comparing_nutrients.Rmd

---
title: "comparing_nutrients"
author: "Hannah Rempel & Abigail Siebert"
date: "4/23/2021"
output:
  html_document: default
  pdf_document: default
---
This script is used to compare the nutritional value of Brown chromis (Chromis multilineata) fecal pellets quantified in this study to published mean values for tropical red, green, and brown algae (Rhodophyta, Chlorophyta, and Ochrophyta, respectively), epilithic and endolithic algae matrix (EAM), and cyanobacteria (Cyanophyceae). For lipids, proteins, and carbohyrdates we compare all these food sources. Summaries of trace and minor elements only include Brown chromis feces and red, green, and brown algae because we were unable to find data on the trace and minor elements composition of EAM or tropical, naturally occuring cyanobacteria.

#Setup
The following chunk installs required packages and loads them.
```{r loading packages, include=FALSE}
#installs packages if not on the users computer
if (!require(tidyverse)) install.packages("tidyverse")
if (!require(here)) install.packages("here") 
if (!require(purrr)) install.packages("purrr") 
if (!require(plotrix)) install.packages("plotrix") 
if (!require(ggplot2)) install.packages("ggplot2") 
if (!require(grDevices)) install.packages("ggpubr")
if (!require(grDevices)) install.packages("cowplot") 
if (!require(grDevices)) install.packages("magick") 

#creating a list of packages to load at once
packages <- c("tidyverse", "here", "purrr", "plotrix", "ggplot2", "ggpubr", "cowplot", "magick")
sapply(packages, require, character.only = T)  
```

The following reads in the data on the nutrient content of various food items (Chromis multilineata feces, red algae, green algae, brown algae, epilithic algae matrix (EAM), and cyanobacteria).
```{r reading in data, include=FALSE}
#literature values for red, green, and brown algae, EAM, and cyanobacteria
nutrients_literature <- read_csv(here("data/algae_eam_cyanobacteria_nutrients.csv")) %>%
  mutate(category_common=case_when(category_common=="Epilithic algae matrix"~"EAM",
                                   TRUE~category_common)) %>%
  #dropping unnecessary variables
  select(-reference, -journal, -doi, -category_scientific, -conversions)

#feces nutrional values from this study
nutrients_feces <- read_csv(here("data/chromis_multilineata_fecal_nutrients.csv")) 

#checking that all the columns are the same
names(nutrients_literature) == names(nutrients_feces)

#merging the datasets
nutrients_df <- nutrients_feces %>% bind_rows(nutrients_literature) 
```

#Summarizing data: mean, median, min, max

The following chunk summarizes the mean, median, minimum, maximum of values for for each food category (feces, red algae, green algae, brown algae, EAM, and cyanobacteria).
```{r data summary for each food category}
#mean values
summary_mean <- nutrients_df %>%
  group_by(category_common) %>%
  dplyr::summarise(across(calories_kcalperg:zinc_ppm, mean, na.rm= TRUE))

summary_mean

#SEM values, for feces only (since other food soruces are published means, not raw data)
summary_SEM_feces <- nutrients_df %>% 
  filter(category_common=="Feces") %>%
  dplyr::summarise(across(c("proteins_percent", "lipids_percent", "carbohydrates_percent", "ash_percent","calories_kcalperg", "calcium_percent", "magnesium_percent", "phosphorus_percent", "copper_ppm", "iron_ppm", "zinc_ppm"), ~ plotrix::std.error(.x, na.rm = TRUE))) %>%
  mutate_all(round, 1)
    
summary_SEM_feces

#median values
summary_median <- nutrients_df %>%
  group_by(category_common) %>%
  dplyr::summarise(across(calories_kcalperg:zinc_ppm, median, na.rm= TRUE))
  
summary_median

#min values
summary_min <- nutrients_df %>%
  group_by(category_common) %>%
  dplyr::summarise(across(calories_kcalperg:zinc_ppm, min, na.rm= TRUE))
  
summary_min

#max values
summary_max <- nutrients_df %>%
  group_by(category_common) %>%
  dplyr::summarise(across(calories_kcalperg:zinc_ppm, max, na.rm= TRUE))

summary_max
```

