GK Lab slides – slides

Belowground regulation of
aboveground dynamics

Gaurav Kandlikar
Louisiana State University, Baton Rouge
slides: https://talks.gklab.org/tulane-25

Sal (Shorea) dominated forests in northern India, Photo by Dhritiman Mukherjee

Intercropping of soybean and wheat; image from U. Iowa

“Christmas tree scientists work to manage grinchy fungal foes”; imaege from Washington State Univ

Research goals

Explain patterns of diversity in nature
Predict responses to perturbations

Approach

Approach

Collaboration, caring, and equity

Today’s talk

Ecological theory for microbial effects on plant coexistence
Microbial effects on plant communities in variable environments

Mathematical underpinning of plant–soil feedback

See also Bever, Westover, and Antonovics (1997) and Kandlikar (2024)

Plant dynamics

where is the relative frequency of microbial community in the soil (e.g. )

Interpretation: Plants have exponential population growth at a rate determined by the composition of the soil community.

Microbe dynamics

Interpretation: Soil communities grow at a rate determined by the frequency of each plant (and the relative strength of each plant’s conditioning effect, ).

R code to simulate model dynamics

Code

psf_model <- function(time, init, params) {
  with (as.list(c(time, init, params)), {
    # description of parameters (see Bever et al. 1997)
    # N1 and N2: abundance of the plant species 1 and 2
    # p1: frequency of plant species 1; p2 = 1-pA
    # m1A, m1B, m2A, m2B: conspecific and heterospecific effects of microbial community A or B on the growth of plant 1 or 2
    # pA: frequency of the soil microbial community A
    # v: influence of plant species 2 on the microbial community relative to that of plant 1
    
    # Differential equations
    dN1 <- (m1A*pA + m1B*(1-pA))*N1
    dN2 <- (m2A*pA + m2B*(1-pA))*N2
    dp1 <- p1*(1-p1)*((m1A-m2A)*pA + (m1B-m2B)*(1-pA))
    dpA <- pA*(1-pA)*(p1-v*(1-p1))
    
    
    # Return dN1 and dN2
    return(list(c(dN1, dN2, dp1, dpA)))
  })
}

Lasting influence of the Bever model

Code

# Dataset downloaded from Figshare: 
# curl::curl_download("https://figshare.com/ndownloader/files/14874749", destfile = "../crawford-supplement.xlsx")

read_excel("../data/crawford-supplement.xlsx", sheet = "Data") |> 
  mutate(stab = -0.5*rrIs) |> # show in terms of stabilization instead of I_s
  filter(stab > -3) |> # Remove the pair with Stab approx. 6 -- according to McCarthy Neuman (original author), this is likely due to bad seed survival
  ggplot(aes(x = stab)) + 
  geom_histogram(color = "black") +
  geom_histogram(fill = "#56A0D3") +
  geom_rug(color = "grey25") + 
  geom_vline(xintercept = 0) + 
  annotate("text", x = Inf, y = Inf, label = "N = 1037 pairwise comparisons\n
           57% experience positive stabilization",
           vjust=1, hjust = 1, size = 6) +
  xlab("Stabilization") + ylab("Count")

Data from Crawford et al. (2019)

A puzzle: same stabilization, different model dynamics

Code



# Define a function used for making the PSF framework
# schematic, given a set of parameter values
make_params_plot <- function(params, scale = 1.5) {
  
  color_func <- function(x) {
    ifelse(x < 0, "#EE6677", "#4477AA")
  }
  df <- data.frame(x = c(0,0,1,1),
                   y = c(0,1,0,1),
                   type = c("M", "P", "M", "P"))
  
