Newly weaned pigs are very sensitive to the quality of their feed. The wrong feed can result in a lower feed intake and a reduced growth rate due to the incomplete development of their digestive system (Zhang et al., 2003). In particular, certain sources of dietary proteins can cause these problems by producing an allergic response (Bimbo and Crowther, 1992; Maxwell et al., 2003). At this point in their development, pigs require a good source of protein and amino acids in their diets; and there are increasing concerns about finding the most practical protein source for newly weaned pigs that will have the greatest positive influence on the function and structure of their digestive tract. An animal protein source that is widely used for early-weaned pigs (as early as 19-21 days) is fish meal. Fish meal is a very digestible protein source, with a high mineral content and low fiber, that is commonly used to stimulate feed intake (FAO, 2001; Kim and Easter, 2001; Jones et al., 2015). The amino acid composition of fish meal protein is very similar to both milk of sow and body tissue of piglets (Fowler, 1997). Wang et al. (2009) showed clear improvements in growth performance when weanling piglets were fed diets containing fish meal. Other reports have also suggested that adding fish meal to the diet of farmed animals could provide several advantages to animal health, including improved immunity against disease, higher survival and growth, and reduced incidence of deformities (FAO, 1986).

However, falling supply of and rising demand for fish meal has led to uneconomically high prices. In part, this is because the quality of the fish meal depends on the type and species of fish, its freshness, and the processing of methods. Recent approaches advocating the use of fish byproduct or green algae have focused on fermentation to improve the quality of the animal feed for pig production. Thus, we hypothesized that fermented fish meal with microorganisms would improve the growth performance in weanling pigs as the partial replacement of high-quality protein. The objective of our study was to further evaluate the effect of dietary supplementation with fermented fish meal on growth performance, blood metabolites, and fecal microflora in weanling pigs.

The animal experimental protocols were conducted according to the guidelines of the Animal Care and Use Committee of Gyeongsang National University, Jinju, South Korea. A total of 180 weaned pigs ((Landrace × Yorkshire) × Duroc; mixed sex; weaned at 21±1d of age) with initial body weight of 6.0±0.2 kg were randomly distributed among three dietary treatments of three replicate pens (20 weanling pigs per pen) in a completely randomized trial over three weeks. The dietary treatments comprised corn-soybean meal-based diets with 0% (control), 0.2%, and 0.5% fermented fish meal. The commercial diet was used throughout the experiment period (Table 1).

Fish meal, mainly containing cod (Gadus macrocephalus) byproduct, was purchased from Dongchang, Busan, South Korea. To obtain fermented fish meal, fresh fish meal (50%), rice bran (30%), and wheat bran (20%) were mixed and 1% of micro-organisms (Lactobacillus acidophilus, Bacillus subtilis, and Saccharomyces cerevisiae) was applied to meet the application ratio at 1.2 × 10⁵ cfu/g (Table 2). Then, the mixture was fermented at 39 ^oC for 72 h. The fermented fish meal was then dried by enforced-air fan and ground into powder.

The pigs were weaned and housed in an environmentally controlled and slatted-floor facility in 20 pens (2.0 m × 1.8 m). Each pen had one self-feeder and a nipple waterer to allow for ad libitum access to feed and water. The temperature was maintained at about 32 °C during the first week and lowered by 2 °C per week down to 28 °C in the third week post-weaning. The relative humidity was kept at 60~70% during the whole period. Each pig was weighed at the beginning (1 d) and the end (21 d) of the experiment period to calculate average daily gain (ADG). Feed intake was also recorded weekly for each pen to measure average daily feed intake and gain:feed ratio (G:F). To determine the amounts of dry matter (DM), crude protein, ether extract, and crude ash, the experimental diets were analyzed in accordance with the methods described by the AOAC (1995) (Table 1).

After three weeks, two pigs from each pen were bled via jugular venipuncture. Blood samples were collected in 5-mL non-heparinized vacuum tubes and 5-mL vacuum tubes containing K₃EDTA (Becton Dickinson Vacutainer Systems, Franklin Lakes, NJ, USA) to obtain whole blood and serum, respectively. The counts for red blood cells (RBC), white blood cells (WBC), hematocrit (Hct), hemoglobin (Hgb), lymphocytes, and monocytes in the whole blood samples were determined, using an automatic blood analyzer (ADVIA 120, Bayer, Tarrytown, NY, USA). In addition, the samples were centrifuged at 3,000 × g for 15 min at 4 °C to separate the serum. Serum immunoglobulin G (IgG) level was determined using ELISA kits, while the concentrations of insulin-like growth factor 1 (IGF-1) and insulin were determined using radioimmunoassay kits. Serum samples were also used to determine blood glucose and blood urea nitrogen (BUN) levels.

To measure Salmonella enterica and Escherichia coli loads, fecal samples (100 g) were collected weekly from each pen at four random locations and immediately analyzed. Ten grams of samples were weighed and placed into a stomacher bag with 100 mL of phosphate-buffered saline buffer (pH 7.4). Samples were homogenized thoroughly for 1 min. Serial 10-fold dilutions (in 0.1% peptone solution) of the fecal samples were then plated onto Difco TM SS agar (Becton, Dickinson and Company, Sparks, MD, USA) and Difco TM Violet Red Bile agar (Becton, Dickinson and Company, Sparks, MD, USA), respectively. Difco TM SS agar and Difco TM Violet Red Bile agar plates were incubated for 24 h at 37 °C. After being removed from the incubator, the colonies, as average colony-forming units (cfu)/g litter, were counted immediately at week 1, 2, and 3, respectively.