This chunk creates a table of summary statistics of major nutrients and total calories (Table S6)
```{r supplementary table of summary statistics of major nutrient values}
mean_vals <- summary_mean %>%
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  add_column(statistic=rep("mean", ncol(summary_mean)-1)) %>%
  select(var, statistic, "Feces", "Cyanobacteria", "EAM", "Brown algae", "Green algae", "Red algae")

median_vals <- summary_median %>%
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  add_column(statistic=rep("median", ncol(summary_mean)-1)) %>%
  select(var, statistic, "Feces", "Cyanobacteria", "EAM", "Brown algae", "Green algae", "Red algae")

min_vals <- summary_min %>%
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  add_column(statistic=rep("min", ncol(summary_mean)-1)) %>%
  select(var, statistic, "Feces", "Cyanobacteria", "EAM", "Brown algae", "Green algae", "Red algae")

max_vals <- summary_max %>%
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  add_column(statistic=rep("max", ncol(summary_mean)-1)) %>%
  select(var, statistic, "Feces", "Cyanobacteria", "EAM", "Brown algae", "Green algae", "Red algae")

combined_summary_stats <- mean_vals %>% bind_rows(median_vals) %>% bind_rows(min_vals) %>% bind_rows(max_vals) %>%
  #reformatting NaN, Inf, and -Inf values as NAs (for variables of a given food item category for which there was no data)
  replace(is.na(.), NA) %>% 
  mutate_if(is.numeric, list(~na_if(., Inf)))%>% 
  mutate_if(is.numeric, list(~na_if(., -Inf)))

#taking feces SEM values and formatting to add to the macronutrient summary table below
SEM_to_add_macros <- summary_SEM_feces %>%
  gather(var, value) %>% 
  add_column(statistic=rep("SEM", ncol(summary_SEM_feces)),
             category_common=rep("Feces", ncol(summary_SEM_feces))) %>%
  spread(category_common, value) %>%
  select(var, statistic, Feces) %>%
  #filtering to only macronutrients
  filter(var %in% c("proteins_percent", "lipids_percent", "carbohydrates_percent", "ash_percent", "calories_kcalperg")) %>%
  #adding in "dummy columns" for other food items
  add_column("Cyanobacteria"=rep("NA", 5),
             "EAM"=rep("NA", 5),
             "Brown algae"=rep("NA", 5),
             "Green algae"=rep("NA", 5),
             "Red algae"=rep("NA", 5)) %>%
  #formatting columns as doubles to bind to the other dataframe
  mutate(Cyanobacteria=as.double(Cyanobacteria),
         "EAM"=as.double("EAM"),
         "Brown algae"=as.double("Brown algae"),
         "Green algae"=as.double("Green algae"),
         "Red algae"=as.double("Red algae"))

table_macronutrient_summary <- combined_summary_stats %>% 
  #filtering to select the relevant variables
  filter(var %in% c("carbohydrates_percent", "lipids_percent", "proteins_percent", "ash_percent", "calories_kcalperg")) %>%
  mutate_if(is.numeric, round,1) %>%
  bind_rows(SEM_to_add_macros) %>%
  #arranging by macronutrient type
  mutate(var = fct_relevel(var, c("proteins_percent", "lipids_percent", "carbohydrates_percent", "ash_percent", "calories_kcalperg"))) %>%
  arrange(var)

#writing the table to a csv
write_csv(table_macronutrient_summary, here("tables/table_s6.csv")) 
```