  params_plot <- 
    ggplot(df) +
    geom_label(x = 0, y = 1.1, size = 3.25, label = "Plant 1", color = "white", fill = "#9970ab", fontface = "bold") + 
    geom_label(x = 1, y = 1.1, size = 3.25, label = "Plant 2", color = "white", fill = "#5aae61", fontface = "bold") +
    geom_label(x = 0, y = -0.15, size = 3.25, label = "Soil\nmicrobes A", label.size = 0, fill = "transparent") + 
    geom_label(x = 1, y = -0.15, size = 3.25, label = "Soil\nmicrobes B", label.size = 0, fill = "transparent") +
    geom_label(x = 0.45, y = -0.4, size = 3.5, fill = "transparent",
               label.size = 0.05,
               label = TeX(paste0("Stabilization = ",
                                  0.5*(params["m1B"] + params["m2A"] - params["m1A"] - params["m2B"])))) +
    geom_segment(aes(x = 0, xend = 0, y = 0.1, yend = 0.9),
                 arrow = arrow(length = unit(0.03, "npc")),
                 linewidth = abs(params["m1A"])*scale, 
                 color = alpha(color_func(params["m1A"]), 1)) +
    geom_segment(aes(x = 0.05, xend = 0.95, y = 0.1, yend = 0.9),
                 arrow = arrow(length = unit(0.03, "npc")),
                 linewidth = abs(params["m2A"])*scale, 
                 color = alpha(color_func(params["m1B"]),1)) +
    geom_segment(aes(x = 0.95, xend = 0.05, y = 0.1, yend = 0.9),
                 arrow = arrow(length = unit(0.03, "npc")),
                 linewidth = abs(params["m1B"])*scale, 
                 color = alpha(color_func(params["m2A"]), 1)) +
    geom_segment(aes(x = 1, xend = 1, y = 0.1, yend = 0.9),
                 arrow = arrow(length = unit(0.03, "npc"),),
                 linewidth = abs(params["m2B"])*scale, 
                 color = alpha(color_func(params["m2B"]), 1)) +
    
    # Plant cultivation of microbes
    geom_segment(aes(x = -0.25, xend = -0.25, y = 0.9, yend = 0.1), linewidth = 0.15, linetype = 1,
                 arrow = arrow(length = unit(0.03, "npc"))) +
    geom_segment(aes(x = 1.25, xend = 1.25, y = 0.9, yend = 0.1), linewidth = 0.15, linetype = 1,
                 arrow = arrow(length = unit(0.03, "npc"))) +
    
    annotate("text", x = 0, y = 0.5, 
             label = TeX(paste0("m1A = ", params["m1A"])), 
             angle = 90, vjust = -0.25, size = 3) + 
    annotate("text", x = 1.15, y = 0.5, 
             label = TeX(paste0("m2B = ", params["m2B"])), 
             angle = -90, vjust = 1.5, size = 3) + 
    annotate("text", x = 0.75, y = 0.75, 
             label = TeX(paste0("m2A = ", params["m2A"])), 
             angle = 45, vjust = -0.25, size = 3) + 
    annotate("text", x = 0.25, y = 0.75, 
             label = TeX(paste0("m1B = ", params["m1B"])), 
             angle = -45, vjust = -0.25, size = 3) + 
    xlim(c(-0.4, 1.4)) + 
    coord_cartesian(ylim = c(-0.25, 1.15), clip = "off") + 
    theme_void() +
    theme(legend.position = "none",
          plot.caption = element_text(hjust = 0.5, size = 10))
  return(params_plot)
}

# Define a function for simulating the dynamics of the 
# PSF model with deSolve
params_coex <- c(m10 = 0.16, m20 = 0.16,
                 m1A = 0.11, m1B = 0.26, 
                 m2A = 0.27, m2B = 0.13, v = 1)

time <- seq(0,50,0.1)
init_pA_05 <- c(N1 = 3, N2 = 7, p1 = 0.3, pA = 0.3)
out_pA_05 <- ode(y = init_pA_05, times = time, func = psf_model, parms = params_coex) |> data.frame()


params_to_plot_coex <- c(m1A = unname(params_coex["m1A"]-params_coex["m10"]),
                    m1B = unname(params_coex["m1B"]-params_coex["m10"]),
                    m2A = unname(params_coex["m2A"]-params_coex["m20"]),
                    m2B = unname(params_coex["m2B"]-params_coex["m20"]), v=1)
param_plot_coex <- 
  make_params_plot(params_to_plot_coex, scale = 6) + 
  annotate("text", x = 1.375, y = 0.5, 
           label = paste0("v = ", params_coex["v"]), 
           angle = -90, vjust = 1.5, size = 3)  

panel_abund_coex <- 
  out_pA_05 |> 
  as_tibble() |> 
  select(time, N1, N2) |> 
  pivot_longer(N1:N2) |>
  ggplot(aes(x = time, y = value, color = name)) +
  geom_line(linewidth = 0.9) + 
  scale_color_manual(values = c("#9970ab", "#5aae61"),
                     name = "Plant species", label = c("Plant 1", "Plant 2"), guide = "none") +
  scale_y_continuous(breaks = c(2e4, 4e4, 6e4, 8e4), labels = scales::scientific) +
  ylab("Abundance") +
  theme(axis.title = element_text(size = 10))

