Overview

This report explores adult fish passage of coho, jack coho and steelhead salmon recorded from January 2001 to December 2010 at the Willamette Falls fish ladder on the Willamette River (Oregon). This document consists of 3 tabs: a time series, a seasonplot, and annual counts by species. This data was recorded by and made accessible through the Columbia River DART project. Fish counting through Willamette Falls fishway was recorded using video cameras and time lapsed video recorders to record fish passage 24 hrs/day, 365 days/year.

Data source: Columbia River DART (Data Access in Real Time), 2010. Data Courtesy of U.S. Army Corps of Engineers, NWD and Chelan, Douglas, and Grant County PUDs, Yakima Klickitat Fisheries Project, Colville Tribes Fish & Wildlife (OBMEP), Oregon Department of Fish & Wildlife, Washington Department of Fish & Wildlife.

Sockeye Salmon (Oncorhynchus nerka) Credit: Jason Ching/U. Washington

Credit: US Army Corp of Engineers (link: https://www.nwp.usace.army.mil/willamette/locks/) Figure 1. Map of Willamette Fish Ladder.

Data preparation

1. Package & theme setup

knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
library(here)
library(lubridate)
library(tsibble)
library(janitor)
library(feasts)
library(slider)
library(patchwork)

2. Data wrangling

#read in and wrangle the data
fish_main <- read_csv(here("data", "willamette_fish_passage.csv")) %>%
  replace(is.na(.), 0) %>%
  clean_names() %>% 
  select("date", "coho", "jack_coho", "steelhead") %>% 
  mutate(date = lubridate::mdy(date)) %>% 
  group_by(date, coho, jack_coho, steelhead) %>% 
  summarize(total_count = sum(coho, jack_coho, steelhead))

#prepare time series dataframe 
fish_ts <- fish_main %>% 
  as_tsibble(key = NULL, index = date)
  
#convert time series to tsibble and use pivot_longer to reformat
fish_ts_longer <- fish_main %>% 
  as_tsibble(key = NULL, index = date) %>% 
  pivot_longer("coho":"steelhead",
               names_to = "species",
               values_to = "value") %>% 
  mutate(species = 
         ifelse(species == "coho", "Coho",
         ifelse(species == "jack_coho", "Jack Coho", 
         ifelse(species == "steelhead", "Steelhead", 
                "" )))) %>% 
  select(-total_count) %>% 
  rename(total_count = value)

#formate data for tab 3
fish_longer <- fish_main %>% 
  pivot_longer("coho":"steelhead",
               names_to = "species",
               values_to = "value")

3. Data familiarizing and initial plotting

#explore the data

#individually plotting over time
ggplot() +
  geom_line(data = fish_ts, aes(x = date, y = jack_coho)) +
  geom_line(data = fish_ts, aes(x = date, y = steelhead)) +
  geom_line(data = fish_ts, aes(x = date, y = coho))
#explore data by month and year using index_by() 

#by month
fish_month <- fish_ts_longer %>% 
  index_by(yr_mo = ~yearmonth(.)) %>% 
  group_by(yr_mo) %>% 
  summarize(month_mean_count = mean(total_count))

ggplot(data = fish_month, aes(x = yr_mo, y = month_mean_count)) +
  geom_line() +
  facet_wrap(~month(yr_mo, label = TRUE))
#by year
fish_year <- fish_ts_longer %>% 
  index_by(yearly = ~year(.)) %>% 
  summarize(yearly_mean_count = mean(total_count))

ggplot(data = fish_year, aes(x = yearly, y = yearly_mean_count)) +
  geom_col() +
  facet_wrap(~yearly)

Tab 1 - Time Series (Hannah)

# explore data with initial graph
# ggplot(data = fish_ts_longer, aes(x=date, y=total_count)) +
#   geom_line() +
#   facet_wrap(~species)

# get species alone and use patchwork
# coho
coho_tab1 <- fish_ts_longer %>% 
  filter(species == "Coho")

coho_graph <- ggplot(data=coho_tab1, aes(x=date, y=total_count)) +
  geom_line() +
  theme_minimal() +
  labs(title = "Coho", y="total count")

