With the increasing demand of protein-rich forage, alfalfa (Medicago sativa L.) is widely cultivated and has become a major portion of the forage in diets for dairy cows in southern China, which is characterized by high temperature and rainy climate. Directly ensiling fresh alfalfa is a preferred method to conserve forage in these regions, however, fresh alfalfa silage is more prone to clostridial spoilage due to the low dry matter (DM) in addition to low sugar contents and high buffering capacity [1]. It is necessary to exploit silage additives with antimicrobial activity to ensilage high moisture alfalfa under hot and humid weather.

Attempts to develop additives with antimicrobial activity were made mainly focused on some organic acids, in particularly short-chain fatty acids [2]. Relative to other short-chain fatty acids, propionic acid (PA) used as silage additive has been considered in many studies in terms of its excellent antifungal activity and sparing impact on the activity of lactic acid bacteria (LAB) [3,4]. As an antifungal agent, PA is more effective at low pH (i.e., undissocated form) with the greatest effectiveness at or below its pKa. Inhibition of microbial growth by PA involves rapid diffusion of undissociated molecules through the plasma membrane. The dissociation of these molecules within cells liberates protons, thus acidifying the cytoplasm and preventing microbial growth [5]. Furthermore, PA could affect the nitrogen transformation and reduce the proteolysis during ensiling through acidifying silage mass or restricting activity of undesirable bacteria [6]. However, the application of PA also faces some problems due to its corrosive and hazardous nature, thus its salt derivatives have been developed and are thought to be safer and easier to handler. Calcium propionate (CAP), the salt of PA, possesses the antimicrobial effect, and offers the same efficacy as its corresponding acid after being ionized in water [7]. Therefore, the use of CAP as silage additive is assumed to be preferable compared with its acid during silage making in practice.

Evaluations of the application of PA/salt have been made mainly on high-moisture corn silages, which observed that PA/salt could suppress undesirable bacteria and improved the aerobic stability of silages [8]. However, as the effectiveness of chemical additives may vary hugely among ensilage materials, limited information is available about the recommended application rate of CAP for alfalfa silage. In addition, considering the hot and humid environmental conditions encountered during summer, it is difficult to achieve sufficient DM content in a by wilting. We hypothesized that CAP with antimicrobial attributes could be used as an additive to modulate the fermentation and inhibit the undesirable bacteria during the ensiling.

This study was aimed to evaluate the effects of varying con centrations of CAP on nitrogen compound compositions, fer-mentation quality and aerobic stability of high moisture alfalfa.

With the purpose of better modeling farmers’ systems of practices in southern China, where it is difficult to achieve the ideal DM by wilting a cut crop, fresh alfalfa having a DM content of 244 g/kg was ensiled directly without wilting. The DM content of a crop is crucial for the rate and extent of the resulting fermentation, and a low DM content coupled with low WSC is readily subjected to clostridial fermentation and subsequent poor acceptance of the silage by the animals [15]. In the experiment, low DM and low WSC (49.8 g/kg DM) of fresh alfalfa were limiting factors for a successful fermentation (Table 1). However, the epiphytic LAB populations were found adequate (5.76 log₁₀ cfu/g FW), considering the recommendation of 5 log₁₀ cfu/g FW as a minimum required for the good fermentation before ensiling [16].

After 30 days of ensiling, although no statistical difference in pH among treatments, there was an interesting tendency: the pH was first decreased and then rebounded to higher values when increasing the application rate in excess of 10 g/kg FW. The low solubility nature of this acid salt may be responsible for the diminished effectiveness on pH decline at high levels of application. In addition, McDonald et al [1] summarized the most of buffering properties of herbage can be attributed to the anions present, such as organic acid salt and orthophosphates. The higher pH in silages treated with higher levels of CAP may be also partly attributed to the increased buffering capacity, which brought by CAP per se.

The AA concentration in control was 30.2 g/kg, which is higher than that in other studies [17]. The high numbers of enterobacteria observed in the fresh alfalfa prior to ensiling caused the AA production, since enterobacteria is facultatively anaerobic and ferments sugars to acetic acid and other products before the silo becomes anaerobic [1]. The addition of CAP linearly decreased AA and ethanol concentrations, and this reduction may be attributed to their inhibition effect on undesirable microorganisms. Woolford [18] assayed the antimicrobial spectra of PA salts, and found that salts have desirable antimicrobial properties, which is superior to the corresponding free acid.

Alfalfa is difficult to ensile successfully mainly due to high buffering capacity and low WSC, while lower DM content resulted in the ensiling of alfalfa being more challenging compared with wilted alfalfa because of the increased possibility of clostridial fermentation [1]. The butyric acid concentrations in control silages were higher than the acceptable level (2.0 g/kg DM) for well-preserved silages [19], while it decreased as increasing proportion of CAP, which suggested that clostridia growth was suppressed by the antimicrobial activity of CAP. The restriction of clostridial fermentation by CAP is also typically reflected by lower ammonia N, which is another metabolic product of clostridia. The ammonia N in all CAP treated silages were decreased to a satisfactory level (<100 g/kg TN) [20], while that in control was 109 g/kg TN.

