Forage potential of three wild species of genus Lupinus (Leguminosae) from Mexico

Legume crops are a significant component of the human diet and animal feed due to high crude protein content and also represent an economical source of digestible fiber and minerals (Jensen et al., 2010).
       
In Mexico, the cultivation of forage legumes has been limited, likely due to lack of knowledge and inadequate economic incentives. Only the lucerne (Medicago sativum) and chickpea (Cicer arietinum L) are sown in a considerable area, but sometimes these species have problems in adapting to certain environmental and management conditions. Therefore, there is a crucial need to increase the production of forage legumes.
       
An important alternative to increase the diversity of forage legumes may be incorporation into the cultivation of native species with forage potential, which are well adapted to local conditions. Among the most promising legumes, as alternative sources and economic fodder for feeding ruminants are some Lupinus species (Fabaceae). This family comprises 200 to 500 species (Gladstones, 1974); mostly grown in America, from Alaska to Argentina and only 13 species grow in the highlands of Africa and the Mediterranean basin (Ainouche and Bayer, 1999). Although these species are generally characterized by a high protein content in both forage and grain, only information is available on the forage of cultivated species, such as Lupinus albus, Lupinus angustifolius and Lupinus mutabilis (Cernay et al., 2016; McKenzie and Spaner 1998; Zamora, et al., 2017).
       
Mexico has 110 wild species distributed in several localities, but few studies have examined their forage potential. Ruiz et al., (2006) carried out a preliminary study of the chemical composition of Lupinus exaltatus forage, whereas Herrera-Velazco et al., (2010) reported that Lupinus rotundiflorus and L. exaltatus green forage can be preserved by silage. In this regard, it is necessary to study the adaptation of these species under cultivation conditions in terms of forage production and quality with the purpose of making adequate use of forage in feeding ruminants. Therefore, the objective of this research was to evaluate the yield and nutritional quality of forage in three wild lupin species (L. exaltatus, L. mexicanus, and L. rotundiflorus) harvested at different maturity stages under irrigation conditions.

The experiments were conducted from November 2016 to March 2017 at the Agricultural Experimental Farm and Nutritional Animal Laboratory located in the Centro Universitario de Ciencias Bioìlogicas y Agropecuarias (CUCBA) of the Universidad de Guadalajara in Zapopan, Jalisco, Mexico. This site is located at latitude 20°43’ N, longitude 103°23’ W, 1560 m above mean sea level (Fig 1). The climate is temperate and humid, with summer rains. Rainfall varies from 700 to 1400 mm per year, with an average annual temperature of 12.0 to 18.0°C, with frosts. The field experiment was established in sandy loam (SL) soil (classified as Regosol), with pH 5.1, 1.7% organic matter content, 0.08% total nitrogen content and 68.5 mg kg^-1 phosphorus. The Ca, Mg, K and Na content was 1.36, 0.18, 0.83 and 0.35 cmol kg^-1, respectively. During the field experimental period, we recorded only the temperature (Fig 2) because rain occurs very infrequently during this time of the year under irrigation conditions.




       
Seeds were collected from natural populations of L. exaltatus, L. mexicanus and L. rotundiflorus. The seeds were subject to a scarification process before sowing, consisting of a 30 min soaking in concentrated sulfuric acid. After the scarification process, the seeds of three species were sown manually in rows in prepared soil without inoculants on 6 November 2016. The experiment was laid out as a randomized complete block design with factorial arrangement (three species and three harvest dates) and three replicates. Each experimental unit consisted of five rows 3 m long. Plant to plant and row to row distance were maintained at 10 cm and 70 cm, respectively. The estimated sowing density for each species was 12 plants/m². Irrigation was applied via subsurface dripping for 15 days during the trial, starting 2 days after the sowing and ending when 80% of the pods in the main stem of the species with the longest time to reach maturity were ripped. Weed control was performed by hand just before the flowering stages. Plots were inspected regularly, but disease control was not necessary.
       
The wild lupin were harvested on three harvest to different growth stages as described in Table 1: i.e. 26 February 2017, 3 March 2017 and 28 March 2017, respectively. The age of the plants was of 120, 135 and 150 days after sowing (DAS).


       
At each harvest, forage yields were determined by harvesting a 2.0 x 0.8 m area in the center of the plots. All plants were cut approximately 3 cm above the level of the soil with pruning scissors. The fresh whole forage plants were dried at 70°C in a forced draught oven for 48 h and weighed to determine the forage yield, expressed as dry matter (DM). In addition, forage samples of each species were ground using a Wiley mill to a particle size of 1 mm for subsequent analysis.
       
Kjeldhal-N was determined and crude protein (CP) content calculated by multiplying N by 6.25 (Khalil, 1994). The other constituents, ether extract (EE), crude fiber (CF), and ash (A), were estimated by the methods of the A.O.A.C. (1995). Acid detergent fiber (ADF), neutral detergent fiber (NDF) and lignin were analyzed for quality using the procedure published by Goering and Van Soest (1970).
       
For the mineral content analysis, an acid digest was prepared by oxidizing each sub-sample with a 2:1 mixture of nitric/perchloric acid (2:1). Aliquots were used to estimate K by flame photometry, P by spectrophotometric methods (Khalil and Manan 1990), and Ca, Mg, Mn, Fe, Cu and Zn by atomic absorption spectrophotometry (A.O.A.C., 1995). The analysis of significant treatment effects and interactions (P<0.05) was performed by analysis of variance (ANOVA) using the Analysis Statistical System (SAS Inst., 1985). When significant treatment effects occurred, means were separated using the Tukey test (P<0.05).