This chunk creates a table of summary statistics of minor and trace element values (Table S7)
```{r supplementary table of summary statistics of minor and trace element values}
#taking feces SEM values and formatting to add to the macronutrient summary table below
SEM_to_add_micros <- summary_SEM_feces %>%
  gather(var, value) %>% 
  add_column(statistic=rep("SEM", ncol(summary_SEM_feces)),
             category_common=rep("Feces", ncol(summary_SEM_feces))) %>%
  spread(category_common, value) %>%
  select(var, statistic, Feces) %>%
  #filtering to only macronutrients
  filter(var %in% c("calcium_percent", "magnesium_percent", "phosphorus_percent", "copper_ppm", "iron_ppm", "zinc_ppm")) %>%
  #adding in "dummy columns" for other food items
  add_column("Brown algae"=rep("NA", 6),
             "Green algae"=rep("NA", 6),
             "Red algae"=rep("NA", 6)) %>%
  #formatting columns as doubles to bind to the other dataframe
  mutate("Brown algae"=as.double("Brown algae"),
         "Green algae"=as.double("Green algae"),
         "Red algae"=as.double("Red algae"))

table_micronutrient_summary <- combined_summary_stats %>% 
  #dropping these variables because theres no trace or minor element data for these food items
  select(-Cyanobacteria, -"EAM") %>%
  #filtering to select the relevant variables
  filter(var %in% c("calcium_percent", "magnesium_percent", "phosphorus_percent", "copper_ppm", "iron_ppm", "zinc_ppm")) %>%
  mutate_if(is.numeric, round,2) %>%
  bind_rows(SEM_to_add_micros) %>%
  #arranging by macronutrient type
  mutate(var = fct_relevel(var, c("calcium_percent", "magnesium_percent", "phosphorus_percent", "copper_ppm", "iron_ppm", "zinc_ppm"))) %>%
  arrange(var)

#writing the table to a csv
write_csv(table_micronutrient_summary, here("tables/table_s7.csv")) 
```

#Sample size summary

The following chunk summarizes the sample size for each food category (feces, red algae, green algae, brown algae, EAM, and cyanobacteria), where feces are samples from the present study, and values for all other food items are refrenced from published studies. 
```{r referenced article sample size per category and nutrient type}
n_carbs <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(carbohydrates_percent) %>%
  drop_na(carbohydrates_percent) %>%
  dplyr::summarize(carbs_n=n())

n_ash <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(ash_percent) %>%
  drop_na(ash_percent) %>%
  dplyr::summarize(ash_n=n())

n_lipids <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(lipids_percent) %>%
  drop_na(lipids_percent) %>%
  dplyr::summarize(lipids_n=n())

n_protein <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(proteins_percent) %>%
  drop_na(proteins_percent) %>%
  dplyr::summarize(protein_n=n())

n_cal <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(calories_kcalperg) %>%
  drop_na(calories_kcalperg) %>%
  dplyr::summarize(cal_n=n())

n_mg <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(magnesium_percent) %>%
  drop_na(magnesium_percent) %>%
  dplyr::summarize(mg_n=n())

n_p <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(phosphorus_percent) %>%
  drop_na(phosphorus_percent) %>%
  dplyr::summarize(p_n=n())

n_ca <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(calcium_percent) %>%
  drop_na(calcium_percent) %>%
  dplyr::summarize(ca_n=n())

n_fe <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(iron_ppm) %>%
  drop_na(iron_ppm) %>%
  dplyr::summarize(fe_n=n())

n_cu <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(copper_ppm) %>%
  drop_na(copper_ppm) %>%
  dplyr::summarize(cu_n=n())

n_zn <- nutrients_df %>% 
  group_by(category_common) %>%  
  select(zinc_ppm) %>%
  drop_na(zinc_ppm) %>%
  dplyr::summarize(zn_n=n())

#tibble with the number of refrences per category and nutrient type
reference_count <- n_cal %>% left_join(n_ash) %>% left_join(n_carbs) %>% left_join(n_lipids) %>% left_join(n_protein) %>% 
  left_join(n_mg) %>% left_join(n_p) %>% left_join(n_ca) %>% left_join(n_fe) %>% left_join(n_cu) %>% left_join(n_zn) %>%
  #replaces NA values with 0 (i.e., 0 references)
  replace(is.na(.), 0)
reference_count 
```