Over the process of ensiling, DM and nutrient losses seem likely to be unavoidable and irreversible, and mainly three factors may be involved in the DM consumption and dissipation of energy, including plant respiration, aerobic microorganism growth, and clostridia growth [21]. In the present study, the addition of CAP inhibited the growth of undesirable bacteria including enterobacteria, molds and clostridia, resulted in the lower DM losses. Arbabi et al [8] found that corn silages treated with CAP showed a significantly higher DM recovery than silage without additives.

The DM loss was related to the depletion of WSC. The linear increase in preservation of WSC with increasing rate of CAP might be attributed to the antimicrobial properties of CAP, which suppressed undesirable bacteria and reduced consumption of WSC. Silva et al [22] found higher WSC contents in corn silages treated with antimicrobial properties than control after 21 days of ensiling.

Degradation of herbage protein during ensiling is a compli cated biochemical process involving a series of plant and microbial enzymes. The linear increase in peptide-N suggested that CAP may depress the hydrolyzation of peptides brought by the metabolism of microorganisms. This statement is in agreement with Yuan et al [13], who reported the silages treated with CAP had higher peptide-N compared with that of control, and they ascribed it to the suppression on activity of detrimental microorganisms at later stage of ensiling.

The FAAs are one of the primary intermediate products of proteolysis by plant enzymes during the initial stage of ensiling and represent a pool of amino acids newly degraded by plant protease and amino acids rather than being metabolized by microorganisms [23]. In this experiment, CAP reduced ammonia N because of suppression of undesirable bacteria as referred above, suggesting less FAAs were degraded to ammonia. Combined with the lower FAA-N in CAP treated silages, weaker proteolysis by plant enzymes occurred during the initial stage of ensiling. It is known that proteases are deactivated at low pH, and Jones et al [24] reported that 67% of its original activities were lost after 24 h at pHs 4 and 5. It is speculated that the lower pH in CAP treated silages was responsible for the lower activity of proteases. Li et al [6] found FAA-N released from proteolysis decreased with decreasing silage pH from 6.23 to 3.40, and they ascribed it to the deactivation of plant proteases by low pH.

Silages are prone aerobic deterioration once they are exposed to air as the result of aerobic microbial activity: lactate-assimilating yeasts initialize aerobic deterioration, indicative of the degradation of lactic acid and production of excessive heat; following opportunistic bacteria (e.g., Bacilli) and molds (e.g., Aspergillus, Fusarium, and Penicillium) grow and further aggravate the spoilage in the silage mass due to the increase of silage pH resulted from degradation of lactic acid [25]. Studies showed that alfalfa silage can remain stable and does not spoil for more than one week during exposure to air [26]. However, silages deteriorated at a faster rate in our experiment compared with the results reported in alfalfa silages from other studies [26], especially in control, which had an aerobic stability of only 130 h. This poor stability was likely due to the high ambient temperature. Because environmental conditions encountered in the tropics facilitates the aerobic deterioration of silages resulting from the intensive proliferation of undesirable microorganisms. Ashbell et al [27] compared the aerobic stability of wheat and maize silages at different temperatures, and found that the silages stayed at 30°C were more prone to aerobic deterioration than those at 10°C, 20°C, or 40°C. It is deduced that the optimal temperature for growth of yeasts, which are known as the initiator of aerobic deterioration, may be about 30°C. In the experiment, the ambient temperature (28°C to 33°C) for aerobic stability test is close to 30°C, coupled with the high yeast counts at silos opening, all these together explained the fact that the alfalfa silages without CAP were more susceptible to aerobic spoilage. Addition of CAP delayed the rise in temperature of alfalfa silages: 130 h for control vs 140, 166, >168, >168 h for silage with CAP at 5, 10, 15, and 20 g/kg FW respectively. Kung et al [28] reported additive containing buffered PA substantially improved the aerobic stability of corn silages, while Kleinschmit et al [25] found that there was no effect on improving aerobic stability of corn silages treated with the same additive, and ascribed it to the lower additive applied rates, which was not sufficient to inhibit the growth of yeasts. Our data supported this fact, marginal improvement of alfalfa silages treated with low level (CAP5) was found, while silages treated with higher levels of CAP kept more aerobically stable over the periods of aerobic test.

Flieg’s point, based on the DM content and pH found in the silages, has been used for judging the quality of silages [14]. In the present study, control silage had a Flieg’s point of 60, characterized by a typical poor fermentation quality of legume; silages with 10 g/kg FW of CAP had the greatest score (75.9), represented the best silages among all treatments.

Addition of CAP can suppress the growth of undesirable microorganism during the ensiling and aerobic to air, thereby improving the fermentation quality, aerobic stability as well as retarding the proteolysis of alfalfa silages. Thus, adding CAP to alfalfa silage may be an alternative to silage additives used in southern China and the application rate of CAP is recommended at 10 g/kg FW when producing alfalfa silages.