1 Why GD and GL?

Capblancq et al. (2021) observed strong positive association between genetic diversity (GD) and Early-Life fitness, and negative assoaciation between genetic load (GL) and Early-Life fitness of red spruce. We use this information for the range edge of red spruce to select for source combinations with high GD and low GL. The fitness was estimated based on the mean family level fitness for each populations. The mean of the resulting pop combinations are then used as fitness estimate for the respective combinations.

1.1 Genetic diversity and Genetic load

require(tidyverse)
require(lmerTest)
require(sjPlot)
require(ggpubr)

GDGL_list <- readRDS("./data/Genetic_diversity_and_Genetic_load/GDGL_list")
MD_GDGL <- GDGL_list$Maryland_GDGL
WV_GDGL <- GDGL_list$West_Virginia_GDGL
VA_GDGL <- GDGL_list$Virginia_GDGL

MD_GDGL <- MD_GDGL[-1,]
WV_GDGL <- WV_GDGL[-1,]
VA_GDGL <- VA_GDGL[-1,]

1.2 Early-Life fitness

fitness <- read.table("./data/FitnessTraits_GeneticParameters_RedSpruce.txt",
                      header = T, stringsAsFactors = T)
fitness$Region <- factor(fitness$Region, levels = c("C","M","E"),
                                         labels = c("Core","Margin","Edge"))

1.3 Model stats

# model
modelFitness <- lm(Fitness ~ Population_Homozygosity + Family_Homozygosity + Genetic_Diversity + Genetic_Load + SeedWeight, data=fitness %>% filter(Region=="Edge"))

# modelFitness <- lm(Fitness ~ Genetic_Diversity + Genetic_Load + SeedWeight + Region + Region*Genetic_Diversity + Region*Genetic_Load, data=fitness)
summary(modelFitness)


Call:
lm(formula = Fitness ~ Population_Homozygosity + Family_Homozygosity + 
    Genetic_Diversity + Genetic_Load + SeedWeight, data = fitness %>% 
    filter(Region == "Edge"))

Residuals:
     Min       1Q   Median       3Q      Max 
-10.8125  -3.8816  -0.0977   3.0421  16.1853 

Coefficients:
                         Estimate Std. Error t value Pr(>|t|)   
(Intercept)               171.160    114.954   1.489  0.13981   
Population_Homozygosity -1502.700    555.629  -2.705  0.00811 **
Family_Homozygosity         9.763     45.974   0.212  0.83229   
Genetic_Diversity         658.994    196.546   3.353  0.00115 **
Genetic_Load             -222.175    106.385  -2.088  0.03944 * 
SeedWeight               2031.839    688.135   2.953  0.00397 **
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 5.968 on 95 degrees of freedom
Multiple R-squared:  0.1985,    Adjusted R-squared:  0.1563 
F-statistic: 4.705 on 5 and 95 DF,  p-value: 0.0006982

1.4 Model Viz.

# model marginal effects
p1 <- plot_model(modelFitness, type = "pred", terms = c("Genetic_Diversity"),
                 title = "", axis.title = "") + 
  # aes(linetype=group,color=group)+
  geom_line(alpha=1, linewidth=1) +
  # scale_linetype_manual(values = c("solid", "dashed", "dotted"),
                        # name = "Region", 
                        # labels = c("Core", "Margin","Edge"))+ 
  facet_wrap(~"Genetic Diversity") + 
  # ylim(-20,40) + 
  labs(y="Early-Life Fitness")+
  theme_bw(base_size = 11, base_family = "Times") +
            theme(axis.text=element_text(size=14), 
                  axis.title=element_text(size=18),
                  panel.background = element_blank(), 
                  legend.background = element_blank(), 
                  panel.grid = element_blank(), 
                  plot.background = element_blank(), 
                  legend.text=element_text(size=rel(.8)), 
                  strip.text = element_text(size=30)) +
  theme(axis.title.x=element_blank(), panel.grid = element_blank(),
        legend.key=element_blank(), legend.background=element_blank())

p2 <- plot_model(modelFitness, type = "pred", terms = c("Genetic_Load"), 
                 title = "", axis.title = "") + 
  # aes(linetype=group,color=group)+
  geom_line(alpha=1, size=1) +
  # scale_linetype_manual(values = c("solid", "dashed", "dotted"),
  #                       name = "Region", 
  #                       labels = c("Core", "Margin","Edge"))+
  facet_wrap(~"Genetic Load") + 
  # ylim(-20,40) + 
  # labs(color  = "Region", linetype = "Region")+
  theme_bw(base_size = 11, base_family = "Times") +
            theme(axis.text=element_text(size=14), 
                  axis.title=element_text(size=18),
                  panel.background = element_blank(), 
                  legend.background = element_blank(), 
                  panel.grid = element_blank(), 
                  plot.background = element_blank(), 
                  legend.text=element_text(size=rel(.8)), 
                  strip.text = element_text(size=30)) +
  theme(axis.title.x=element_blank(), axis.text.y = element_blank(), axis.ticks.y = element_blank(), panel.grid = element_blank())

ggarrange(p1,p2, ncol = 2, common.legend = T, legend = "right", widths = c(1.1,1,1,1))

Early-Life fitness of seedlings from range edge against different genetic metrics