基因组选择利用表型与密集的全基因组单核苷酸多态性(Single nucleotide polymorphisms, SNPs)来估计个体的基因组估计育种值(Genomic estimated breeding value, GEBV)。目前，基因组最佳线性无偏预测(Genomic best linear unbiased prediction, GBLUP)是预测复杂性状最广泛的工具，随着技术不断发展，使用情境日下迁变，研究者们对预测的准确性提出更高要求。侧重于对模型本身进行拓展的相关研究将传统模型中的随机效应部分重新划分为多个子集并赋权，以此来提高预测精度，而倾向于利用多组学数据的研究则捕获基因组测序水平之外的变异来辅助提高预测精度。在本研究中，为了在大动物牛中提高基因组选择的预测精度，我们将具有基因组与转录组数据的群体设置为训练群体，以华西牛的生长性状中的背最长肌重量性状(Longissimus dorsi muscles, LDM)，与肉质性状中的系水力(Water holding capacity, WHC)、剪切力(Shear force, SF)以及pH的数据作为表型数据，使用贝叶斯稀疏线性混合模型(Bayesian sparse linear mixed model, BSLMM)、全转录组关联分析(Transcriptome-wide association study, TWAS)及表达数量性状基因座(Expression quantitative trait locus, eQTL)映射的方法，根据βb>0、前1%表型方差解释率(Percent of phenotypic variation explained, PVE)、与表达关联的单核苷酸多态性位点(expression-associated single nucleotide polymorphisms, eSNPs)及基因(egenes)错误发现率(False discovery rate, FDR)< 0.01的条件进行基因组特征预筛选，而后将这些显著的预筛选位点分别设置为额外的固定效应(GBLUP-Fix)与随机效应(GFBLUP)，用以改进模型并在验证群体中进行验证，同时，将传统GBLUP方法以及基于随机筛选的位点进行设置的GFBLUP与GBLUP-Fix进行比较。结果表明，在GFBLUP与GBLUP-Fix模型下，不同策略预筛选位点的加入，将LDM、WHC、SF、pH性状的预测精度平均提高了2.14%至8.69%。其中GFBLUP-TWAS在SF方面相较于GBLUP模型，预测精度提高了13.66%。此外结果也表明这些方法能够捕获比GBLUP模型更多的遗传变异。我们的研究验证并强调了多组学辅助的大效应位点预筛选策略在提高大动物的基因组预测准确性上的可行性，这为筛选位点并用于华西牛低密度SNP芯片序列的设计工作奠定基础。

本研究依托于29年的黑土长期定位试验，供试作物为玉米。肥料试验包括不施肥对照（CK）、施氮、钾肥（NK）、氮磷钾平衡施肥（NPK）、氮磷钾+有机肥（NPKM）共4 个处理。选取了2018年0-20，20-40，40-60 cm的土壤样品，测定土壤性质，并采用Hedley连续浸提法分析了土壤磷形态。试验结果表明，长期有机无机配施处理（NPKM）促进了磷在土壤剖面的积累，与初始值相比，其全磷含量增加了0.6-1.6倍。NPKM处理中，其剖面的有效磷含量也远远超过了黑土的环境阈值（50.6 mg kg^-1）。此外，NPKM处理中土壤剖面各活性和中活性态无机磷（NaHCO₃-Pi，NaOH-Pi和diluted HCl-Pi）的含量与比例显著高于NPK处理。这表明，NPKM处理可促进土壤稳定态磷向活性较高的磷形态转化。这可能是由于NPKM降低了土壤对磷的固定。冗余分析结果表明，由施肥引起的土壤有机质、Mehlich3-Fe和络合态铝氧化物含量的增加是影响黑土0-20 cm磷形态差异的主要因素。土壤矿物组分，如游离态的铁氧化物、碳酸钙是影响黑土深层土壤磷形态差异的主要因素。这表明黑土磷的转化过程同时受土层和施肥的影响。由于磷在不同土层中的积累和转化导致的高含量的有效磷以及有机肥的施用增加的活性较高的磷形态含量，我们在生产实践中应适当的减少或优化化肥的投入，以实现农业与环境的可持续发展。

In order to reveal the impact of various fertilization strategies on carbon (C) and nitrogen (N) accumulation and allocation in corn (Zea mays L.), corn was grown in the fields where continuous fertilization management had been lasted about 18 years at two sites located in Central and Northeast China (Zhengzhou and Gongzhuling), and biomass C and N contents in different organs of corn at harvest were analyzed. The fertilization treatments included non-fertilizer (control), chemical fertilizers of either nitrogen (N), or nitrogen and phosphorus (NP), or phosphorus and potassium (PK), or nitrogen, phosphorus and potassium (NPK), NPK plus manure (NPKM), 150% of the NPKM (1.5NPKM), and NPK plus straw (NPKS). The results showed that accumulated C in aboveground ranged from 2 550–5 630 kg ha–1 in the control treatment to 9 300–9 610 kg ha–1 in the NPKM treatment, of which 57–67% and 43–50% were allocated in the non-grain organs, respectively. Accumulated N in aboveground ranged from 44.8–55.2 kg ha–1 in the control treatment to 211–222 kg ha–1 in the NPKM treatment, of which 35–48% and 33–44% were allocated in the non-grain parts, respectively. C allocated to stem and leaf for the PK treatment was 65 and 49% higher than that for the NPKM treatment at the both sites, respectively, while N allocated to the organs for the PK treatment was 18 and 6% higher than that for the NPKM treatment, respectively. This study demonstrated that responses of C and N allocation in corn to fertilization strategies were different, and C allocation was more sensitive to fertilization treatments than N allocation in the area.

Soil organic carbon (SOC) and soil Olsen-P are key soil fertility indexes but information on their relationships is limited particularly under long-term fertilization. We investigated the relationships between SOC and the percentage of soil Olsen-P to total P (PSOPTP) under six different 15-yr (1990-2004) long-term fertilizations at two cropping systems in northern China. These fertilization treatments were (1) unfertilized control (control); (2) chemical nitrogen (N); (3) N plus chemical P (NP); (4) NP plus chemical potassium (NPK); (5) NPK plus animal manure (NPKM) and (6) high NPKM (hNPKM). Compared with their initial values in 1989 at both sites, during the 11th to 15th fertilization years annual mean SOC contents were significantly increased by 39.4-47.0% and 58.9-93.9% at Gongzhuling, Jilin Province, and Urumqi, Xinjiang, China, under the two NPKM fertilizations, respectively, while no significant changes under the no-P or chemical P fertilization. During the 11th to 15th fertilization years, annual mean PSOPTP was respectively increased by 2.6-4.2 and 5.8-14.1 times over the initial values under the two chemical P fertilizations and the two NPKM fertilizations, but was unchanged in their initial levels under the two no-P fertilizations at both sites. Over the 15-yr long-term fertilization SOC significantly positively correlated with PSOPTP (r2=0.55-0.79, P<0.01). We concluded that the combination of chemical P plus manure is an effective way to promote SOC accumulation and the percentage of soil Olsen-P to total P at the two mono-cropping system sites in northern China.