# Panel D: plot for frequencies of both plants when growing 
# in dynamic soils
panel_freq_coex <- 
  out_pA_05 |> 
  as_tibble() |> 
  mutate(p2 = 1-p1) |> 
  select(time, p1, p2) |> 
  pivot_longer(p1:p2) |>
  ggplot(aes(x = time, y = value, color = name)) +
  geom_line(linewidth = 1.2) + 
  scale_color_manual(values = c("#9970ab", "#5aae61"),
                     name = "Plant species", label = c("Plant 1", "Plant 2")) + 
  scale_y_continuous(limits = c(0,1), breaks = c(0, 0.5, 1)) + 
  ylab("Frequency") +
  theme(axis.title = element_text(size = 10))

param_plot_coex +  {panel_abund_coex+panel_freq_coex &
  theme(axis.text = element_text(size = 8), legend.position = 'none')} + 
  plot_layout(widths = c(1/3, 2/3))

Code

# Define a function for simulating the dynamics of the 
# PSF model with deSolve
params <- c(m10 = 0.16, m20 = 0.16,
              m1A = 0.02, m2A = 0.33, 
              m1B = 0.18, m2B = 0.20, v = 1)

time <- seq(0,200,0.1)
init_pA_05 <- c(N1 = 3, N2 = 7, p1 = 0.3, pA = 0.3)
out_pA_05 <- ode(y = init_pA_05, times = time, func = psf_model, parms = params) |> data.frame()


params_to_plot <- c(m1A = unname(params["m1A"]-params["m10"]),
                    m1B = unname(params["m1B"]-params["m10"]),
                    m2A = unname(params["m2A"]-params["m20"]),
                    m2B = unname(params["m2B"]-params["m20"]), v=1)
param_plot <- 
  make_params_plot(params_to_plot, scale = 6) + 
  annotate("text", x = 1.375, y = 0.5, 
           label = paste0("v = ", params["v"]), 
           angle = -90, vjust = 1.5, size = 3)  

panel_abund <- 
  out_pA_05 |> 
  as_tibble() |> 
  select(time, N1, N2) |> 
  pivot_longer(N1:N2) |>
  ggplot(aes(x = time, y = value, color = name)) +
  geom_line(linewidth = 0.9) + 
  scale_color_manual(values = c("#9970ab", "#5aae61"),
                     name = "Plant species", label = c("Plant 1", "Plant 2"), guide = "none") +
    scale_y_continuous(breaks = c(0,8e17,1.6e18), labels = c("0", "7e17","1.4e18")) +
  ylab("Abundance") 

# Panel D: plot for frequencies of both plants when growing 
# in dynamic soils
panel_freq <- 
  out_pA_05 |> 
  as_tibble() |> 
  mutate(p2 = 1-p1) |> 
  select(time, p1, p2) |> 
  pivot_longer(p1:p2) |>
  ggplot(aes(x = time, y = value, color = name)) +
  geom_line(linewidth = 1.2) + 
  scale_color_manual(values = c("#9970ab", "#5aae61"),
                     name = "Plant species", label = c("Plant 1", "Plant 2")) + 
  scale_y_continuous(limits = c(0,1), breaks = c(0, 0.5, 1)) + 
  ylab("Frequency")

param_plot + {panel_abund + panel_freq & 
  theme(axis.text = element_text(size = 8), legend.position = 'none')} +
  plot_layout(widths = c(1/3, 2/3))

How can we more accurately infer predicted plant dynamics under plant–soil feedback?

Build on insights from coexistence theory (Chesson 2018):

Stablization is necessary for coexistence but doesn’t guarantee coexistence
Stable coexistence is possible only when stabilization overcomes competitive imbalances (fitness differences)

How can we more accurately infer predicted plant dynamics under plant–soil feedback?

How can we more accurately infer predicted plant dynamics under plant–soil feedback?

