4.1. Chemical Composition, In Vitro Fermentation, and In Vitro Intestinal Digestibility of TP Samples
Chemical composition of TP was within the range of values previously reported [
4,
26,
27], although other studies [
2,
28] reported lower lignin and greater EE values. Tomato pomace is a fibrous by-product with a relatively high content of CP and EE. As reported by others [
8,
29,
30], most of fiber and CP comes from the peel fraction of TP, whereas seeds contain most of ash and EE; therefore the relative proportions of peels and seeds in TP can help to explain the observed differences in lignin and EE due to sampling time in the present study, as well as variations in chemical composition among studies [
31]. In addition, chemical composition of tomato fruits has been reported to be highly variable, depending on factors such as cultivars, crop conditions, origin, etc. [
8], which also contributes to variability in TP composition. Additional factors, such as tomato industrial processing characteristics, can increase variability. However, in the present study, there were no differences in any chemical fraction between the TP samples from the two processing plants, which might be due to a similar processing type in both of them. Finally, the application of high temperatures during tomato processing may result in the formation of indigestible compounds via the Maillard reactions between sugar aldehyde groups and free amino groups [
32] that resulted in high ADIN values (16%–26%) reported in some studies [
2,
33,
34]. However, ADIN concentrations in our study were relatively low, ranging from 6.6% to 8.8% of total N. Differences in the ADIN content could be attributed to variability in tomato fruits or processing [
8], but also to different processing temperatures.
Despite the relatively low variability in TP chemical composition observed in our study, there were some differences between sampling times in PGP, AGPR, and DEMD, with samples taken at week 6 having lower values than those taken at week 2. Abbeddou et al. [
26] pointed out that CP and non-structural carbohydrates of TP are easy and rapidly degraded, whereas NDF has low degradability. The low lag values (≤2.63 h) agree with this observation. These results indicate that TP was rapidly fermented in vitro.
Values of total VFA production agree well with gas production parameters, as TP samples taken at week 6 had lower values than the rest of the samples, except those taken at week 1. Despite the fact that there were significant effects of sampling period on molar proportions of all individual VFA, changes were of minor importance and the VFA profile remained relatively constant over the sampling period.
The main fatty acids present in TP are oleic and linoleic acids [
35] and the unsaturated fatty acids which are produced by the hydrolysis of triglycerides are toxic to the fibrolytic bacteria [
36]. However, the lack of correlation (
p ≥ 0.339) between EE content and any gas production or in vitro fermentation parameter indicates that EE had no negative effect on TP fermentation, despite the fact that its concentrations were greater than the maximal level of 60 g of fat per kg of DM recommended in the diet of ruminants to avoid reductions in fiber digestibility [
37]. There were some differences between samples in CH
4 production, but these differences disappeared when the CH
4/total VFA ratio was calculated.
Although OM content showed low variability among TP samples, it was the chemical fraction more closely related to gas production parameters and VFA and CH
4 production, with the correlation being negative; this might be explained because of the close relationship between OM and the lignin/NDF ratio (r = 0.830;
p < 0.001), as lignification is one of the main factors limiting NDF degradability [
38]. In fact, the lignin/NDF ratio was negatively correlated with PGP, DMED, and total VFA, indicating that NDF lignification was a major factor involved in the differences observed in fermentation parameters.
Values of 16 h in situ rumen DM degradability were similar to those reported by Abbeddou et al. [
26] and Gasa et al. [
28] after 16 and 24 h of ruminal incubation, respectively (476 and 471 g/kg), but greater than the 389 g/kg reported by Fondevila et al. [
2] after 24 h of incubation in the rumen. Values of in situ degradability of CP were greater than the 406 g/kg reported by Fondevila et al. [
2] for 24 h of incubation, but lower than the 649 g/kg reported by Abbeddou et al. [
26] for 16 h of incubation. Differences among studies are probably due to variability in TP chemical composition, but may be also related to the performance of the in situ procedure which is influenced by many factors, such as animals, diet, incubation procedure, etc. [
39]. In situ degradability of both DM and CP was negatively correlated with ADF content, reflecting the low degradability of this fraction, which is the main constituent of fiber in TP; this might also restrict the access of microorganisms to protein.
The values of IDCP revealed that only nearly half of the CP undegraded in the rumen can be digested in the intestine, and total CP digestibility reached 737 g/kg as average. These values agree with previous observations reporting lower protein digestibility in TP compared with protein concentrates such as soybean meal [
2,
40]. The positive correlation of IDCP with ADF fraction is in accordance with the observation that intestinal digestibility of CP increases as rumen degradability of CP decreases [
41]; this would explain the fact that ADF was negatively correlated with in situ CP degradability and positively with IDCP. The negative correlation between IDCP and TSP content may indicate that TSP formed less-digestible complexes with dietary proteins.
4.2. In Vitro Fermentation of Diets Containing TP
The increased fractional rates of gas production indicate that diets including TP were fermented more rapidly than the control diet, which is in agreement with the increased VFA productions observed at 24 h of incubation, although only TP18 diet showed significant differences with the control diet. These results are consistent with the lower amounts of barley straw in the TP-diets than in the control diet, as barley straw is slowly degraded in the rumen and greater VFA productions and FOM were observed for the TP-diets after 24 h of incubation. Including TP in the diet also changed the VFA profile. Our results are in agreement with those of Soto et al. [
42], who observed that in vitro molar proportion of acetic acid increased and butyric and minor VFA proportions decreased when tomato fruit wastes were included in the diet. Similarly, Arco-Perez et al. [
7] reported an increase in propionate proportions and a decrease in the acetate/propionate ratio in the rumen of goats receiving tomato silage compared with those fed a control diet.
The reduced NH
3-N concentrations observed with increasing TP levels could be due to the lower degradation of TP protein, as NH
3-N is one of the main final products of protein degradation in the rumen. The decreased proportions of minor VFA as the level of TP in the diet increased are consistent with this hypothesis, but NH
3-N concentrations were adequate for rumen microorganism growth in all diets [
43]. Tomato pomace replaced increased amounts of soybean meal in the experimental diets and rumen degradability of TP has been reported to be lower than that of soybean meal [
2], which would indicate greater bypass protein in TP than in soybean meal. According to Drouliscos [
44], the essential amino acids profile in TP protein is similar to that in soybean meal, but the low values of IDCP observed in our study indicate that about half of the protein reaching the small intestine may not be digested, whereas intestinal digestibility of soybean meal is greater than 90% [
43]. In agreement with these results, Fondevila et al. [
2] and Yuangklang et al. [
40] reported a significant decrease in CP digestibility when replacing soybean meal by TP in the diet of lambs and beef cattle, respectively.
Although some studies have shown that the inclusion of tomato fruit wastes in the diet reduced CH
4 emission both in goats [
7,
9,
10] and in in vitro fermentations [
11], no changes in CH
4 production were observed in our study. The anti-methanogenic mechanisms of tomato fruit ingredients remain yet unknown [
7], but either the extraction of the bioactive compounds during the industrial processing of tomato or their destruction by the heat applied in the process might help to explain the lack of anti-methanogenic properties of TP. The reduction of the CH
4/total VFA ratio observed in TP-diets after 8 h of incubation could be due to the fermentation of soluble and rapidly degraded fractions, which also contributed to greater propionate proportions.