Herbivore-Diet Analysis Based on Illumina MiSeq Sequencing: the Potential Use of an ITS2-Barcoding Approach to Establish Qualitative and Quantitative Predictions of Diet Composition of Mongolian Sheep
Yanping Guo,Hao Zhang,Wenqing Chen,Yingjun Zhang
DOI: https://doi.org/10.1021/acs.jafc.8b02814
IF: 6.1
2018-01-01
Journal of Agricultural and Food Chemistry
Abstract:DNA-barcoding approaches to estimate the diet compositions of grazing animals have received significant attention, and particularly when combined with next-generation sequencing, these techniques have substantially improved in recent years. In this study, the identity and species composition of plant material ingested by Mongolian sheep were estimated through the use of 350 bp ITS2 gene sequences of the vegetation found in fecal samples. Four diets were formulated using varying amounts of eight plant species that are common in the grasslands of northern China. Sixteen Mongolian sheep were taken from pastures and randomly assigned to four groups, and each group received one of four diets. Each sheep was randomly assigned to one of 16 confinement pens and fed its respective diet for 12 consecutive days. Fecal samples were removed from each pen from days 7-12, preserved, and composited for each pen. All herbage species included in the daily diets were detected in each fecal sample, with the exception of Phragmites australis. Moreover, 12 additional different plant species were retrieved from feces of the experimental sheep. The obtained data provided preliminary support for the use of the ITS2 barcode to determine which plants were consumed. Moreover, the proportions of the herbage DNA sequences recovered from sheep feces and those of the herbage masses in the daily diets did not completely match. These results indicate that the non-Gramineae DNA sequences amplified with ITS2 primers (including those of Chenopodium album, Artemisia scoparia, Artemisia tanacetifolia, and Medicago sativa) far exceeded those of the Gramineae species (including Leymus chinensis and Puccinellia distans), which constitute the largest share of the experimental diets. A significant positive correlation (Spearman's ρ = 0.376, P = 0.003) between the actual herbage mass proportions in the experimental diets and the herbage-DNA-sequence proportions provided sufficiently favorable support for the further investigation of DNA barcoding for the quantification of plants in feces. A significant regression coefficient was found between the relative DNA-sequence proportions of L. chinensis ( R2 = 0.82, P < 0.0001), P. distans ( R2 = 0.64, P = 0.0017), and C. album ( R2 = 0.98, P < 0.0001) and their respective herbage mass proportions. The quantitative relationship can be expressed by the linear-regression equations y = 0.90 x - 0.22, y = 0.98 x - 0.03, and y = 5.00 x - 0.25, respectively. Thus, these results demonstrate that dietary-DNA-barcoding methods exhibited potential in providing valuable quantitative information regarding food-item components. However, it should be noted that this explorative data needs to be further improved by using additional genes and by creating a sophisticated reference database, thus enhancing both quality and accuracy of the obtained results.