Code

base <-
  tibble(sd = seq(-1,1,0.1),
       fd = c(seq(1,0,-0.1), seq(0.1,1,0.1))) |> 
  ggplot(aes(x = sd, y = fd)) + 
  geom_line(linetype = "dashed") +
  geom_hline(yintercept = 0, linewidth = 1) + 
  geom_vline(xintercept = 0, linewidth = 1) + 
  xlab("Stabilization") + 
  ylab("Fitness difference") +
  theme(axis.text = element_blank(),
        axis.ticks = element_blank(),
        axis.line = element_blank(),
        axis.title = element_text(size = 25)) + 
  scale_x_continuous(expand = c(0,0)) + 
  scale_y_continuous(expand = c(0.04,0)) 
base

Code

base <- 
base + 
  annotate("text", x = Inf, y = 0, hjust = 1, vjust = -1, label = "Coexistence", size = 7, family = "italic") + 
  annotate("text", x = -Inf, y = 0, hjust = 0, vjust = -1, label = "Priority effects", size = 7, family = "italic") + 
  annotate("label", x = 0, y = Inf,  vjust = 1.5, label = "Species exclusion", size = 7, fill = "#F0F1EB",family = "italic") 

base

Code

metrics <- function(params) {
  IS <- with(as.list(params), {m1A - m2A - m1B + m2B})
  FD <- with(as.list(params), {(1/2)*(m1A+m2A) - (1/2)*(m1B+m2B)})
  SD <- (-1/2)*IS
  return(c(IS = IS, SD = SD, FD = FD))
}

coex_metrics <- metrics(params_coex)
base_coex <- 
  base + 
  inset_element({panel_freq_coex + theme(legend.position = 'none',
                                         axis.text = element_blank(),
                                         axis.ticks = element_blank(),
                                         axis.title = element_blank(),
                                         plot.background = element_rect(color = "grey"))},
                0.75,0.15,1,0.5)
base_coex

Code

base_coex + 
  inset_element({panel_freq + theme(legend.position = 'none',
                                    axis.text = element_blank(),
                                    axis.ticks = element_blank(),
                                    axis.title = element_blank(),
                                    plot.background = element_rect(fill = "#F0F1EB", color = 'grey'))},
                0.375,0.4,.625,0.75)

Do microbes generate fitness differences in nature?

Sedgwick Reserve (unceded territory of Chumash people)

Code

base <- 
  base +
  ylim(-0.25,1.35) 
base

Code

# Example point, using UR_FE values
base +
  annotate("pointrange", x = 0.275, y = 0.0591, ymin = 0.0591-0.307,ymax = 0.0591+0.307, shape = 21, stroke = 1.5, size = 1, linewidth = 0.1, fill = "#4477aa") + 
    annotate("pointrange", x = 0.275, y = 0.0591, xmin = 0.275-0.152,xmax = 0.275+0.152, shape = 21, fill = "#4477aa", linewidth = 0.1, size = 1, stroke = 1.5) + 
    annotate("text", x = 0.275 + 0.3, y = 0.0591 + 0.15, color = "#4477aa", label = "Festuca - Plantago", size = 8, fontface = "italic") +
  theme(axis.title = element_blank())

Code

# Dataset downloaded from Figshare: 
# curl::curl_download("https://datadryad.org/downloads/file_stream/367188", destfile = "../data/kandlikar2021-supplement.csv")

biomass_wide <- 
  read_csv("../data/kandlikar2021-supplement.csv") |> 
  mutate(log_agb = log(abg_dry_g)) |> 
  mutate(source_soil = ifelse(source_soil == "aastr", "str", source_soil),
                source_soil = ifelse(source_soil == "abfld", "fld", source_soil),
                pair = paste0(source_soil, "_", focal_species)) |> 
      select(replicate, pair, log_agb)  |> 
    pivot_wider(names_from = pair, values_from = log_agb) |> 
  filter(!is.na(replicate))

