2  Experimental design and climate condition

Code 
#pkg 
library(dplyr)
library(ggplot2)
library(statgenHTP) #remotes::install_github("Biometris/statgenHTP", ref = "develop", dependencies = TRUE)
library(SpATS)
library(tidyr)
library(patchwork)
library(readxl)
library(plotly)
library(lubridate)
library(gridExtra)
library(ggpubr)

# src 
source(here::here("src/function/card_compilation.R"))

# cosmetics
temperature_color=read_excel(here::here("data/color_palette.xlsm")) %>%
      filter(set == "greenhouse") %>%
      dplyr::select(color, treatment) %>%
      pull(color) %>%
      setNames(read_excel(here::here("data/color_palette.xlsm")) %>%
                 filter(set == "greenhouse") %>%
                 pull(treatment)
               )

my_theme=theme_classic()+theme(
  plot.title = element_text(size = 14, face = "bold"),
  axis.title = element_text(size = 12),
  axis.text = element_text(size = 10),
  legend.text = element_text(size = 10),
  panel.grid.major = element_blank(),
  panel.grid.minor = element_blank()
)

2.1 Experimental design

Two French genotypes of soybean (Glycine max L.) belonging to the same maturity group I were selected for their similar root or aerial biomasses and their contrasted root system architecture: Stocata (obtained from RAGT Semences) and Wendy (obtained from Caussade Semences) (Maslard et al. 2021). Seeds were calibrated and pre-germinated for 4 days at 20 °C in a Fitoclima S600 germinator (Aralab, Rio de Mouro, Portugal) before transplantation. Seedlings were grown in RhizoTubes® (Figure 7.1) filled with sandy soil, allowing the visualization of the root system (Chapter 7). The soil was collected from a grassland at Flammerans near Dijon (Burgundy). Plants were grown under controlled conditions, in the Plant Phenotyping Platform for Plant and Microorganism Interaction (4PMI) at INRAE in Dijon (France) (47°32’N, 5°02’E) during 23 days in RhizoTube® (Figure 2.1). Two seedlings were transplanted in each of the 144 RhizoTubes®, in two different identical greenhouses (Figure 2.4). Two plants grew in each RhizoTubes® and were destined for four different types of analysis (Figure 2.2). At the same time, each seedling was inoculated with 1mL of Bradyrhizobium diazoefficiens corresponding to 108 rhizobia.

Figure 2.1: Experimental design of the experiment
Figure 2.2: Harvest diagram for the experiment. RT for Rhizotube. Harvest 1 took place 5 days after sowing. Harvest 2 was carried out on the day after sowing. For harvest 1, there are two plants of the same genotype per rhizotube (RT0). For each plant, ecophysiological measurements and the content of elements in each compartment were conducted. For harvest 2, in each rhizotube (RT1, RT2), there were two plants of the same genotype that have undergone the same stress conditions. Ecophysiological measurements and the element content were performed on one of the two plants. The other plant was designated for molecular analyses. Transcriptomic measurements were conducted on a pool of two roots from two plants from two different RhizoTubes (RT1 and RT2) of the same genotype having received the same stress conditions. Metabolomic analyses of leaves, stems, and roots were performed on the same plant combination. The soil near this plant was collected for metabolomic analyses, pH analyses, and microbial community analyses (Rhizosphere) (no assembly of these soils). The soil between these two plants was used for metabolomic analyses as well as for microbial community analyses (Bulk soil). During this experiment, other RhizoTubes (RT3) also containing two plants of the same genotype were in the same greenhouse. One was designated for microbial community analyses, the other for metabolomic measurements of exudates. On the plant used for the root exudates, ecophysiological measurements were also conducted. Root system architecture (RSA) and total evapotranspiration were measured on all RhizoTubes during the experiment. WW: Well Watered; WS: Water Stress; OT: Optimal temperature; HS: Heat Stress.

Air relative humidity was controlled to 45% and the photoperiod was set to 16h thanks to an artificial lighting (PAR of 272µmol.m-2. s-1) supplied with sodium lamps (400W lamp, HPS Plantastar, OSRAM, Munich, Germany). The PAR data as well as the other climatic data (humidity, temperature) are presented below (Figure 2.3).

Every Rhizotube® was autonomously weighed both before and after each watering session, if applicable, three times a day. This process was conducted to establish the evolving water evapotranspiration (ml)(Figure 3.2) and precise solution quantities (Chapter 3). The cumulative evapotranspiration during experiment was computed by summing and halving (accounting for the two plants within each Rhizotube®) the individual daily differences in Rhizotube® weight between consecutive watering events, or weight variation (Chapter 3). A N-free nutrient solution as described in (Voisin et al. 2003) was applied 3 times a day to allow 20% SWC (80% of field capacity) of water content in the RhizoTube® (Figure 3.1). Mean day/night temperatures were 32°C/24°C in two greenhouses (Figure 2.3).

Following the first week of growth, 12 plants per genotype were harvested (Figure 2.2). Subsequently, we implemented a factorial design consisting of eight different conditions (Sto_WW_OT, Sto_WS_OT, Sto_WW_HS, Sto_WS_HS, Wen_WW_OT, Wen_WS_OT, Wen_WW_HS, Wen_WS_HS) involving the two genotypes, Stocata (Sto) and Wendy (Wen), and four climatic conditions, composed of two stress factors. One factor was related to watering, with half of plants being subjected to either well-watered (WW) conditions with a SWC of 20 percent (field capacity of 80%) soil water content of 20%, or water stress (WS) conditions achieved by withholding water to reach 10% soil water content (Figure 3.1). The second factor was related to temperature. One greenhouse was maintained at an optimal temperature (OT) of 32°C/24°C (day/night), while the other greenhouse experienced two progressive heat waves of 40°C for 3 days (Figure 2.3 A), representing heat stress conditions (HS). Five days after sowing some plants were harvested for physiological analysis (including ionomic analysis) (H1). Then 20 days after sowing, (after the period of stress) plant were harvest (H2) and each organ was collected separately (leaves, stems and roots).