This chunk filters and formats the 'reference_count' tibble to create a supplementary table of the number of references per food item category considered for major nutrients (defined here as proteins, lipids, carbohydrates, ash, and total calories)
```{r supplementary table of sample size for major nutrients}
#supplementary table of the number of references per category for proteins, lipids, carbs, ash, and total calories
table_major_nutrients <- reference_count %>% 
  #selecting variables of interest
  select(category_common, protein_n, lipids_n, carbs_n, ash_n, cal_n) %>% 
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  #specifying variable order
  mutate(variable = fct_relevel(var, c("protein_n", "lipids_n", "carbs_n", "ash_n", "cal_n"))) %>%
  #rearranging column order
  select(variable, "Cyanobacteria", "EAM", "Brown algae", "Green algae", "Red algae") %>%
  #rearranging row order
  arrange(variable)

table_major_nutrients

#writing the table to a csv
write_csv(table_major_nutrients, here("tables/table_s2.csv")) 
```

This chunk filters and formats the 'reference_count' tibble to create a supplementary table of the number of references per food item category considered for minor and trace minerals (magnesium, phosphorus, calcium, iron, copper, zinc). We were unable to find references for cyanobacteria and EAM, thus these are excluded from subsequent visualization of the minor and trace minerals in food items.
```{r supplementary table of sample size for minor and trace elements}
#supplementary table of the number of references per category for minor and trace elements
table_minor_nutrients <- reference_count %>%  
  #selecting variables of interest
  select(category_common, mg_n, p_n, ca_n, fe_n, cu_n, zn_n) %>% 
  #transposing the data
  gather(var, value, -category_common) %>% 
  spread(category_common, value) %>%
  #specifying variable order
  mutate(variable = fct_relevel(var, c("mg_n", "p_n", "ca_n", "fe_n", "cu_n", "zn_n"))) %>%
  #rearranging column order
  select(variable, "Brown algae", "Green algae", "Red algae") %>%
  #rearranging row order
  arrange(variable)

table_minor_nutrients

#writing the table to a csv
write_csv(table_minor_nutrients, here("tables/table_s3.csv")) 
```

