MRS Long-term Monitoring

DT::datatable(MRS_LMM)

Download: https://github.com/RandomForestz/MRS_Subalpine/blob/main/MRS_LMM.RData

Metadata

Variable <- c("ID", "Time", "Species", "DBH", "BA", "ABAI", "Height", "Comp", "Dead", "Dead_Census", "Elevation",
              "Moisture_Class", "Stand", "TWI")
Description <- c("Individual Tree ID (Plot + Tree Number)",
                 "A time stamp for data collection with 1 = 1982,1983,1986, 2 = 2016, 3 = 2022",
                 "Lodgepole pine (PICO), Limber pine (PIFL), Subalpine fir (ABLA), Engelmann spruce (PIEN)",
                 "Diameter at breast height for specfic data collection period",
                 "Diameter at breast height converted to basal area",
                 "The difference of diameter at breast height converted to basal area. Annualized according to time between data collection periods",
                 "Subjective height class given at data collection",
                 "Competition index based on Hegyi's Index (1970). Higher values represent higher competition within a 6m buffer on each tree",
                 "1 = Dead, 0 = Alive",
                 "Indicating whether the tree died during time period 2 (Census 1) or 3 (Census 2)",
                 "Elevation (m) of plot center",
                 "Subjective moisture class",
                 "Stand age",
                 "Topographic wetness index")

metadata <- cbind.data.frame(Variable, Description)
DT::datatable(metadata)

MRS Plots

# growth ~ height
MRS_LMM_height <- MRS_LMM %>% 
  subset(Height %in% c("Suppressed", "Co-Dominant", "Dominant"))
level_order1 <- c("Suppressed", "Co-Dominant", "Dominant")
ggplot(data = MRS_LMM_height) +
  geom_boxplot(mapping = aes(x = factor(Height, level = level_order1), y = log(ABAI), fill = Height)) +
  scale_fill_brewer(palette = "YlOrBr") +
  labs(x = "Height Class", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Height", fill = "Height Class") +
  theme_bw()

Smaller trees have slightly higher growth across the dataset, though distributions look very similar.

# growth ~ moisture
ggplot(data = MRS_LMM) +
  geom_boxplot(mapping = aes(x = as.factor(Moisture_Class), y = log(ABAI), fill = Moisture_Class)) +
  scale_fill_brewer(palette = "RdYlBu", direction = -1) +
  labs(x = "Moisture Class", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Moisture", fill = "Moisture Class") +
  theme_bw()

Wetter sites have slightly higher growth though distribution of growth is very similar across the dataset.

# growth ~ elevation
ggplot(data = MRS_LMM) +
  geom_boxplot(mapping = aes(x = as.factor(Elevation), y = log(ABAI), fill = as.factor(Elevation))) +
  scale_fill_brewer(palette = "Blues") +
  labs(x = "Plot Elevation", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Elevation", fill = "Elevation") +
  theme_bw()

Growth seems to fluctuate across elevation but not significantly.

# growth ~ Stand type
z <- ggplot(data = MRS_LMM) +
  geom_boxplot(mapping = aes(x = as.factor(Stand), y = log(ABAI), fill = as.factor(Stand))) +
  scale_fill_brewer(palette = "Blues") +
  labs(x = "Stand", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Stand Type", fill = "Stand") +
  theme_bw()
z + theme(axis.text.x = element_text(angle = 300, vjust = 0.5, hjust=.1))

Growth is similar across stand age/type with MRS showing over less growth when compared to the rest.

# Growth ~ TWI | Species
ggplot(data = MRS_LMM) +
  geom_smooth(mapping = aes(x = TWI, y = log(ABAI)), method = "lm") +
  labs(x = "TWI", y = "Annualized Basal Area Increment (log)", title = "Growth ~ TWI | Species") +
  facet_wrap(~Species) +
  theme_bw()

PICO and PIFL really favor wetter sites with growth showing a positive trend with topographic wetness. ABLA and PIEN are pretty flat with a slight decrease.

