Mapping the root systems of individual trees in a natural community using genotyping-by-sequencing

The architecture of root systems is an important driver of plant fitness, competition and ecosystem processes. However, the methodological difficulty of mapping roots hampers the study of these processes. Existing approaches to match individual plants to belowground samples are low-throughput and species-specific. Here, we developed a scalable sequencing-based method to map the root systems of individual trees across multiple species. We successfully applied it to a tropical dry forest community in the Brazilian Caatinga containing 14 species. We sequenced all 42 individual shrubs and trees in a 14 by 14 m plot using double-digest restriction-site associated sequencing (ddRADseq). We identified species-specific markers and individual-specific haplotypes from the data. We matched these markers to ddRADseq data from 100 mixed root samples from across the centre (10 by 10 m) of the plot at four different depths, using a newly developed R package. We identified individual root samples for all species and all but one individual. There was a strong significant correlation between below and aboveground size measurements, and we also detected significant species-level root-depth preference for two species. The method is more scalable and less labour-intensive than current techniques, and is broadly applicable to ecology, forestry and agricultural biology.

根系构型是影响植物适合度、种间竞争以及生态系统过程的关键驱动因素。然而，根系定位的方法学难题阻碍了相关研究的推进。现有将单株植物与地下样本进行匹配的方法存在通量低且仅适用于特定物种的局限。本研究开发了一种可扩展的基于测序的方法，可实现多物种下单株树木根系的定位。研究团队将该方法成功应用于巴西卡廷加（Caatinga）热带旱林群落，该群落包含14个物种。研究人员利用双酶切限制性位点关联测序（double-digest restriction-site associated sequencing, ddRADseq）技术，对14m×14m样地内的全部42株灌木与乔木进行了测序。从测序数据中筛选得到物种特异性标记与个体特异性单倍型。研究团队借助自主开发的R软件包，将上述标记与取自样地中心10m×10m区域、四个不同土层的100份混合根系样本的ddRADseq数据进行匹配。最终为所有物种的根系样本匹配到了对应单株，仅1株个体未能完成匹配。分析结果显示，地下与地上部分的生物量指标存在显著强相关；同时研究还发现2个物种存在显著的物种水平根系深度偏好性。相较于现有技术，本方法具有更高的可扩展性与更低的人力投入成本，可广泛应用于生态学、森林学以及农业生物学研究领域。