Four different types of analysis were performed:

  1. Eight plants per condition were dedicated to ecophysiological analyses (Chapter 3) including ionomics (Chapter 10) (8 plants per condition (only 6 plants over the 8 plants were used for ionomic analysis).
  2. Eight plants per condition were dedicated to transcriptomic and metabolomic analyses. After being harvested in liquid nitrogen, four replicates of two plants each were pooled together for each of the four treatments, for each genotype and each plant compartment (leaf, stem, and root). These pooled samples were then ground for analysis. Only root samples were analysed for transcriptomic analysis (?sec-rnaseq) and all three compartments were analyzed for metabolomics (?sec-metabo). Plant soil located close to or far from the roots was collected for metabolomics (?sec-metabo) analyses as well as for microbial community (?sec-mcom) and pH analyses.
  3. Five plants per condition were dedicated to exudate collection and exudation analyses by metabolomics (?sec-metabo) and to biomass measurement of each compartment.
  4. Five plants per condition were dedicated to the analysis of microbial communities (fungi or bacteria) in different compartments (phylloplane, leaf endosphere, root endosphere, and rhizoplane) (?sec-mcom)

Plants A and B were harvested from the same rhizotube. Plants C and D were harvested from the same rhizotube. All plants were imaged three times a week to measure aerial and root architecture.

2.2 Climate in the greenhouse

Code 
#temperature compilation
data_global_climat_select=climat_function()
data_global_climat_select_moy=mean_interval_function(data_global_climat_select,5,"min")

#graphique
ggplotly(ggplot(data_global_climat_select_moy,aes(x=as.POSIXct(Date_Time),y=Hygrometry,col=CardID))+geom_line()+#geom_point(size=0.5)+
           facet_grid(~Unit))

# Specify cutting interval in minutes (e.g. 60 minutes)
interval_minutes <- 60

data_mean <- data_global_climat_select_moy %>%
  filter(CardID!="ARIA") %>% 
  mutate(Date_Time =as.POSIXct(Date_Time, format="%Y-%m-%d %H:%M")) %>%
  filter(Date_Time>as.POSIXct("2021-09-22 10:00:00")& Date_Time<as.POSIXct("2021-10-14 00:00:00")) %>% 
  mutate(interval_id = cut(
    as.POSIXct(Date_Time),
    breaks = seq(min(Date_Time), max(Date_Time) + interval_minutes * 60, by = interval_minutes * 60),
    labels = FALSE,
    include.lowest = TRUE
  )) %>% 
  dplyr::group_by(Unit, interval_id) %>%
  dplyr::summarise(mean_temp = mean(Temperature, na.rm = TRUE),
            mean_humidity = mean(Hygrometry, na.rm = TRUE),
            mean_PAR = mean(PAR, na.rm = TRUE)) %>%
  ungroup() 

plot_mean_temperature=data_mean %>% 
ggplot(aes(x = interval_id/24, y = mean_temp, color = Unit)) +
  geom_line() +
  labs(x = "Day", y = "Average temperature (°C)", color = "Unit") +
  scale_fill_manual(values=temperature_color)+
  scale_color_manual(values=temperature_color)+
  my_theme 
plot_mean_temperature

plot_mean_humidity=data_mean %>% 
  ggplot(aes(x = interval_id/24, y = mean_humidity, color = Unit)) +
  geom_line() +
  labs(x = "Day", y = "Average humidity (%)", color = "Unit") +
  scale_y_continuous(labels = scales::percent_format(scale = 100))+
  scale_fill_manual(values=temperature_color)+
  scale_color_manual(values=temperature_color)+
  my_theme
plot_mean_humidity

plot_mean_PAR=data_mean %>% 
  ggplot(aes(x = interval_id/24, y = mean_PAR, color = Unit)) +
  geom_line() +
  labs(x = "Day", 
       y=expression(atop("Photosynthetically active",paste( "radiation ", "(µmol/ ",m^-2," /",s^-1,")"))),
       color = "Unit") +
  scale_fill_manual(values=temperature_color)+
  scale_color_manual(values=temperature_color)+
  my_theme
plot_mean_PAR

combined_plot=ggarrange(plot_mean_temperature,plot_mean_humidity,plot_mean_PAR,ncol=1, nrow=3,common.legend = TRUE, legend="right",labels = c("A","B","C"),align='v')

# Create a title and subtitle
combined_plot_title=annotate_figure(
  annotate_figure(combined_plot,
                  top=text_grob("The measurement represents an average each hour derived from four sensors positioned at \n various locations within the unit for each unit.", size = 12,hjust=0,x=0.1,y=0.4),
  ),
  top=text_grob("Temperature, humidity and PAR during the experiment",face = "bold", size = 14,hjust=0,x=0.1,y=0.4)
) ; combined_plot_title

# export
ggsave(here::here("report/intro/plot/combined_plot.svg"), combined_plot_title, width = 21, height = 25, units = "cm")
Figure 2.3: Temperature (A), humidity (B) and PAR (C) during the experiment. The measurement represents an average each hour derived from four sensors positioned at various locations within the unit for each unit.
Bonnus: Correction for spatial trends over time