# Calculate the Stabilization between each species pair ----
# Recall that following the definition of the m terms in Bever 1997,
# and the analysis of this model in Kandlikar 2019,
# stablization = -0.5*(log(m1A) + log(m2B) - log(m1B) - log(m2A))
# Here is a function that does this calculation 
# (recall that after the data reshaping above, 
# each column represents a given value of log(m1A))
calculate_stabilization <- function(df) {
  df |> 
    # In df, each row is one repilcate/rack, and each column
    # represents the growth of one species in one soil type.
    # e.g. "FE_ACWR" is the growth of FE in ACWR-cultivated soil.
    mutate(AC_FE = -0.5*(AC_ACWR - AC_FEMI - FE_ACWR + FE_FEMI),
           AC_HO = -0.5*(AC_ACWR - AC_HOMU - HO_ACWR + HO_HOMU),
           AC_SA = -0.5*(AC_ACWR - AC_SACO - SA_ACWR + SA_SACO),
           AC_PL = -0.5*(AC_ACWR - AC_PLER - PL_ACWR + PL_PLER),
           AC_UR = -0.5*(AC_ACWR - AC_URLI - UR_ACWR + UR_URLI),
           FE_HO = -0.5*(FE_FEMI - FE_HOMU - HO_FEMI + HO_HOMU),
           FE_SA = -0.5*(FE_FEMI - FE_SACO - SA_FEMI + SA_SACO),
           FE_PL = -0.5*(FE_FEMI - FE_PLER - PL_FEMI + PL_PLER),
           FE_UR = -0.5*(FE_FEMI - FE_URLI - UR_FEMI + UR_URLI),
           HO_PL = -0.5*(HO_HOMU - HO_PLER - PL_HOMU + PL_PLER),
           HO_SA = -0.5*(HO_HOMU - HO_SACO - SA_HOMU + SA_SACO),
           HO_UR = -0.5*(HO_HOMU - HO_URLI - UR_HOMU + UR_URLI),
           SA_PL = -0.5*(SA_SACO - SA_PLER - PL_SACO + PL_PLER),
           SA_UR = -0.5*(SA_SACO - SA_URLI - UR_SACO + UR_URLI),
           PL_UR = -0.5*(PL_PLER - PL_URLI - UR_PLER + UR_URLI)) |>
    select(replicate, AC_FE:PL_UR) |> 
    gather(pair, stabilization, AC_FE:PL_UR)
}

stabilization_values <- calculate_stabilization(biomass_wide)
# Seven values are NA; we can omit these 
stabilization_values <- stabilization_values |> 
  filter(!(is.na(stabilization)))

# Calculating the Fitness difference between each species pair ----
# Similarly, we can now calculate the fitness difference between
# each pair. Recall that FD = 0.5*(log(m1A)+log(m1B)-log(m2A)-log(m2B))
# But recall that here, IT IS IMPORTANT THAT
# m1A = (m1_soilA - m1_fieldSoil)!

# The following function does this calculation:
calculate_fitdiffs <- function(df) {
  df |> 
    mutate(AC_FE = 0.5*((AC_ACWR-fld_ACWR) + (FE_ACWR-fld_ACWR) - (AC_FEMI-fld_FEMI) - (FE_FEMI-fld_FEMI)),
           AC_HO = 0.5*((AC_ACWR-fld_ACWR) + (HO_ACWR-fld_ACWR) - (AC_HOMU-fld_HOMU) - (HO_HOMU-fld_HOMU)),
           AC_SA = 0.5*((AC_ACWR-fld_ACWR) + (SA_ACWR-fld_ACWR) - (AC_SACO-fld_SACO) - (SA_SACO-fld_SACO)),
           AC_PL = 0.5*((AC_ACWR-fld_ACWR) + (PL_ACWR-fld_ACWR) - (AC_PLER-fld_PLER) - (PL_PLER-fld_PLER)),
           AC_UR = 0.5*((AC_ACWR-fld_ACWR) + (UR_ACWR-fld_ACWR) - (AC_URLI-fld_URLI) - (UR_URLI-fld_URLI)),
           FE_HO = 0.5*((FE_FEMI-fld_FEMI) + (HO_FEMI-fld_FEMI) - (FE_HOMU-fld_HOMU) - (HO_HOMU-fld_HOMU)),
           FE_SA = 0.5*((FE_FEMI-fld_FEMI) + (SA_FEMI-fld_FEMI) - (FE_SACO-fld_SACO) - (SA_SACO-fld_SACO)),
           FE_PL = 0.5*((FE_FEMI-fld_FEMI) + (PL_FEMI-fld_FEMI) - (FE_PLER-fld_PLER) - (PL_PLER-fld_PLER)),
           FE_UR = 0.5*((FE_FEMI-fld_FEMI) + (UR_FEMI-fld_FEMI) - (FE_URLI-fld_URLI) - (UR_URLI-fld_URLI)),
           HO_PL = 0.5*((HO_HOMU-fld_HOMU) + (PL_HOMU-fld_HOMU) - (HO_PLER-fld_PLER) - (PL_PLER-fld_PLER)),
           HO_SA = 0.5*((HO_HOMU-fld_HOMU) + (SA_HOMU-fld_HOMU) - (HO_SACO-fld_SACO) - (SA_SACO-fld_SACO)),
           HO_UR = 0.5*((HO_HOMU-fld_HOMU) + (UR_HOMU-fld_HOMU) - (HO_URLI-fld_URLI) - (UR_URLI-fld_URLI)),
           SA_PL = 0.5*((SA_SACO-fld_SACO) + (PL_SACO-fld_SACO) - (SA_PLER-fld_PLER) - (PL_PLER-fld_PLER)),
           SA_UR = 0.5*((SA_SACO-fld_SACO) + (UR_SACO-fld_SACO) - (SA_URLI-fld_URLI) - (UR_URLI-fld_URLI)),
           PL_UR = 0.5*((PL_PLER-fld_PLER) + (UR_PLER-fld_PLER) - (PL_URLI-fld_URLI) - (UR_URLI-fld_URLI))) |>
    select(replicate, AC_FE:PL_UR) |> 
    gather(pair, fitdiff_fld, AC_FE:PL_UR)
}
fd_values <- calculate_fitdiffs(biomass_wide)
# Twenty-two values are NA; let's omit these.
fd_values <- fd_values |> filter(!(is.na(fitdiff_fld)))