All data were subjected to ANOVA using the GLM procedure of the SAS package program (Statistical Analysis System, version 8). The IML procedure was used to develop linear and quadratic coefficients for data with unequal spacing. Tukey's test was also used to identify treatment means to compare the difference between means. Significance of treatment effects was declared at P<0.05.

Adding fermented fish meal to weanling pig diets had a linear effect (P<0.05) on average feed intake and a quadratic effect (P<0.05) on final body weight (FBW), ADG, and G:F throughout the period. There was no significant difference among diets (linear or quadratic; P>0.05) with the initial body weight (Table 3). Growth performance was greater in the diet with 0.2% fermented fish meal than in the other treatments.

Overall, values on blood metabolites showed differences among all supplemented diets (Table 4). Hematocrit, monocyte, IgG, and BUN responded linearly and quadratically (P<0.05) with increasing levels of dietary fermented fish meal. Moreover, differences (P<0.05) were detected linearly among diets for lymphocytes and insulin. However, RBC, WBC, Hgb, IGF-1, and glucose were not affected by diets with different levels of fermented fish meal over the three-week period (linear and quadratic; P>0.05).

During the experimental period, diets with 0.2% and 0.5% fermented fish meal showed reductions in fecal microbes that were linear, quadratic, or both, compared with controls (Table 5). In particular, there was a significant reduction in S. enterica and E. coli populations when weanling pigs were fed the 0.5% fermented fish meal at week 1 and week 2. However, no differences (linear and quadratic; P>0.05) among treatments were found in E. coli populations at week 3.

Our study finds direct evidence that the inclusion of fermented fish meal improves the growth performance of weanling pigs. Similarly, Kim and Easter (2001) reported that the ADG in young pigs, between 3 and 5 weeks old, was increased by diets with either mackerel or herring fish meal compared with other feed supplements. Improvement in the feed efficiency of weanling pigs due to 5.0% fish byproduct supplementation was also reported by Noh et al. (2014). In our study, the reasons for the improved growth performance of weanling pigs appears to be that fish meal is a protein source with high nutrient digestibility (Jones et al., 2015). In addition, fermented fish meal has been reported to offer benefits for weaning pigs by improving flavor and enriching the feed with desirable metabolites produced by the microorganisms (Buckenhüskes et al., 1990; Cho and Kim, 2011). There were no effects on FBW and ADG between supplementation levels of fermented fish meal. However, pigs fed diet with 0.5% fermented fish meal had greater average feed intake compared with pigs fed 0.2%. These results lead to the greater G:F in 0.2% fermented fish meal compared to 0.5%. This improvement might be due to the nutrient digestibility enhance at 0.2% fermented fish meal. In our study, the parameters we chose for the blood profile (Hct, lymphocyte, monocyte, IgG, insulin, and BUN) had a greater effect on the metabolism and immune status in weanling pigs fed on diets with 0.2% and 0.5% fermented fish meal. In other words, elevated blood metabolites might allow better utilization of fermented fish meal due to improved gut flora in the weanling pig, which might affect lymphocyte, monocyte, and IgG levels. For example, lymphocyte proliferation is an important phase in determining cell immunity and clinical immune function of the animal body (Lafuente et al., 2003). According to Zinnerman (1998), the presence of immunoglobulin had a positive effect on immunity. Likewise, glucose and BUN are important for growth. However, there is little or no research evaluating the effect of fermented fish meal on blood profiles; therefore, more evidence is needed to confirm our results. In another study using fermented soybean meal, Liu et al. (2007) reported decreased serum IgG and lowered whole blood and spleen lymphocyte proliferation in weaned piglets. In the case of fermented fish meal, the blood parameters selected appear to be correlated with growth performance of weanling pigs. It has been well documented that gastrointestinal microflora affects animal production because the activation of the gastrointestinal immune system significantly affects the intestinal morphology and the ability to digest and absorb nutrients in pigs (Yan et al., 2012; Liu, 2015). Consequently, the results in our study indicate that fecal microflora is closely related to the production performance in weanling pigs. Fermented fish meal reduces the populations of S. enterica and E. coli by creating gut micro-ecological conditions that suppress harmful microorganisms or favor beneficialmicroorganisms, as previously suggested by Lee et al. (2014). Our results support the findings of Noh et al. (2014), who indicated that diets supplemented with 5.0% citrus pulp, fish byproduct, and B. subtilis fermentation biomass have the potential to improve fecal microflora of weanling pigs.

Feeding fermented fish meal in increasing dietary levels (0.2% and 0.5%) to weanling pigs is an effective protein source to improve growth performance and reduce harmful microorganisms (S. enterica and E. coli) in the feces. In addition, hematocrit, lymphocyte, monocyte, immunoglobulin G, insulin, and blood urea nitrogen have a greater effect on metabolism and immune status in weanling pigs fed diets with 0.2% and 0.5% fermented fish.