Forage yield
 
The effects of species and harvest date on the forage yield (expressed as DM) were significant (P<0.05), but no significant interactions were found between species and harvest dates. The forage yields for all three lupin species at different harvest dates are presented in Fig 3.
        
In general, the DM gradually increased with advancing maturity in all three lupin species. Higher yields are a natural phenomenon as structural members increase and new tissues are formed with the progression of ripening. A similar trend has already been reported in other forage legume species, including Lupinus species (Romero et al., 1993; Prusiński, 2014; Kitessa, 1992). L. exaltatus, the high yielding species, always maintained the highest DM at all three growth stages, followed by L. mexicanus species. L. rotundiflorus, the low yielding species, maintained the lowest DM over its growth period. This can be explained by it being the first species to reach maturity (early maturity), followed by L. mexicanus, and then L. exaltatus. Only L. exaltatus had forage yields similar to those found in different genetically improved varieties of L. angustifolius, L. albus, L. mutabilis and L. luteus cultivated in Lithuania, Chile, and Poland (Maknickienë and Ražukas, 2007; Romero et al., 1993; Prusiński, 2014).
        
However, in New Zealand, Kang et al., (2008), reported dry forage yields higher than those found in this study with L. exaltatus, which presented the best behavior for forage production. Although forage yields in L. mexicanus and L. rotundiflorus were generally lower than those recorded in other cultivated legume species, with appropriate management practices, such as seed inoculation, fertilization, date sowing, and plant density, it should be possible to increase the yields. Significant differences were found among lupin species and date harvests in CP, EE, CF, Ash, ADF, NDF and lignin content (Table 2). The two most influential factors that affect forage quality and utilization are the forage species and forage maturity (Arthington and Brow, 2005). On the other hand, the CP content of forage has been reported to be one of the most important criteria for forage quality evaluation (Haj-Ayed et al., 2000). In this study, the CP content ranged from 14.0 to 17.8% in L. exaltatus, from 12.0 to 15.2 % in L. mexicanus, and from 14.6 to 19.8% in L. rotundiflorus, independent of the harvest date.


 
Nutritive value
 
The highest CP concentration was obtained in the first harvest for three species, whereas the lowest was found on the third harvest date. The CP content decreased with advancing maturity, but with different values over the three species, which has already been reported in other forage species harvested at different phenological stages (Kaplan, et al., 2014; Borreani, et al., 2007). The rate of decrease in the CP was more pronounced in L. mexicanus, with values of 19.8% on the first harvest and 14.6% on the last harvest. The average protein content found in our study was below the values reported in lucerne (Medicago sativa), which is considered the most important forage species in the world. However, with respect to other annual legumes, the CP values in the present study are higher than those reported in soybean varieties harvested at the full seed stage (Kökten, 2014) but similar or lower than those obtained in different bitter vetch (Vicia ervilia L.) lines harvested at flower-set, full-bloom and pod set periods of maturity (Kaplan, et al., 2014). Compared to other lupin species, the protein contents measured in our species were similar or slightly less than those reported in L. angustifolius, L. albus, L. mutabilis, and L. Lupinus polyphyllus x Lupinus arboreus harvested at different stages (Romero et al., 1993; Zamora et al., 2017; Kitessa, 1992).
       
Other important quality characteristics for forage are NDF and ADF (Caballero et al., 1995), which should be low in quality forage because they obstruct the digestibility and consequently decrease the quality of forage (Kaplan, et al., 2014). In our study, the NDF and ADF content ranged from 34.3 to 51.2% and 20.1 to 33% respectively, depending on the species and harvest date. NDF and ADF in the present study were within the ranges reported in other annual legumes used as forage, such as Vicia ervillia L., Pisum sativum, Vicia sativa, and Lupinus angustifolius (Kaplan, et al., 2014; Borreani, et al., 2007; Reboléet_al2004; Faligowska et al., 2014). The NDF of the whole crop increased over the growth cycle; as the plant grows, the need for structural tissue increases. Although an increase in NDF and ADF was observed with the later harvest, in general all of the species tested here had good fiber quality. On the other hand, in all three species studied, the lignin content of the whole crop also increased over the growth cycle (harvest date). Data from this study support the findings of previous studies with different forage legumes and no legumes. The highest average lignin 5.4 was measured in L. exaltatus, whereas the lowest lignin 2.8 was measured in L. rotundiflorus. With respect to mineral content, the stage of plant maturity generally affected the content of a number of minerals in lupin forage. Among the three lupin species, the forage of L. rotundiflorus had the highest concentrations of P, K and Fe, whereas L. exalatus had the lowest concentrations of P, K Ca and Mg (Table 3). Zn and Mg were found in appreciable concentrations in L. exaltatus and L. rotundiflorus. Rapid uptake of P, K and Mg was usually observed during the early stages of growth. P, K and Mg markedly declined with increasing maturity, whereas Ca, Mn and Zn were not altered by the stage of maturity. The results indicated that only Ca and Zn concentrations did not change appreciably with advancing maturity of the forage in any of the three species. Although there was significant variation in the mineral content among species and maturation stages, the concentrations of some of the minerals analyzed in the present study were comparable to those measured in other annual or perennial forage species, such as medicago sativa, Vicia villosa and Vigna unguiculata (Marković, et al., 2009; Lanyasunya et al., 2008).

 The forage yields (DM) in all three species were generally greater on the third harvest date, but in the CP content was lower. The species with the highest yield of forage was L. exaltatus, followed by L. mexicanus and L. rotundiflorus, but the species with the highest protein content on average was L. rotundiflorus.