# Now, generate statistical summaries of SD and FD

stabiliation_summary <- stabilization_values |> group_by(pair) |>
  summarize(mean_sd = mean(stabilization),
            sem_sd = sd(stabilization)/sqrt(n()),
            n_sd = n())

fitdiff_summary <- fd_values |> group_by(pair) |>
  summarize(mean_fd = mean(fitdiff_fld),
            sem_fd = sd(fitdiff_fld)/sqrt(n()),
            n_fd = n())

# Combine the two separate data frames.
sd_fd_summary <- left_join(stabiliation_summary, fitdiff_summary)

# Some of the FDs are negative, let's flip these to be positive
# and also flip the label so that the first species in the name
# is always the fitness superior.
sd_fd_summary <- sd_fd_summary |> 
  # if mean_fd is < 0, the following command gets the absolute
  # value and also flips around the species code so that
  # the fitness superior is always the first species in the code
  mutate(pair = ifelse(mean_fd < 0,
                       paste0(str_extract(pair, "..$"),
                              "_",
                              str_extract(pair, "^..")), 
                       pair),
         mean_fd = abs(mean_fd))
sd_fd_summary <- 
  sd_fd_summary |> 
  mutate(
    # outcome = ifelse(mean_fd - 2*sem_fd >
    #                    mean_sd + 2*sem_sd, "exclusion", "neutral"),
    outcome2 = ifelse(mean_fd > mean_sd, "exclusion", "coexistence"),
    outcome2 = ifelse(mean_sd < 0, "exclusion or priority effect", outcome2)
)

base +
  geom_pointrange(data = sd_fd_summary, 
                  aes(x = mean_sd, y = mean_fd,
                      ymin = mean_fd-sem_fd,
                      ymax = mean_fd+sem_fd, fill = outcome2), size = 1, linewidth = 0.1,
                  shape = 21, stroke = 1.5) +
    geom_pointrange(data = sd_fd_summary, 
                  aes(x = mean_sd, y = mean_fd,
                      xmin = mean_sd-sem_sd,
                      xmax = mean_sd+sem_sd, fill = outcome2), size = 1, linewidth = 0.1, shape = 21, stroke = 1.5) +
  scale_fill_manual(values = c("#4477aa", "#ee6677", "#ccbb44")) + 
  theme(legend.position = "none")+ 
  theme(axis.title = element_blank())

Detailed in Kandlikar et al. (2021)

Do microbes generate fitness differences in nature?

In California grasslands, microbes tend to drive stronger fitness differences than stabilization.
Can we answer this question more generally?

Meta-analysis overview

Data for 518 pairwise comparisons of stabilization and fitness differences (results shown for 72 pairs here)

New sampling approach for quantifying uncertainty in coexistence outcomes

Results in Yan, Levine, and Kandlikar (2022); Analyses available on Zenodo

Do microbes generate fitness differences in nature?

Fitness differences represent a dominant avenue through which microbes control plant coexistence

Microbially-mediated intransitivity in pine-oak forests in Spain, Pajares-Murgó et al. (2024)

Today’s talk

Ecological theory for microbial effects on plant coexistence

2. Microbial effects on plant communities in variable environments

Today’s talk

Ecological theory for microbial effects on plant coexistence

2. Microbial effects on plant communities in variable environments

Coexistence in fragmented tropical forests
Woody encroachment in longleaf pine understories
Plant eco-evolutionary dynamics under drought

Hassan District, Karnataka, India

Gradients in soil moisture from forest edge to interior

Detailed in Krishnadas et al. (2018)

Can variable plant–soil feedback explain this result?

