter and 31.3 cm in height. Each pot was filled with 18 kg of a substrate mixture with a 50:30:20 ratio of soil:compost:sand. The substrate had a sandy loam texture, with a proportion of 52 % sand, 26 % silt, and 22 % clay, and a pH of 8.69, an EC of 10.76 dS m^-1, and a bulk density of 1.46 g cm^-3. A digital ORIA thermometer/hygrometer placed inside the shade net recorded the daily temperature (°C) and relative moisture (%) during the evaluation period.

Irrigation was applied every four days, and the soil moisture contents were measured by gravimetry, for which purpose the weight of the pots in the OSMC was maintained at 23.9 kg, and that of SSMC, at 23.0 kg. An average of 0.6 L of water per irrigation was added to both moisture contents, restoring the OSMC to 27.5 % and the SSMC to 23.5 % as upper limits of soil moisture, and leaving both values to decrease to 24.5 % and 20.5 % as lower limits, respectively. A margin of 3.5 % (20.5 - 17.5) was considered a usable moisture range so that the plant did not reach PWP.

A total of seven fresh material cuts were made: the first one in July 2021 and the last one in May 2022. They previously had an adaptation period of 60 d after transplanting, and a standardization cut was made 45 d before the first cut. For the cuttings, growth periods were considered according to the seasons and the intermediate periods: spring-summer (Sp-Su), summer (Su), summer-autumn (Su-A), autumn (A), winter (W), winter-spring (W-Sp), and spring (Sp). The time interval between cuts was 45 d, except for W, when it was extended to 90 d due to the slow growth of the plant due to the decrease in temperature.

Measured variables

The leaf area index was calculated. For this purpose, the leaf surface area was first determined by randomly selecting 10 complete stems per plant at each cutting date; the leaves were then separated from the stems and spread and photographed on a white paper surface; the photographs were processed with ImageJ for each treatment and repetition according to the experimental design. Subsequently, Equation 1, adapted to the conditions of the experiment, was used to obtain the leaf area index¹⁸.

Where: LA= leaf area of a stem (cm²); NS= number of stems, and TSA= total soil surface area in cm² (pot surface area= 962.11 cm²).

The leaves harvested at each cutting date per treatment were dried in a HAFO^® (model 1600, USA) forced-air oven at 60 °C for 24 h or until attaining a constant weight; the dry material was weighed on a Shimadzu analytical balance (model AY220M), and the dry matter (DM) production for each cutting was determined.

The dry fodder increase rate (DFIR) was estimated by dividing the dry weight of forage harvested by the number of days of growth elapsed from one cutting period to the next, with the following equation:

The leaf/stem ratio (L/S) was obtained from a representative subsample of 10 stems from each treatment, for which purpose the leaf and stem components were separated and placed separately in a HAFO^® (model 1600, USA) forced-air oven at 60 °C for 24 h. Subsequently, the leaf/stem ratio was calculated as the quotient between the leaf dry weight (g DM) and the stem dry weight (g DM).

Data analysis

The Statistical Analysis System software¹⁹ was utilized to process the database, performing an analysis of variance and a Tukey multiple range test (P≤0.05) to identify the effect of each treatment effect. In addition, Excel version 6.0 was used for regression analysis.

Temperature and relative humidity

During the period from June 2021 to May 2022, a mean maximum temperature of 30 °C and mean minimum of 20 °C, as well as a maximum of 46.9 °C and minimum of -4.6 °C, were recorded inside the shade net (Figure 1), with mean and maximum temperatures per day of 16.6 to 40.1, 19.8 to 32.7, 9.8 to 37 and 4.5 to 30 °C during the spring, summer, fall and winter seasons, respectively. The average relative moisture recorded ranged between 44 and 73 %, with a minimum of 5-10 % in the months of May through July , and a maximum of 100 % during the rainy season, in July, August and September, with a regional historical annual average rainfall of 258 mm¹⁶. In order to avoid alterations of the soil moisture content in the pots due to the effect of rain, the experimental area occupied by the pots was covered with plastic during these periods.

Figure 1 Maximum, and minimum temperatures and monthly relative moisture (RM) recorded during the evaluation of various genetic materials of L. corniculatus under shade-mesh conditions from July 2021 to May 2022 in Bermejillo, Dgo. 

Leaf area index

The leaf area index (LAI) was significantly higher (P≤0.05) in accession 255305 at OSMC, with a value of 4.7, followed in importance by accession 255301 and the bird’s-foot trefoil variety, with values of 4.1 and 3.7, respectively. In SSMC, the bird’s-foot trefoil variety stood out with 3.9 and was followed in importance by the other accessions, except for 226792, which registered the lowest value of 2.6 (Table 2). The leaf surface area achieved by a plant during its development defines the capacity of the plant canopy to intercept photosynthetically active radiation, the primary source for the proper development of organs and tissues²⁰. According to these results, the LAI in general was slightly negatively affected by the suboptimal soil moisture condition, although the bird’s-foot trefoil variety showed an above-average performance of the LAI attained under optimal soil moisture conditions.