#Visualizing data

The following chunk visualized data on the protein, lipid, and carbohydrate content of red, green, and brown algae, EAM, cyanobacteria, and Chromis multilineata feces.
```{r fig 3: proteins lipids carbs}
#ordering the y axis values for the figure
y_axis_ordered <- factor(nutrients_df$category_common, levels=c("Red algae", "Green algae", "Brown algae", "EAM", "Cyanobacteria","Feces"))

#generating a color palette for the figure
food_palette <- c("#A07272", "#8AAB7E", "#BF8E4C", "#8BC2BD", "#A11932", "#9D9D9D")

#Proteins
fig_protein <- nutrients_df %>%
  ggplot(aes(y = y_axis_ordered, x = proteins_percent, fill = y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Proteins (%)",
       y="") +
  theme(legend.position = "none",
        axis.title.x=element_text(size=10, color="black"),
        axis.text=element_text(size=10, color="black")) +
  scale_fill_manual(values=food_palette)

#Lipids
fig_lipid <- nutrients_df %>%
  ggplot(aes(y = y_axis_ordered, x = lipids_percent, fill = y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Lipids (%)",
       y="") +
  theme(legend.position = "none",
        axis.title.x=element_text(size=10, color="black"),
        axis.text=element_text(size=10, color="black")) +
  scale_fill_manual(values=food_palette)

#Carbohydrates
fig_carb <- nutrients_df %>%
  ggplot(aes(y = y_axis_ordered, x = carbohydrates_percent, fill = y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Carbohydrates (%)",
       y="") +
  theme(legend.position = "none",
        axis.title.x=element_text(size=10, color="black"),
        axis.text=element_text(size=10, color="black")) +
  scale_fill_manual(values=food_palette)

#Calories
fig_cal <- nutrients_df %>%
  ggplot(aes(y = y_axis_ordered, x = calories_kcalperg, fill = y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x=expression(paste("Energy (kcal ", g^-1, ")")),
       y="") +
  theme(legend.position = "none",
        axis.title.x=element_text(size=10, color="black"),
        axis.text=element_text(size=10, color="black")) +
  scale_fill_manual(values=food_palette)
  
#combining graphs (first step)
fig_macronutrients_initial <- ggarrange(fig_protein + theme(plot.margin = margin(0.1,0.1,2,0.1, "cm"),
                                                     axis.title.y = element_blank()), 
                   fig_lipid + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    axis.title.y = element_blank(),
                                    plot.margin = margin(0.1,0.1,2,0.1, "cm")), 
                  fig_carb + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    axis.title.y = element_blank(),
                                    plot.margin = margin(0.1,0.1,2,0.1, "cm")), 
                   fig_cal + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    axis.title.y = element_blank(),
                                    plot.margin = margin(0.1,0.1,2,0.1, "cm")), 
                   nrow = 1,
                   labels = "",
                 widths=c(1.5,1,1,1),
                  align = "h")

#reading in images of food items to incorperate in figure
feces <- magick::image_read(path=here("figures/icons/Chromis_feces.png"), density = NULL, depth = NULL, strip = TRUE)
cyano <- magick::image_read(path=here("figures/icons/Cyanobacteria.png"), density = NULL, depth = NULL, strip = TRUE)
EAM <- magick::image_read(path=here("figures/icons/EAM.png"), density = NULL, depth = NULL, strip = TRUE)
brown_algae <- magick::image_read(path=here("figures/icons/Dictoyota_algae.png"), density = NULL, depth = NULL, strip = TRUE)
green_algae <- magick::image_read(path=here("figures/icons/Halimeda_algae.png"), density = NULL, depth = NULL, strip = TRUE)
red_algae <- magick::image_read(path=here("figures/icons/Laurencia_algae.png"), density = NULL, depth = NULL, strip = TRUE)

#creating finalized figure with images of food items
fig_macronutrients <- cowplot::ggdraw() +
  draw_plot(fig_macronutrients_initial, scale=0.9, hjust = -0.02)+ #adds the plot (order matters to overlay the layers correctly)
  draw_image(feces, x = -0.46, y = 0.375, scale=0.08)+
  draw_image(cyano, x = -0.46, y = 0.28, scale=0.07)+
  draw_image(EAM, x = -0.46, y = 0.185, scale=0.06)+
  draw_image(red_algae, x = -0.46, y = -0.115, scale=0.087)+
  draw_image(green_algae, x = -0.46, y = -0.012, scale=0.082)+ 
  draw_image(brown_algae, x = -0.46, y = 0.09, scale=0.075)

fig_macronutrients

#saving figure
ggsave(filename=(here::here("figures/fig3_macronutrients.png")),
      plot=fig_macronutrients,
      width = 174, height = 115,
      units = "mm",
      dpi = 300)
```