Quantify pairwise plant–soil feedback under high and low water conditions in a shadehouse

Exaggerated microbially mediated fitness differences in dry soils could contribute to erosion of diversity at fragment edges.

Analyses ongoing

Applications of basic ecology to habitat management

Do soil microbes contribute to invasion of coffee into forest edges?
Do soil microbes in abandoned coffee orchards affect restoration?

Applications of basic ecology to habitat management

Extremely preliminary analyses!

Coffee growth enhanced by forest soil microbes
Microbes in coffee soil suppress growth of native seedlings

Today’s talk

Ecological theory for microbial effects on plant coexistence

2. Microbial effects on plant communities in variable environments

Coexistence in fragmented tropical forests
Woody encroachment in longleaf pine understories
Plant eco-evolutionary dynamics under drought

Longleaf pine savanna understory at Palustris Experimental Forest

Across ecosystems, grasses tend to experience negative feedback

Code

# curl::curl_download("https://datadryad.org/downloads/file_stream/2789266", destfile = "../data/jiang2024-dataset.csv")

jiang_dat <- read_csv("../data/jiang2024-dataset.csv")



# jiang_metamod  <- metafor::rma.mv(rr, var, data = jiang_dat,   mods = ~ Life.form-1)
# saveRDS(jiang_metamod, "../data/jiang-metamod.rds")

metamod <- readRDS("../data/jiang-metamod.rds")

metamod_s <- broom::tidy(metamod)

jiang_dat |> 
  ggplot(aes(x = rr, y = Life.form, color = Life.form)) +
  ggbeeswarm::geom_quasirandom(shape = 21) +
  scale_color_manual(values = c("#ccbb44","#228833")) + 
  annotate("point",
           x = metamod_s$estimate[1], y = 1, size = 3, shape = 21, stroke = 1.5) +
  annotate("point",
           x = metamod_s$estimate[2], y = 2, size = 3, shape = 21, stroke = 1.5) +
  ylab("") + 
  xlab("Growth in self vs. growth in non-self") +
    xlim(-2.5,2.5)+
  geom_vline(xintercept = 0, linetype = 'dashed') + 
  theme(legend.position = "none",
        axis.text.y = element_text(size = 12, color = 'black'))

Data from Jiang et al. (2024)

Do soil microbes play a role in driving shrubification under fire suppression?

What are the direct impacts of fire on soil microbial dynamics?

Does the history and severity of fire mediate the impact of soil microbes on plant performance and coexistence?

Approach

Field experiment: evaluate fire effects on soil communities
Hoophouse experiment: evaluate shrub and grass growth in shrub- and grass-conditioned soils with different burn histories
Mathematical modeling: experimentally-informed dynamics of plant community dynamics

LSU AgCenter’s Lee Memorial Forest, FranklintonLA

Video Player is loading.

Current Time 0:00

Duration 0:25

Loaded: 80.54%

Stream Type LIVE

Remaining Time 0:25

Plot set to burn next week

How does burn severity affect biotic and abiotic soil properties?
What are the dynamics of recovery?

Mathematical modeling of plant dynamics

Goal: translate results from short-term experiments into long-term projections of community dynamics

Today’s talk

Ecological theory for microbial effects on plant coexistence

2. Microbial effects on plant communities in variable environments

Coexistence in fragmented tropical forests
Woody encroachment in longleaf pine understories
Plant eco-evolutionary dynamics under drought

Past work: Soil microbial legacies mediate plant response to drought across generations

Code

# curl::curl_download("https://datadryad.org/downloads/file_stream/50543", destfile = "../data/lau2012-dataset.csv")

lau_dat <- read_csv("../data/lau2012-dataset.csv")
lau_dat |> 
  group_by(Contemp_Water, Plant_History, Microbe_History) |> 
  filter(!is.na(Fruit_number)) |> 
  summarize(mean_fruit = mean(Fruit_number),
            sd_fruit = sd(Fruit_number),
            nreps = n(),
            sem = sqrt(sd_fruit/(nreps-1))) |> 
  mutate(Microbe_History = 
           ifelse(Microbe_History == "Dry microbes", "Dry-adapted\nmicrobes", "Wet-adapted\nmicrobes")) |> 
  filter(Plant_History == "Wet") |> # Only show wet history plants for simplicity
  ggplot(aes(x = Contemp_Water, y = mean_fruit, 
             ymin = mean_fruit-sem,
             ymax = mean_fruit+sem,
             fill = Microbe_History)) +
  geom_pointrange(shape = 21, 
                  size = 1.25,
                  linewidth = 1,
                  position = position_dodge(width = 0.25)) + 
  scale_fill_manual(name = "",
                      values = c("#cfe2f3ff", "#0b5394ff")) + 
  scale_y_continuous(#limits = c(0,6),
                     name = "Fruit number") + 
  xlab("Contemporary watering") + 
  # facet_wrap(.~Plant_History, scales = "free") +f
  theme(axis.text = element_text(size = 12, color = "black"),
        legend.text = element_text(size = 12),
        legend.position = "inside",
        legend.position.inside = c(0.8,0.2),
        legend.background = element_rect(color = "black"))

