INTRODUCTION

World fisheries production in 2016 reached approximately 171 million tons, of which aquaculture accounted for 47% of the total and fishing for 53%, not including what was destined for the production of fishmeal and fish oil (^{FAO, 2018}). More than 60% of residues made up of fins, scales, heads, viscera, skeleton, skin, roe and remains of meat derive from the industrial processing of fish (^{Ghosh et al., 2016}; ^{Renuka et al., 2016}). These wastes in many parts of the world are discarded, which causes a great loss of nutrients such as proteins, lipids and minerals, in addition to polluting the environment (^{Olsen and Toppe, 2017}). With fish waste, fertilizers, concentrates and protein hydrolysates, as well as fish meal and oil can be produced (^{Renuka et al., 2016}; ^{Ozyurt et al., 2017}). Biotechnologies have also been developed for the fish waste conversion into high-value products such as polyunsaturated fatty acids, physiologically important peptides, carbohydrates and other bioactive compounds (^{Ghaly et al., 2013}; ^{Smichi et al., 2016}). However, most of these technologies are not economically attractive because they require high investment (^{Ozyurt et al., 2017}). The most important input in animal production is fishmeal due to its high protein content, but due to its high cost, alternative sources of protein are demanded in the market (^{Castillo et al., 2019}). Fish silage is a product resulting from the preservation of whole fish or parts by the addition of organic or inorganic acids (chemical silage) or by bacterial fermentation (biological silage) (^{Ghaly et al., 2013}; ^{Olsen and Toppe, 2017}). The fish silage produced by the biological method is a viable technological alternative from the economic and environmental point of view, which consists of mixing the ground fish waste with molasses or another carbohydrate source and a starter culture of lactic acid bacteria (LAB ) (^{Ramírez-Ramírez et al., 2018}; ^{Castillo et al., 2019}). During fermentation, LABs increase the production of acids, mainly lactic, so the pH decreases and microbial deterioration slows down. Furthermore, fish proteases are activated, accelerating proteolysis and consequently the digestibility of the product increases (^{Ghaly et al., 2013}). Likewise, LABs generate compounds such as bacteriocins and hydrogen peroxide that help preservation and diacetyl, an aroma and flavor enhancing substance (^{Jini et al., 2011}). In this sense, to achieve control of fermentation, the selection of LAB strains is important. According to various studies, fish silage is an excellent source of proteins, lipids and minerals with great biological properties for animal feed (^{Geron et al., 2007}; ^{Ghaly et al., 2013}; ^{Ramírez-Ramírez et al., 2016}; ^{Land et al., 2017}), in addition, biological fish silage has antibacterial and antioxidant benefits and it is a possible source of probiotics (^{Jini et al., 2011}; ^{Ozyurt et al., 2017}). On the other hand, pineapple (Annanas comosus Merr.) Ranks third as the most popular and most economically important fruit in the world, with a production of 24'785,762 tons (FAOSTAT, 2018). From pineapple industrialization, approximately 75% of the weight of the fruit is obtained as waste, which is a valuable fiber source, soluble sugars, protein, ascorbic acid, vitamins, minerals, water and bioactive compounds such as bromelain from multiple applications (^{Damasceno et al., 2016}; ^{Ketnawa et al., 2012}). Pineapple waste can be used as a good quality substrate for microorganisms in fermentation processes; however, they are often thrown away causing serious contamination problems (^{Ketnawa et al., 2012}). For all the above, the use of fish and pineapple waste could be a viable and relevant alternative from an economic and environmental point of view. To our knowledge there are no reports on the use of fish waste and pineapple peel together. Therefore, the objective of this research was to evaluate the use of these industrial wastes in the silage production by lactic fermentation.

MATERIAL AND METHODS

Raw material preparation

Fish waste and pineapple peel

Fish waste was obtained from commercial marine species such as Bagre panamensis, Peprilus snyderi, Sphyraena ensis, Trachynotus ovatus, Argyrosomus regius, Diplodus vulgaris and Bagre panamensis (estuarine fish) from San Blas port, Nayarit, Mexico. The waste was processed in a meat mill (Torrey Brand, model 32-3, Mexico) using the 0.5 cm diameter sieve and stored at -20 °C until use. The peel of pineapple (Annanas comosus Merr.), Cayenne smooth variety, was obtained manually with a knife and homogeneously ground in a food processor. The following table shows the chemical composition of these ingredients.

Cane molasses and corn stubble

Cane molasses with a moisture content of 25.15%, 10.37% ash and 55.73% total soluble carbohydrates was used. To improve the consistency of the silages, corn stubble processed in a knife mill was added using the 2 mm sieve (Willey, model 4, Philadelphia, USA).

Inoculums

Lactobacillus B2 and Lactobacillus sp. strains were evaluated, the latter being isolated from mango waste in our laboratory. The initiators were cultured in MRS broth (from Man Rogosa and Sharpe, MRS, Merck Darmstadt) at 30 °C for 24 h until a final concentration of 1 X 109 cfu/mL was recorded.

