4.3 | Additional methodological considerations
Dietary metabarcoding techniques are being increasingly applied to investigate the diets of herbivores and omnivores, including the diets of pigeons, bears, monkeys, deer, zebras, and goats (Ando et al., 2013; Quemere et al., 2013; De Barba et al., 2014; Kartzinel et al., 2015; Gebremedhin et al., 2016; Erickson et al., 2017). Much higher resolution is provided by the molecular analysis in comparison with the traditional alkane analysis of scats (Zhang, Jin, Badgery, & Tana, 2017). Both frequency of occurrence (FOO) data of food taxa and relative read abundance (RRA) of sequences are commonly used to convert sequence reads to dietary data. Both metrics have inherent biases when the number of food taxa in samples is large. In the current study, the diet data were presented using these two approaches to reduce any biases.
Accumulating evidence suggests that the number of sequence reads is approximately proportional to relative biomass of prey in a diet sample (Elbrecht, & Leese, 2015; Evans et al., 2016; Hänfling et al., 2016; Clarke, Beard, Swadling, & Deagle, 2017). However, the relationship between organism biomass and number of sequences can be variable due to PCR amplification bias of barcode primers (Albaina, Aguirre, Abad, Santos, & Estonba, 2016), and such biases were observed in our previous studies (Guo, Zhang, Chen, & Zhang, 2018). The utilization of different sets of barcode primers amplifying different gene regions could further improve the taxonomic resolution of the dietary items. Besides the consideration of primers, the prey taxa are important factors associated with the amplification bias. Our previous research showed that the sequencing process consistently overrepresented some herbage taxa (e.g., Leguminosae and Compositae) and underestimated others (e.g., Gramineae) relative to actual biomass. Nevertheless, a significant positive correlation between the DNA sequence proportion acquired from the sequencing procedure of feces samples and the actual biomass proportion in the diet was observed in this study, providing support for quantitative investigation using metabarcoding dietary analysis. Additionally, the determination of relative abundances of target items could be improved through post-sequencing quantification methods or correction factors (Matesanz et al., 2019).
Although metabarcoding is considered as a robust technique for assessing a species’ presence and its relative abundance, there are some shortcomings. Some common Poaceae species (such as Zea mays ) were missing from the diet of lambs in the time-limited group, thus being consistent with the findings from an earlier study (Guo, Zhang, Chen, & Zhang, 2018). The lack of certain food taxa may have resulted from methodological biases or from a draconian cutoff in the PCR protocols, as proposed by Vesterinen et al. (2016) and Clarke et al. (2014). In addition, incidental consumption of non-food DNA in the drinking water and fungi in the digestive tract can hinder the identification of certain groups. Furthermore, metabarcoding depends on single gene regions to detect food items, and this may cause that numerous related species to be indistinguishable (e.g., Lespedeza potaninii andLespedeza chinensis ). Having said this, the utilization of primer sets targeting composite barcodes (e.g., ITS2 and trn H-psb A) should be taken into consideration in future studies to further improve the taxonomic resolution and obtain a more thorough analysis (Pang, Shi, Song, Chen, & Chen, 2013). Likewise, establishment of reference databases targeting regional location-based items has been suggested as an important consideration of diet studies.
In conclusion, our results demonstrated that grazing systems could affect the fatty acid composition and promote n-3 PUFAs deposition in Tan sheep. Grazing and time-limited grazing feeding systems had lower ratios of n-6/n-3 fatty acids in the muscle than stall-fed modes, a result that was more in line with human health needs. Moreover, this research systematically studied the correlations between herbage taxa and omega-3 fatty acids and thereby has provided an initial view of the herbage effect on PUFAs of lambs. The significant correlations found in this study provide experimental evidence for future feeding research. The diet-PUFAs in the meat quality regulation of sheep may be one of the key research points for in-depth understanding of the flavor accumulation of beneficial health-related PUFAs.