Add files via upload

pigments/hplc-pigments-analysis.Rmd

@@ -1,7 +1,11 @@
 ---
-title: "Line P Pigments"
+title: "Line P Pigment Data Analysis"
 author: "Colleen Kellogg, Hakai Institute"
-output: html_document
+output: 
+  rmarkdown::html_document:
+    theme: lumen
+    toc: true
+    toc_float: true
 ---
 
 ```{r setup, include=FALSE}
@@ -20,23 +24,10 @@ library(ggfortify)
 
 ```
 
-### Questions we are asking of the dataset:
-
-#### There are three oceanic states we are considering in this analysis. 
-1. Non-marine heatwave [before 2015, 2016-May 2018]
-2. MHW1 (high particles). 2015
-3. MHW2 (high particles). March-August 2019 (with moderate heatwaves beginning in June 2018 but most severe was in July-Sept 2019)
-
-#### Particle dynamics differences:
-1. first MHW, high export belonged to particles exported during Spring, then export pulses along the year until early fall [2015]
-2. second MHW, export was shorter and more intense (observe the z-axis from the plot above and below), with most of export happening during spring to early summer. [2019]
-
 #### We are using 2 types of biological data: HPLC pigments and metabarcoding (16S rRNA gene and 18S rRNA gene amplicon sequencing) data. 
 
-This file was used for the analysis of the HPLC pigment data
-
-### Pigment Data
-This data is derived from HPLC pigment data generated by Angelica Peña ([email protected]) at Fisheries and Oceans Canada. Pigment concentrations were passed through CHEMTAX, a program used to predict phytoplankton functional groups present based on ratios of pigments present in a water sample. 
+#### Pigment Data
+This data analyzed here is derived from HPLC pigment data generated by Angelica Peña ([email protected]) at Fisheries and Oceans Canada. Pigment concentrations were passed through CHEMTAX, a program used to predict phytoplankton functional groups present based on ratios of pigments present in a water sample. 
 
 ```{R load hplc data, warning = FALSE}
 linep20_hplc0 <-tibble(read.csv("LineP-HPLC-data_wP20.csv", header = TRUE))
@@ -71,7 +62,29 @@ tchla_ave_no2019<-linep20_hplc_long %>%
   filter(Year != "2019") %>%
   group_by(Phytoplankton_Group, Month) %>% 
   summarise(mean = mean(Concentration, na.rm=TRUE)) 
+```
+
+### PERMANOVA to examine dataset for station and month differences
+
+```{R permanovas, warning = FALSE}  
+
+#PERMANOVA to see if there a significant differences in pigment among stations
+linep20_hplc_nm<-na.omit(linep20_hplc) %>% mutate(SampleID = paste(Station, Year, Month, sep = '_'))
+
+adonis2(linep20_hplc_nm[7:13] ~ Station, data = linep20_hplc_nm) #R2 = 0.01082, p = 0.99
+adonis2(linep20_hplc_nm[7:13] ~ Month, data = linep20_hplc_nm) #R2 = 0.13017, p = 0.001
+adonis2(linep20_hplc_nm[7:13] ~ Month*Station, data = linep20_hplc_nm) #R2 = 0.17468, p = 0.001
+
+#significant differences by month (and month*station), but not by station alone
+
+
+```
 
+
+### Calculate phytoplankton group averages and plot
+We are averaging the phytoplankton groups across the P20-P26 transect since (1) there is no significant difference by station and (2) the argo floats were moving throughout this whole region, generating a regional dataset (akin to a pigment dataset average)
+
+```{R group averages, warning = FALSE}
 #### phyto group averages for plotting ####
 
 hplc_ave<-linep20_hplc_long %>% 
@@ -131,6 +144,8 @@ hplcp20_rel + annotate("rect", xmin = 12.5, xmax = 15.5, ymin =0, ymax = 1.15, a
 
 ```
 
+### Ordination of pigment data
+
 ```{R HPLC ordinations and multi-panel figure, warning = FALSE}
 
 #ordination
@@ -180,10 +195,18 @@ hplc_pca_legend<-as_ggplot(hplc_pca_leg)
 
 mhw_hplc_pca_nolgnd<-autoplot(res.pca, data = na.omit(linep20_hplc), colour = 'Year', shape = "Season", loadings = TRUE, loadings.label = TRUE, loadings.color = "black", loadings.label.color = "black", size = 3,loadings.label.repel = TRUE) + theme_bw() + scale_color_manual(values = MHWcols_full_BGCmatch_pigs) + theme(axis.text.x = element_text(size = 8),axis.text.y = element_text(size = 8), axis.title.x = element_text(size = 10), axis.title.y = element_text(size = 10), legend.position = "NULL")
 
+mhw_hplc_pca_nolgnd
+
 
 ord_side <- plot_grid(mhw_hplc_pca_nolgnd, NULL, hplc_pca_legend, nrow = 3, rel_heights = c(1,0.05,0.2), labels=c("C",""), label_size = 10, align = "v")
 
+ord_side
+
+```
+
+### Create bar graphs + ordination figure panel for manuscript
 
+```{R figure panel for manuscript, warning = FALSE}
 #now attach ordination and transect lines with the HPLC bar plots - rectangles
 hplc_avebars_ord<-plot_grid(hplc_bars_both, NULL, ord_side, ncol=3, rel_widths = c(1.5,0.05,1), align = "v")
 
@@ -205,16 +228,4 @@ hplc_avebars_ord_line
 
 ```
 
-```{R permanovas, warning = FALSE}  
 
-#PERMANOVA to see if there a significant differences in pigment among stations
-linep20_hplc_nm<-na.omit(linep20_hplc) %>% mutate(SampleID = paste(Station, Year, Month, sep = '_'))
-
-adonis2(linep20_hplc_nm[7:13] ~ Station, data = linep20_hplc_nm) #R2 = 0.01082, p = 0.99
-adonis2(linep20_hplc_nm[7:13] ~ Month, data = linep20_hplc_nm) #R2 = 0.13017, p = 0.001
-adonis2(linep20_hplc_nm[7:13] ~ Month*Station, data = linep20_hplc_nm) #R2 = 0.17468, p = 0.001
-
-#significant differences by month (and month*station), but not by station alone
-
-
-```