Silage production

Six treatments were prepared with different amounts in percentage proportion of fish waste, PP (15, 30 and 45%, w/w), corn stubble, and sugar cane molasses (9%, w/w). As inoculum (I) Lactobacillus sp. or Lactobacillus B2 at 4% (v/p) (Table 2). The mixtures obtained were used to make 100 g mini silos in black plastic bags. Each treatment was prepared in triplicate and the silos were vacuum sealed and incubated at 30 ° C for 14 days.

Chemical analysis

The silages were analyzed every 0, 2, 4, 7 and 14 days (T) to determine the pH with a potentiometer model UB10 Ultra Basic (Denver Instrument, USA) and the content of lactic acid by titration. The silages obtained at the end of fermentation were analyzed to determine their proximal chemical composition (^{AOAC, 2005}) and the contents of neutral detergent fiber (NDF) and acid detergent fiber (ADF) by the method of ^{Van Soest et al. (1991)}.

In Vitro Digestibility of Dry Matter (IVDMD)

IVDMD was determined following the two-step technique of ^{Tilley and Terry (1963)}. The silage samples were dried at 70 °C for 24 h in a forced air oven and processed in a Wiley mill to a particle size of 1 mm. Two uncastrated Blackbelly sheep weighing 35 kg of body weight and equipped with a cannula in the rumen were used to collect the ruminal fluid. The sheep were fed a diet based on 25% corn silage, 25% alfalfa and 50% concentrate. In the second step of the IVDMD technique, pepsin (Sigma P-7012, Sigma) was used.

Statistical analysis

The data of the fermentation parameters were treated by analysis of variance (ANDEVA) for a factorial design 3 x 2 x 5 (level of pineapple peel x inoculum x fermentation time). The data obtained from the chemical composition and IVDMD were analyzed by ANDEVA for a 3 x 2 factorial design (level of pineapple peel x inoculum).

When a significant difference was found, the means of the treatments were compared with the Tukey test (p <0.05). The analysis was done with the Statistica 7 program (Statistica, versión 7.1).

RESULTS AND DISCUSSION

Fermentative parameters

The silage lactic acid content showed significant differences between treatments due to the inoculum factors (I) and fermentation time (T) (Table 3). The unfermented mixtures presented the lowest lactic acid content (average 0.55 ± 0.15%) with no difference between these treatments (p> 0.05). However, between days 2 and 7 of the process, most of the treatments with Lactobacillus B2 presented the maximum concentration of lactic acid (average 3.42 ± 0.17%). The high acidifying power of Lactobacillus B2 has been previously reported (Ramírez-Ramírez et al., 2008; 2018) and in this work it is reaffirmed. The results obtained agree with values of 2.96 to 4.42% of lactic acid recently reported for biological silage added with mixtures of fish and mango waste (^{Ramírez-Ramírez et al., 2018}), as well as with the values of 2.96 and 3.08% of lactic acid for fermented shrimp head silage and fish waste; respectively (^{Castillo et al., 2019}).

The pH was significantly affected by the main factors, pineapple peel level (PP), inoculum (I) and fermentation time (T); confirming the PP x I and PP x T interactions (Table 3). The unfermented mixtures presented the highest pH values and when the PP level increased there was a decrease in this parameter (p <0.05). In silages, it was observed that when the lactic acid production increased, the pH decreased significantly, obtaining the best pH values between days 2 and 7 of fermentation (Table 3), which is due to the fact that LAB reaches the growth curve top. However, the PP x I interaction showed that at 7 days of the process the pH of the treatments with 15 and 30% of PP and inoculated with Lactobacillus B2 was better than with Lactobacillus sp, which was also observed in a similar way to the 14 days in silage with 15% PP. The silage pH was affected by the PP x T interaction during the 0, 7 and 15 days of the process (p <0.05). However, between 2 and 4 days of fermentation, the pH values of the silages did not show significant differences, regardless of PP level and inoculum type (p <0.05). The treatments with 30 and 45% PP at 14 days of fermentation showed a small increase in pH, which was related to a decrease of 0.72 percentage units in the lactic acid production. The increase in pH was due to the buffering effect of fish proteins and peptides derived from their hydrolysis (^{Ramírez-Ramírez et al., 2008}; ^{Ghaly et al., 2013}), as well as the high amount of ashes provided by the bones from fish waste (Tables 2 and 4), since calcium salts act as neutralizers of lactic acid in silage during storage (^{Land et al., 2017}).

However, the silage obtained in this study presented acceptable sensory characteristics and did not show signs of decomposition. The pH values of this study silages agree with those of other reports (^{Castillo et al., 2019}; ^{Ramírez-Ramírez et al., 2018}), although they were higher than the 4.2 obtained for biological silage from residues of tilapia filleting (^{Gerón et al., 2007}).

