Evaluation of new pasture legumes options in the Pampa-Patagonia transitional region of Argentina

Esta obra está bajo una Licencia Creative Commons Atribución-NoComercial 4.0 Internacional.

The semiarid-arid Pampa-Patagonia transitional region of Argentina (PTRA) encompasses nearly 18 million hectares. The predominant farming systems combine cereal cropping with cattle production. Annual precipitation (PP) varies between 370 to 590mm, with 67±13% of the precipitation distributed from autumn until spring, but high evapotranspiration (PET) especially during the summer months. During the spring-summer period, the aridity index (AI=PP/PET) is less than 0.2, corresponding to the arid climate type (UNEP 1993). The mean annual temperature is about 14.8ºC with an average minimum temperature during the winter of 7.9°C and a maximum of 22.4°C during summer. As in other semiarid-arid regions of the world, the soils of the PTRA present high risk of eolic erosion, compaction, as well as low fertility (Burkart et al., 1999; Silenzi et al., 2012). Additionally, this region receives low inputs of fertilization and control of weeds, insects, and diseases.

The diversity of native herbaceous legumes is low and perennial grasses dominate the grasslands of the PTRA (Busso, 1997). Naturalised populations of woolly burr medic (Medicago minima L.) show a notorious adaptation in about 10 million hectares (Fresnillo, 1990). However, this species shows fast senescence, low fodder production and very low drought tolerance (Covas, 1978). Moreover, traditional perennial pasture legumes, such as lucerne (Medicago sativa L.), white clover (Trifolium repens L.), and red clover (Trifolium pratense L.) show lack of adaptation to dry conditions and are not used in the PTRA (Iurman et al., 2008).

Scientific information about the benefits of legumes in cereal rotations is extensive. The Australian farming systems, known as ´ley farming´ and ´phase farming´, are based on legume crops. Australian ´ley farming´ practice, which stands for the inclusion of an annual pasture legume in a cereal crop rotation, is a well-known solution to revert environmental and economic problems derived from monoculture practices (Nichols et al., 2007, 2012). Traditionally, the most common species used for this purpose were subterranean clover (Trifolium subterraneum L.) and annual medics (Medicago spp.) (Loi et al., 2005).

The western Mediterranean-Asiatic region is a major area of legume biodiversity which includes a wide range of annual and perennial species, currently grown under semiarid-arid environments. Among the diverse range of adapted pasture legume species, vetches (Vicia spp.) and vetchlings (Lathyrus spp.) are considered the most agronomically important species. The Onobrychis, Hedysarum and Ornithopus genera have great potential as future pasture legumes (Hayot Carbonero et al., 2011; Ates et al., 2013).

Despite all the scientific advances generated for the conditions in Australia and the Mediterranean basin, no studies have yet been conducted in the PTRA. In this region, legume adaptation also presents some difficulties mainly associated with drought and high rainfall variability. Vicia villosa Roth is an annual legume species, well adapted to different semiarid-arid temperate areas of the world. It is considered as a reference legume in the PTRA due to its confirmed adaptation and a large number of references for its use and management (Renzi et al., 2017, 2019).

The current work aims to explore opportunities for developing new pasture legumes to complement V. villosa (Renzi and Cantamutto, 2013). A key requirement for farmers to adopt a new pasture is that it should present an acceptable biomass production during the first year of establishment. The objective of the present work was to assess the suitability of different legume species compared to V. villosa during the first year of establishment in the PTRA.

Field experiment

Legume germplasm was provided by the Gene Bank of the Research Station Hilario Ascasubi (62º37’W, 39º23’S), located in the south of Buenos Aires province. Twenty-seven legume accessions were cultivated in the area of influence of the Research Station from 2015-2019. The selected legumes include 5 species of the genus Medicago, thirteen of Trifolium, and one of Onobrychis. All species showed antecedents of autumnal sowing and use as forage. Most of them are annual species, except for T. fragiferum, T. hybridum, T. medium, T. montanum, T. ochroleucon, T. pannonicum, and O. viciifolia (Kaiser, 2015). Within the evaluated accessions, the cultivars (cv) indicate a greater breeding effort compared to the rest of the accessions (table 1).

A field experiment was performed each year from 2015-2019 with similar sowing dates (10th of April 2015, 27th of April 2016, 11th of April 2017, 22th of March 2018, 5th of April 2019). The accessions evaluated, taxonomic names, common names, references for cultivars and sowing rates are listed in table 1.