Dry matter production

The DM production was higher with an optimal soil moisture content, amounting to an average of 98 g plant-1, compared to the suboptimal content (SSMC), which recorded an average of 82 g plant-1 without statistical difference (P≤0.05) between the genetic materials tested in this study; while with an OSMC, accession 255305 had the best response, with 131.8 g plant-1 (Table 2). The above suggests that biomass productivity is directly dependent on the soil moisture content, and all the genetic materials of L. corniculatus are negatively affected to the same degree by a suboptimal soil moisture content. These results differ from those reported in a clover adaptability study in which 12 materials were evaluated under temperate field conditions⁸, where accession 202700 and the bird’s-foot trefoil variety were reportedly the most productive, possibly due to the environmental conditions of temperature, ranging between 5 and 32 °C, which are more favorable for this crop

Dry fodder increase rate

The dry fodder increase rate (DFIR) was consistent with the results shown for DM, with a statistical difference (P≤0.05) in the OSMC plot corresponding to accession 255305 as the most outstanding, with a DFIR of 0.43 g plant^-1 d^-1 and no statistical difference between the rest of the genetic materials evaluated. Whereas the SSMC plot had the lowest values, of 0.26 g plant^-1 d^-1 in average, with no statistical difference between the evaluated accessions and varieties (Table 2). Both the yield and the biomass accumulation of different forage crops develop dynamically²¹ due to the formation of new tissue, which is highly influenced by environmental and management conditions, mainly by the temperature and water availability²².

Leaf/stem ratio

The leaf-to-stem ratio (L/S) was similar in both moisture contents, with average values of 2.3 and 2.2 in OSMC and SSMC, respectively, with statistical difference between genetic materials in both cases. Accession 255301 excelled in OSMC with an L/S of 2.9 and 226792 in SSMC with a value of 2.5 (Table 2). The results suggest that, in this variable, the genetic materials are not affected when going from an optimal soil moisture condition to a suboptimal one, which makes it possible to save water, without significantly affecting this productivity indicator. It is desirable that this value be as high as possible, as it is determined by the leaf component; this organ is the most digestible part of the forage and has the highest protein content -much higher than the other organs of the plant- and therefore has the highest nutritional value²³. The L/S results obtained for accessions 255305, 202700 and 226792 at both soil moisture contents were similar to those obtained in a temperate region of Mexico⁸ where values of 2.4, 1.7 and 2.3, respectively, were reported. Additionally, the values obtained in accession 255301 and bird’s-foot trefoil were higher than those obtained in the aforementioned studies, which reported a ratio of 2.0 and 1.5, compared to the 2.9 and 1.9 obtained in the present study in OSMC. This is relevant, given that the study was carried out in a hot dry climate where, even under a shading mesh, extreme weather events occurred that are regarded as very unfavorable conditions compared to the cold temperate climates from which most of the genetic materials in this study originate.

Seasonal dynamics of production indicators

In general, the seasonal behavior of L. corniculatus exhibited variations in terms of the variables LAI, DM, and DFIR, which were higher during spring, summer and summer-autumn -particularly DM- with a production of 17.5, 11.7, and 17.7 g plant^-1, respectively, and a drastic decrease during the winter season (Figure 2A, 2B and 2C); while the L/S remained stable throughout the evaluation period (Figure 2D).

Figure 2 Seasonal dynamics of: A) Leaf area index (LAI); B) Dry matter (DM); C) Dry fodder increase rate (DFIR), and D) Leaf-to-stem (L/S) ratio of L. corniculatus in different seasons of the year with optimal (OSMC) and suboptimal (SSMC) soil moisture contents, during the June 2021 to May 2022 period 

The LAI showed the highest values in the spring, summer and summer-autumn periods, standing out in the OSMC plot during the summer, and then exhibiting an even behavior between the two soil moisture contents (OSMC and SSMC) in the rest of the year (Figure 2A). A similar behavior was observed for DM (Figure 2B) and DFIR (Figure 2C). L/S showed less variation by soil moisture content during the entire evaluation period (Figure 2C). These results coincide with the seasonal behavior of the temperature, which increases with the beginning of spring and reaches its highest values during the summer, being related to a higher incidence of solar radiation, with the consequent increase in the photosynthetic rate, and then begins to decrease in autumn, due to the beginning of the decrease in temperature²⁴. Higher LAI values translate into higher biomass production²⁵.

The productivity results obtained coincide with those obtained in a temperate region of Mexico²⁶, where the highest yields were obtained in spring and the lowest in autumn; however, they do not coincide with the production obtained in summer, which registered the highest values in the present study. This response behavior suggests that it is related to the higher temperature regime, with an average of 22 °C, which favors the growth and development of L. corniculatus⁸. Although there was a decrease in forage production in the winter period, the DM production increased again in spring, which proves the tolerance of the plants to low temperatures, down to -4 °C, and their ability to recover after the thermal stress²⁷.

Knowledge of L. corniculatus forage accumulation per day and its seasonal influence will allow future estimates of forage yield and persistence during the aforementioned periods of the year and will make it possible to establish different management and utilization strategies under field conditions. In this case, the highest DFIR was obtained in the spring, summer and autumn periods for both soil moisture contents, where OSMC exhibited the highest value, of 0.53 g plant^-1 d^-1.

The response obtained for the L/S variable was most stable between cutting periods in the plots with the two established moisture contents, showing differences only in spring-summer, when the ratio was higher in the OSMC plot (2.8) and in the SSMC plot (2.6), followed by the winter and spring periods. This behavior is similar to that observed in a temperate region²⁶, where the highest L/S values were observed in winter and autumn, followed by spring and summer, with values of 2.4, 2.7, 2.0 and 2.1, respectively. This indicator shows that there are no differences between the genetic materials evaluated for the same phenological stage²⁸. Based on this characteristic, it is possible to implement a sequence of forage utilization in future clover farms for the purpose of improving the production and nutritional quality of the plants^29,30.