# jack_coho
jackcoho_tab1 <- fish_ts_longer %>% 
  filter(species == "Jack Coho")

jackcoho_graph <- ggplot(data=jackcoho_tab1, aes(x=date, y=total_count)) +
  geom_line() +
  theme_minimal() +
  labs(title = "Jack Coho", y = "total count")

# steelhead
steelhead_tab1 <- fish_ts_longer %>% 
  filter(species == "Steelhead")

steelhead_graph <- ggplot(data=steelhead_tab1, aes(x=date, y=total_count)) +
  geom_line() +
  theme_minimal() +
  labs(title = "Steelhead", y = "total count")

# patchwork

tab1_comb <- coho_graph / jackcoho_graph / steelhead_graph
tab1_comb
**Figure 2.** Time series of Coho, Jack Coho, and Steelhead counts on the Willamette fish ladder between 2001 and 2010.

Figure 2. Time series of Coho, Jack Coho, and Steelhead counts on the Willamette fish ladder between 2001 and 2010.

Time Series Plots Discussion

  • Steelhead are by far the most abundant fish in the Willamette River, or they are just better at accessing the fish ladder. Steelhead have seasonality of when they pass through the fish ladder, but some years have larger counts than others.There is no apparent trend in Steelhead counts as it varies year to year.

  • Both Coho and Jack Coho counts do not have the same seasonality that Steelhead have. Their counts are relatively low and there is a lack of an overall trend. It seems, though, that as more data gets collected, we may see some seasonality with their counts.

Tab 2 - Season Plots (Vanessa)

#explore total counts with season plots

#fish_month %>% 
  #gg_season(y = month_mean_count)

#----------#

#explore counts per species with season plots

#fish_ts %>%
  #gg_season(y = total_count)

#----------#

# seasonal subseries plot

#fish_month %>% 
  #gg_subseries(month_mean_count)

#----------#

#final sesason plot

fish_ts_longer %>% 
  gg_season(y = total_count) +
  theme_minimal() +
  labs(x = "Date", y = "Total Count", 
       title = "Total Counts for Coho, Jack Coho, and Steelhead at Willamette Falls, Oregon",
       subtitle = "2001-2010")
**Figure 3.** Season plots of Coho, Jack Coho, and Steelhead counts on the Willamette fish ladder between 2001 and 2010.

Figure 3. Season plots of Coho, Jack Coho, and Steelhead counts on the Willamette fish ladder between 2001 and 2010.

Season Plots Discussion

  • Observed seasonal patterns can be seen across all three species. Coho and jack coho have similar seasonal passage of the falls, whereas steelhead have a longer, more sustained seasonal passage.

  • Coho and jack coho have similar seasonal patterns with a serge occurring between late September and early October. Coho appear to have an increase in number of fish passing through in 2009 and 2010. Jack coho had a their highest count in 2008.

  • Steelhead trout pass through the falls over a much longer, sustained period between January through July. There is a slight rise in counts in May and June every year and then much less after July and August. In general, there are more steelhead passing through the falls each year than coho and jack coho.

Tab 3 - Annual Counts by Species (Tom)

fish_annual <- fish_longer %>% 
  mutate(year = as.Date(as.character(date),format = "%Y")) %>%
  group_by(year, species) %>% 
  summarize(fish_yearly_total = sum(value))

fish_annual %>% 
  ggplot(aes(x = year, y = fish_yearly_total)) +
  geom_line(aes(color=species)) +
  theme_minimal() +
  scale_x_date(date_labels = "%Y") +
  labs(
    title = "Annual totals for fish passage by species (2000-2010)",
    x = "Year",
    y = "Count of fish passage"
  )
**Figure 4:** The line graph above displays changes in fish passage by fish species between 2000 and 2010.

Figure 4: The line graph above displays changes in fish passage by fish species between 2000 and 2010.

Annual Trends Discussion

  • Steelhead observations appear to be most volatile over time with large spikes in passage over the first 5 years eventual stabilization from 2005 to 2009 followed by a large population spike.

  • Jack coho’s population remains relatively low and stable over time, seeing a small peak in population in 2008.

  • Coho population has mild volatile over the first 8 years followed by a spike in population grow after 2008.