The experiment was arranged in a randomized block design (n=3). The experimental units were 4x1.4m. A small plot cone seeder or row manual seeder was used for sowing at 20cm row spacing. During the growing season, observations of flowering time were recorded (2015, 2016, and 2019). Growing degree days (GDD; threshold temperature = 0°C) for the duration between sowing and flowering were estimated. A single cut per year of the forage dry matter (DM) production was measured at the end of winter (mid-September) and in late spring (November) by cutting plant shoots at ground level in 0.16 m² samples. The aboveground biomass harvested was oven-dried at 65°C for 72 hours.

The legume seed was inoculated with a commercially available rhizobial inoculum. Seed germination testing was previously carried out according to ISTA rules (International Seed Testing Association, 2019) to adjust the sowing rates (table 1). Fertilizers were avoided to exclude chemical interactions with bacterial activity. The seedbed was disked lightly and then levelled in the late summer before sowing the plots. Small grains were cultivated before establishing the field experiments with legumes.

The soil surface layer (0-20 cm) of the experimental sites were entic haplustoll, sandy loam, slightly alkaline (pH = 7.9±0.2), with a low organic matter content (1.7±0.5%) and a high P content (15.6±5.9 ppm P Bray & Kurtz). Weather data from each year were gathered from a meteorological station located less than 800 m from the experimental site (http://inta.gob.ar/documentos/informes-meteorologicos).

Statistical analysis

A linear mixed model (LMM) was performed to test if the accessions differed in the winter and spring DM production, and GDD. We considered the accession as fixed factors and the year as a random factor. The blocks were nested within the year. The LMM was performed with R version 3.6.3 using Infostat software (Di Rienzo et al., 2019) interface to R. The coefficient of variation (CV%) for each accession was calculated only in those that were evaluated for more than one year (table 1).

The relationship between spring DM and GDD was plotted using GraphPad Prism Software version 8.0 (GraphPad, San Diego, California, USA).

Table 1

Sample size and serology results.

Analysis of the 3’UTR of SLC11A1 bovine gene

*unhulled seed

** kg of clean, germinable seed ha^-1

***Inoculant group: AM: Sinorhizobium medicae, B and C: Rhizobium leguminosarum bv. trifolii, E: R. leguminosarum bv. viciae, and Special: Rhizobium sp.

Nodulaid

Pampa-Patagonia transitional region of Argentina displays strong climate variability with favourable periods for plant growth during autumn and spring. The annual precipitation from 2015-2017 was similar to the long-term mean (>400 mm), while <400 mm year^-1 was registered in 2018 and 2019 (table 2).

Table 2

Monthly precipitation (mm), average monthly potential evapotranspiration (PET, mm) and temperatures (°C) during the experiments in comparison to the historic average (1966-2018).

Medicago rigidula, Trifolium arvense, T. montanum, T. retusum and T. striatum L. failed to establish (death of seedlings) during 2015. Therefore, they were not evaluated in subsequent years. Similarly, M. minima, T. ochroleucon, T. fragiferum, T. hybridum, T. medium and T. pannonicum showed very low winter DM yield during 2015 (figure 1a), thus they were excluded from further evaluations.

Both winter and spring DM mean values are presented in figure 1a, showing a wide range of oscillation (0 to 2.1 t ha^-1 and 0 to >6 t ha^-1 during winter and spring seasons, respectively). Interaction between accessions per year was observed both in winter (F _46,127 = 4.9; P<0.001) and spring DM (F _46,127 = 12.6; P<0.001). Figure 1b and c show the relationship between the coefficient of variation and winter and spring DM among accessions as measured of stability between years.

V. villosa (cv. Patagonia INTA) had a higher winter DM followed by M. truncatula (cv. Jester and Chetah) and T. subterraneum (cv. Goulburn) (figure 1b). Spring DM production contributed approximately 80% of the total DM per year under PTRA conditions. The best performance in spring DM were V. villosa (cv. Patagonia INTA), O. viciifolia (AD-INTA>ARHA-059), and T. vesiculosum (cv. Zulu II) (figure 1c). However, the latter showed low stability (> CV%) due to reducing the spring DM yield when annual rainfall was < 400 mm (figure 1c).

No significant relationship between flowering time (GDD) and spring DM was observed, although mid-maturing species, as the case of V. villosa, T. vesiculosum, and O. viciifolia (≈ 2000°Cd) showed higher spring DM production, compared to early-maturing (< 2000°Cd) and late-maturing accessions (> 2000°Cd) (figure 2).