The following chunk visualized data on the minor and trace element content of red, green, and brown algae, EAM, cyanobacteria, and Chromis multilineata feces.
```{r fig 4: minor and trace elements}
#Getting rid of EAM data
elements_categories <- nutrients_df %>%
  #filtering out EAM and cyanobacteria because we were unable to finddata on their minor and trace element content
  dplyr::filter(category_common %in% c("Brown algae", "Green algae", "Red algae", "Feces")) %>%
  select(category_common, magnesium_percent:zinc_ppm)

#ordering the y axis values for the figure
fig2_y_axis_ordered <- factor(elements_categories$category_common , levels=c("Red algae", "Green algae", "Brown algae", "Feces"))

#generating a color palette for the figure
nutrient_palette <- c("#A07272", "#8AAB7E", "#BF8E4C", "#9D9D9D")

#Calcium
fig_calcium <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = calcium_percent, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Calcium (%)",
       y="") +
    scale_x_continuous(labels = scales::number_format(accuracy = 0.1))+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)

#Phosphorus
fig_phosphorus <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = phosphorus_percent, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Phosphorus (%)",
       y="") +
    scale_x_continuous(labels = scales::number_format(accuracy = 0.1))+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)

#Magnesium
fig_magnesium <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = magnesium_percent, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Magnesium (%)",
       y="")+
    scale_x_continuous(labels = scales::number_format(accuracy = 0.1))+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)
  
#Copper
fig_copper <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = copper_ppm, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Copper (ppm)",
       y="") +
  lims(x=c(0,30))+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)

#Iron
fig_iron <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = iron_ppm, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Iron (ppm)",
       y="")+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)

#Zinc
fig_zinc <- elements_categories %>%
  ggplot(aes(y = fig2_y_axis_ordered, x = zinc_ppm, fill = fig2_y_axis_ordered)) +
  geom_boxplot() +
  theme_classic() +
  labs(x="Zinc (ppm)",
       y="")+
  theme(legend.position = "none",
        axis.text=element_text(size=10, color="Black"),
        axis.title=element_text(size=10, color="Black"))+
  scale_fill_manual(values=nutrient_palette)

#Combining graphs into one figure
fig_micronutrients_initial <- ggarrange(fig_calcium + theme(axis.title.y = element_blank(),
                                                plot.margin = unit(c(t=0,r=0.25,b=0,l=0.25), "cm")),
                                
                              fig_magnesium + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    plot.margin = unit(c(t=0,r=0.25,b=0,l=0.1), "cm")),
                              
                              fig_phosphorus + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    plot.margin = unit(c(t=0,r=0.25,b=0,l=0.25), "cm")),
                              
                              fig_copper  + theme(axis.title.y = element_blank(),
                                                       plot.margin = unit(c(t=0.25,r=0.1,b=0,l=0.1), "cm")),
                                
                              fig_iron + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    plot.margin = unit(c(t=0.25,r=0.25,b=0,l=0.25), "cm")),
                            
                              fig_zinc + theme(axis.text.y = element_blank(),
                                    axis.ticks.y = element_blank(),
                                    plot.margin = unit(c(t=0,r=0.25,b=0,l=0.1), "cm")),
                                
                              ncol = 3, nrow = 2,
                            align = "h",
                  widths=c(1, 0.78, 0.78, 1, 0.78, 0.78))

#creating finalized figure with images of food items
fig_micronutrients <- cowplot::ggdraw() +
  draw_plot(fig_micronutrients_initial, scale=0.9, hjust = -0.06)+ #adds the plot (order matters to overlay the layers correctly)
  #upper set
  draw_image(feces, x = -0.45, y = 0.38, scale=0.08)+
  draw_image(brown_algae, x = -0.45, y = 0.3, scale=0.068)+
  draw_image(green_algae, x = -0.45, y = 0.22, scale=0.07)+ 
  draw_image(red_algae, x = -0.45, y = 0.13, scale=0.07)+
  #lower set
  draw_image(feces, x = -0.45, y = -0.08, scale=0.08)+
  draw_image(brown_algae, x = -0.45, y = -0.145, scale=0.068)+
  draw_image(green_algae, x = -0.45, y = -0.23, scale=0.07)+ 
  draw_image(red_algae, x = -0.45, y = -0.315, scale=0.07)

fig_micronutrients

#saving figure
ggsave(filename=here::here("figures/fig4_micronutrients.png"),
      plot=fig_micronutrients,
      width = 174, height = 115,
      units = "mm",
      dpi = 300)
```