# Growth ~ Species | Moisture Class
ggplot(data = MRS_LMM) +
  geom_boxplot(mapping = aes(x = as.factor(Species), y = log(ABAI), fill = Species)) +
  scale_fill_brewer(palette = "BrBg") +
  facet_wrap(~Moisture_Class) +
  labs(x = "Species", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Species", fill = "Species") +
  theme_bw()

Across moisture class, PICO & PIFL growth favors moisture, with higher growth in mesic and lower growth in xeric. ABLA looks similar in all sites. PIEN shows higher distribution of growth in xeric sites than mesic, most likely due to a few big trees doing well in drier sites.

pines <- MRS_LMM %>% 
  subset(Species %in% c("PIFL", "PICO"))
level_order <- c("Xeric", "Mesic")
ggplot(data = pines) +
  geom_boxplot(mapping = aes(x = factor(Moisture_Class, level = level_order), y = log(ABAI), fill = Moisture_Class)) +
  scale_fill_brewer(palette = "BrBg", direction = -1) +
  facet_wrap(~Species) +
  labs(x = "Site Moisture Class", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Moisture Class (Pines)", fill = "Class") +
  theme_bw()

PICO & PIFL favor wetter conditions for growth. An assumption could also be that drier sites have increased competition for pines (MRS 1).

ggplot(data = MRS_LMM) +
  geom_point(mapping = aes(x = Comp, y = log(ABAI))) +
  geom_smooth(mapping = aes(x = Comp, y = log(ABAI))) +
  facet_wrap(~Species) +
  labs(x = "Competition Index", y = "Annualized Basal Area Increment (log)", title = "Growth ~ Competition | Species") +
  theme_bw()

Competition shows a negative trend in growth for all species. PICO & ABLA experience higher values of competition than PIEN & PIFL.

ggplot(data = MRS_LMM) +
  geom_smooth(mapping = aes(x = TWI, y = ABAI, col = Species), method = "lm", se = F) +
  facet_wrap(~Moisture_Class) +
  theme_bw()
## `geom_smooth()` using formula = 'y ~ x'

PIFL: In all sites present, Limber pine shows increasing ABAI with topographic wetness.There are more PIFL in xeric sites, but both appear to show similar trends.

PICO: Lodgepole shows a slight negative trend in ABAI across increasing topographic wetness. However, trends show ABAI is overall higher in Mesic sites.

ABLA: In hydric and mesic sites, trends in ABAI across topographic wetness are meh. Not much going on - around same values, slightly different trends. However, in xeric sites, ABLA ABAI sky rockets across increasing topographic wetness.

PIEN: Who knows what’s going on in mesic sites. I’m assuming there is some sort of other influence like competition or small sample size.

ggplot(data = MRS_LMM) +
  geom_point(mapping = aes(x = TWI, y = log(ABAI))) +
  geom_smooth(mapping = aes(x = TWI, y = log(ABAI)), se = F, method = "lm") +
  facet_wrap(~Moisture_Class) +
  labs(x = "Topographic Wetness Index", y = "Annualized Basal Area Increment (log)", title = "Growth ~ TWI | Moisture Class") +
  theme_bw()

The variation of moisture (TWI) within moisture class is of interest. Wetter micro sites within xeric plots favor growth.

ggplot(MRS_LMM, aes(Comp)) + 
  geom_density(aes(fill=factor(Moisture_Class)), alpha=0.8) + 
  facet_wrap(~Species) +
  theme_bw() +
  labs(x = "Competition", y = "Density", fill = "Moisture Class")

Competition by moisture class. Not very helpful.

MRS Dendrometer Bands

band_mmt <- read.csv("D:/Dissertation/Analysis/Chapter_1/MRS/MRS/data/dendroband/band_MMT.csv")
DT::datatable(band_mmt)

Download: https://github.com/RandomForestz/MRS_Subalpine/blob/main/band_MMT.csv

ggplot(data = band_mmt) +
  geom_boxplot(mapping = aes(x = as.factor(Species), y = MMT, fill = Species)) +
  facet_wrap(~Moisture) +
  ylim(c(0, 10)) + 
  theme_bw()

Distribution of species across moisture classes.

ggplot(data = band_mmt) +
  geom_boxplot(mapping = aes(x = as.factor(Species), y = MMT, fill = Species)) +
  facet_wrap(~Year) +
  ylim(c(0, 10)) + 
  theme_bw()

Bands are starting to fit and gather data over time.

ggplot(data = band_mmt) +
  geom_point(mapping = aes(x = Comp, y = MMT, col = Species), alpha = .2) +
  geom_smooth(mapping = aes(x = Comp, y = MMT, col = Species), se = F, method = "lm") +
  facet_grid(~Year) +
    ylim(c(0, 5)) + 
  theme_bw()
## `geom_smooth()` using formula = 'y ~ x'

ggplot(data = band_mmt) +
  geom_point(mapping = aes(x = TWI, y = ABAI, col = Species), alpha = .2) +
  geom_smooth(mapping = aes(x = TWI, y = ABAI, col = Species), se = F, method = "lm") +
  facet_grid(~Year) +
  ylim(c(0, 0.05)) + 
  theme_bw()
## `geom_smooth()` using formula = 'y ~ x'

About the same as we’re seeing with the MRS plots. ABLA really has done well with fitting bands and showing growth.