Data from: Changes in plant architecture in Brazilian peppertree damaged by the biological control agent, <i>Pseudophilothrips ichini</i> Hood (Thysanoptera: Phlaeothripidae)

These data come from a roughly 4-year study on the growth response of a highly invasive woody plant to the damage induced by its biological control agent (an insect) and to soil fertilizer levels. The study design was an interrupted time series in which data were collected from plants for 2 years ("pre"), agents were introduced, and data collection continued for 2 years with sustained agent releases ("post"). Each month, stem tip counts and other measurements were collected from roughly 200 plants: 100 in each of two garden plots (IPRL and UF/IRREC). As plants grew larger (eventually to ~250 cm tall), data collection was spaced to roughly every 2 months and some replicates within each treatment were dropped. In addition to the "interruption" of introducing the agents, the other treatment was the addition of various concentrations of liquid fertilizer to the soil throughout the study. While there were no control plants that did not receive insects, each plant was followed over the course of the pre-post design experiment and Bayesian mixed modelling was used to interpret the effect of releasing the insect on how plant growth parameters changed over time and under different fertilizer levels. We have included raw data in .csv and .xlsx formats, one set for each garden plot as plots were analyzed separately. The Excel files have notes in the header column providing more information about each variable. However, not all variables were analyzed in the published study. The R code is also attached, which runs the Bayesian analyses for each plot and each response variable of interest in the published study. Raw data are also provided for soil and foliar nitrogen content. Soil nitrogen was analyzed at both plots, but foliar nitrogen was only analyzed at one of the plots (IPRL).本数据集源自一项为期约4年的研究，聚焦于一种高度入侵木本植物对其生物防治媒介（biological control agent，一种昆虫）造成的损伤以及土壤肥料水平的生长响应。该研究采用间断时间序列设计（interrupted time series）：先对植株开展为期2年的数据采集（即"前干预阶段"），随后引入生物防治媒介，之后持续释放媒介并继续采集数据2年（即"后干预阶段"）。研究人员每月对约200株植株采集茎尖计数及其他相关测量数据：两个试验小区（IPRL与UF/IRREC）各设100株供试植株。随着植株生长至最终株高约250厘米，数据采集间隔调整为约每2个月一次，同时剔除了各处理组内的部分重复样本。除引入防治媒介这一"干预环节"外，另一处理变量为在整个研究期间向土壤施加不同浓度的液态肥料。本研究未设置未接触昆虫的对照植株，但所有植株均贯穿前后干预阶段的实验设计，并采用贝叶斯混合模型（Bayesian mixed modelling）解析释放昆虫对不同肥料水平下植株生长参数随时间变化的影响。数据集包含.csv与.xlsx格式的原始数据，两个试验小区分别单独分析，故各对应一套数据文件。Excel文件的表头列中附有各变量的详细说明，但并非所有变量均已在已发表研究中得到分析。附件中还包含R语言代码，可针对已发表研究中的各试验小区及目标响应变量运行贝叶斯分析。研究同时提供了土壤与叶片氮含量的原始数据：两个试验小区均开展了土壤氮含量检测，但叶片氮含量仅在IPRL小区完成检测。