Chemical composition

Table 4 shows the results of the chemical composition and IVDMD of the silages obtained after 14 days of fermentation. The dry matter content (DM) decreased significantly when the PP level increased, therefore the silages with 45% PP presented the lowest DM content, regardless of the inoculating strain (p <0.05). The silages with 15% PP presented the highest DM values (39.3%), despite having a high content of fish waste (57%) in their formulation, whose moisture percentage is high (Tables 1, 2 and 4). This was probably due to the addition of 15% corn stubble, which also improved the silage consistency. The DM results obtained agree with those of other reports (^{Geron et al., 2007}; ^{Castillo et al., 2019}; ^{Ramírez-Ramírez et al., 2018}). The PP level and the PP x I interaction were significant in the ash content. The silages with 15% PP and inoculated with Lactobacillus sp. presented the highest ash content (14.5%) (p <0.05), however, the other treatments presented a good concentration of minerals, because fish waste is an important source of these nutrients Tables (1 and 4). The ash content obtained is similar to other investigation results (^{Castillo et al., 2019}; ^{Ramírez-Ramírez et al., 2018}), although they were lower than 18.7% reported by ^{Geron et al. (2007)}.

The silages showed a significant difference in the crude protein content due to the type of inoculum used. The silages with 45% PP and inoculated with Lactobacillus B2 presented numerically the highest crude protein content (31%), but statistically the PP level and the inoculum did not produce relevant changes in silage crude protein (Table 4). The crude protein content of silages obtained with 45% PP and Lactobacillus B2 coincides with 31.6% of biological silage from tilapia waste (^{Geron et al., 2007}), although it was higher than 28.08% of crude protein from biological silage of fish and mango waste (^{Ramírez-Ramírez et al., 2018}); however, it was lower than the 35.42% reported by ^{Castillo et al. (2019)} for biological silage of fish waste, molasses and yogurt.

Silage ethereal extract showed a significant difference due to the main effects. In this regard, treatments with 30 and 45% PP and inoculated with Lactobacillus B2 showed the highest lipid content (9.85%), which is very important from a nutritional point of view, since essential fatty acids are found in them for animal feed (^{Vidotti et al., 2011}). The ethereal extract results of this study are lower than those reported by ^{Geron et al. (2007}) y ^{Castillo et al. (2019)}; however, they coincide with the findings of another study (^{Ramírez-Ramírez et al., 2018}). The NDF and ADF contents decreased with increasing PP level. Although PP is a good source of fiber (Table 1), the trend in the results was due to the addition of corn stubble in the formula, since the silages with a higher PP content at the same time contained a lower amount of corn stubble and therefore less NDF and ADF content (Tables 2 and 4). In this sense, the results are very important, since the contents of NDF and ADF in forages are negatively correlated with consumption and digestibility. The percentages of NDF and ADF obtained in the silages of the present study were higher than those reported by ^{Ramírez-Ramírez et al. (2018)}.

In Vitro Digestibility of Dry Matter (IVDMD)

The digestibility values of food components are important parameters to evaluate the formulation of diets and to determine the use of a food component (^{Ozyurt et al., 2017}). The IVDMD values of the silages presented a range of 76.2 to 84.4% with significant differences between treatments due to the main effects (p <0.05). By increasing the PP level to 45% and with the addition of Lactobacillus sp., the silages presented an increase of 6.1 percentage units in the IVDMD to reach 82.6% (p <0.05). However, this result was statistically the same when using Lactobacillus B2, regardless of PP level (Table 4). In fish silage production, the proteases present in the acidic medium hydrolyze the proteins into smaller fragments, peptides and amino acids, which affects the total digestibility (^{Ghaly et al., 2013}; ^{Geron et al., 2007}; ^{Ramírez-Ramírez et al., 2016}; ^{Olsen and Toppe, 2017}). The PP effect on the increase in IVDMD was probably due to an increase in nutrient availability, which was related to the decrease in fiber fractions and consequently increased ruminal microbial activity. Likewise, the addition of PP probably increased the activity of the pepsin used in the second step of the digestibility test, which simulates stomach digestion and therefore the IVDMD of the silages increased. LABs are better known as starter cultures due to their versatile metabolic characteristics such as acidifying activity, proteolytic activity, and bacteriocin synthesis (^{Jini et al., 2011}). In general, IVDMD of the silages was higher with Lactobacillus B2 than with Lactobacillus sp., due to the better acidifying power of Lactobacillus B2 and probably to a high production capacity of digestive enzymes. The IVDMD results obtained in this work agree with those of other reports (^{Ozyurt et al., 2017}; ^{Ramírez-Ramírez et al., 2018}).

CONCLUSIONS

The inclusion of pineapple peel in 15 and 30% and Lactobacillus B2 caused the best acidification of silages at 7 days of fermentation. However, at 14 days all silages were stable and presented high nutrient content. Furthermore, the use of Lactobacillus B2, regardless of pineapple peel level produced the highest IVDMD of the silages. Silage production with fish waste and pineapple peel in combination with molasses and corn stubble is a simple, economical and environmentally friendly technological alternative. It is recommended to scale up the production process and evaluate the silage in ruminant feeding.