Brassica rapa’s tolerance of dought is enhanced by growing with a drought-adapted soil microbiome Data from Lau and Lennon (2012)

Ongoing/future work: What is the mechanistic basis, and how commonly do microbes shape transgenerational plant dynamics under abiotic stress?

Currently in the second generation of experimental drought, with additional plants grown to evaluate immediate maternal effects on plant functional traits

Drought imposes a fitness cost to B. rapa

Extremely preliminary analyses!

Drought imposes a fitness cost to B. rapa

Extremely preliminary analyses!

Drought imposes a fitness cost to B. rapa

Evidence of genetic variation in drought tolerance

Extremely preliminary analyses!

Currently in the second generation of experimental drought, with additional plants grown to evaluate immediate maternal effects on plant functional traits

Early indication of maternal effects

Extremely preliminary analyses!

Thanks to an incredible crew of undergrads!

Other projects in the lab

How does nutrient enrichment structure plant–mycorrhizal symbiosis?

Citizen science for invasion ecology

Thank you!

slides: https://talks.gklab.org/tulane-25
contact: gkandlikar@lsu.edu

References

Bever, James D., Kristi M. Westover, and Janis Antonovics. 1997. “Incorporating the Soil Community into Plant Population Dynamics: The Utility of the Feedback Approach.” Journal of Ecology 85 (5): 561–73. https://doi.org/10.2307/2960528.

Chesson, Peter. 2018. “Updates on Mechanisms of Maintenance of Species Diversity.” Journal of Ecology 106 (5): 1773–94.

Crawford, Kerri M, Jonathan T Bauer, Liza S Comita, Maarten B Eppinga, Daniel J Johnson, Scott A Mangan, Simon A Queenborough, et al. 2019. “When and Where Plant-Soil Feedback May Promote Plant Coexistence: A Meta-Analysis.” Ecology Letters 22 (8): 1274–84.

Jiang, Feng, Jonathan A Bennett, Kerri M Crawford, Johannes Heinze, Xucai Pu, Ao Luo, and Zhiheng Wang. 2024. “Global Patterns and Drivers of Plant–Soil Microbe Interactions.” Ecology Letters 27 (1): e14364.

Kandlikar, Gaurav S. 2024. “Quantifying Soil Microbial Effects on Plant Species Coexistence: A Conceptual Synthesis.” American Journal of Botany 111 (4): e16316.

Kandlikar, Gaurav S, Xinyi Yan, Jonathan M Levine, and Nathan JB Kraft. 2021. “Soil Microbes Generate Stronger Fitness Differences Than Stabilization Among California Annual Plants.” The American Naturalist 197 (1): E30–39.

Krishnadas, Meghna, Robert Bagchi, Sachin Sridhara, and Liza S Comita. 2018. “Weaker Plant-Enemy Interactions Decrease Tree Seedling Diversity with Edge-Effects in a Fragmented Tropical Forest.” Nature Communications 9 (1): 4523.

Lau, Jennifer A, and Jay T Lennon. 2012. “Rapid Responses of Soil Microorganisms Improve Plant Fitness in Novel Environments.” Proceedings of the National Academy of Sciences 109 (35): 14058–62.

Pajares-Murgó, Mariona, José L Garrido, Antonio J Perea, Álvaro López-Garcı́a, Jesús M Bastida, Jorge Prieto-Rubio, Sandra Lendı́nez, Concepción Azcón-Aguilar, and Julio M Alcántara. 2024. “Intransitivity in Plant–Soil Feedbacks Is Rare but Is Associated with Multispecies Coexistence.” Ecology Letters 27 (3): e14408.

Yan, Xinyi, Jonathan M Levine, and Gaurav S Kandlikar. 2022. “A Quantitative Synthesis of Soil Microbial Effects on Plant Species Coexistence.” Proceedings of the National Academy of Sciences 119 (22): e2122088119.