Figure 1

Winter and spring dry matter (DM) yield in different legume accessions (a), and relation to the coefficient of variance (%) (b and c). Least significant differences (LSD) (p <0.05) were used to indicate differences among accessions.

Figure 2

Spring dry matter (DM) yields as a function of growing degree days (above 0ºC) necessary to reach flowering. Least significant difference (LSD) (p<0.05) was used to indicate differences among accessions.

This study explored opportunities for developing new pasture legumes to complement the V. villosa (Renzi et al., 2019). M. truncatula and T. subterraneum L. (cv. Goulburn) showed high winter dry matter production, similar to V. villosa, but a poor biomass production in spring. Spring dry matter contributed ≈80% of the total biomass per year. We found evidence that O. viciifolia performs well for spring DM production and moderate stability (CV%) during the establishment (figure 1c). O. viciifolia is a drought tolerant pasture legume with many positive agronomic, environmental, nutritional and nutraceutical attributes, and currently, there has been a resurgence of interest in it (Mora-Ortiz and Smith, 2018). Cultivation as a pasture legume is increasing in semiarid-arid environments, like the western United States (Hybner, 2013; Sintim et al., 2015) and northeast Spain (Delgado et al., 2002; Cirujeda et al., 2019). This situation supports the possibility of its development as a new pasture legume under PTRA conditions.

While in some legume accessions the DM production was very low, these experiments were valid for accession comparison but not for generalizing about the species performance as a whole. However, like in the dryland of West Asia and North Africa, it would seem that vetches would appear to have greater potential use than annual medicines (Ates et al., 2013; Cicek et al., 2020). V. villosa was the best species in DM production due to its breeding and development locally, and O. viciifolia could be a new option for PTRA.

A common outcome in this type of study is the need for earlier accessions for semiarid-arid Mediterranean climates, based on the fact that delaying flowering is counterproductive because it is exposed to terminal drought (Thomson et al., 1997; Berger et al., 2002). Moreover, in Mediterranean climates most rainfall occurs at the same time as the low evaporative demand, improving the rainfall-use efficiency and pasture growth (Cobon et al., 2019). Our results show that mid-maturing accessions present the best performance under PTRA conditions. The mid-maturing accessions would perform better than the early flowering types in low and unpredictable rainfall environments because they show a greater ability to modify growth according to the occurrence of rainfall events. The good adaptation of V. villosa does not correspond to a greater tolerance to drought periods, being higher in O. viciifolia (Schvachsa, 2018). In V. villosa, the shallow root system with good lateral branches as well as the mid-maturity cycle could take advantage of the low and occasional rain events (Renzi and Cantamutto, 2013). V. villosa is well adapted to PTRA farming systems, possibly due to different factors, such as an autumn–spring adjusted growing season, cold hardiness, a long vegetative phase that maximizes productivity, phenotypic plasticity with the ability to modify growth according to the occurrence of rainfall events, and excellent nodulation (Renzi et al., 2017). O. viciifolia shows good behaviour both in the spring DM production and tolerance to drought, with a higher survival rate under the evaluated conditions.

Based on a derivation from current experiments, O. viciifolia is included as a crop for persistence evaluation and on-farm weed control trials. We observed that O. viciifolia with a range of 0.9±0.3 to 5.8±2.0 t DM ha^-1 year^-1, depending on rainfall occurrence (232-to-575 mm year^-1), could persist during more than three years under PTRA (unpublished data). As O. viciifolia has shown adaptation to dryland conditions, it is imperative to evaluate different genotypes for further use in breeding programs for developing cultivars. Also, annual pasture legumes, like V. villosa, could be sown in mixtures with deep-rooted O. viciifolia to improve the forage seasonal distribution, water use, and increase the resilience of pastures to environmental stresses (Dear 2003; Cicek et al., 2020). This study can be useful for a similar semiarid-arid environment considering also the predicted climate change scenarios.

This study explored the dry matter production of 19 selected legume species during the first year of establishment. Besides the well-known hairy vetch, we found evidence to show that O. viciifolia (sainfoin) demonstrated good performance for spring DM production and drought tolerance. This is the first exploratory study of the potential use of sainfoin as a pasture in the PTRA.

The authors declare that they have no competing interests

The data that support the findings of this study are openly available in zenodo at https://doi.org/10.5281/zenodo.5363431.

The study reported in this manuscript was supported by the Federal Institute of Agricultural Technology-INTA (PE-I142, PE-I086, and Convenio MAIBA-INTA). Appreciation is extended to J. Castillo and M. Bruna, agricultural specialists, for assistance in establishing the field experiments.