Data
Data Table
Table 1. Data table showing the research variables including predictor, response, and other design variables.
The predictor variable for this project is Mean Annual Temperature (MAT) (expressed as degrees Celsius) as a continuous variable, measured through time from 1992 to 2019, and the response variable is number of salmon harvested broken down by species (Chinook, Coho, Fall Chum, Pink, and Summer Chum). Design variables include unique identification (ID), Section, District, Community (COMM_NO and COMM_NAME), and GPS data for weather station locations.
Exploratory Graphics
First we looked at raw harvest counts for each species of salmon per community for all years data had been recorded (1992 to 2022). The resulting graph (Figure 7), though messy, does have visible downward trends in harvest counts for most of the salmon species, especially within the past 5 years. This is better shown in Figure 8 after trend lines were added based on each community's average harvest count per year.
Harvest Over Time by Species
MAT Over Time
Mean annual temperatures were gathered from all communities where weather station data was available. This data was grouped into three sections, Yukon Coastal District, Lower Yukon, and Upper Yukon indicating general location along the Yukon River. Overall the trend is very similar for all regions with average annual temperatures rising more than 2˚C, though local average temperatures differ between locations.
Other Factors to Consider
Number of households with fishing permits varied between communities (Figure 9) and years. For the years 2021 and 2022, there was a severe drop in the number of households with fishing permits (Figure 10). If we ignore these two last years, the trend had been that number of households with fishing permits was increasing (Figure 11). There was no explanation provided in the data or Alaska Department of Fish and Game reports for why number of fishing permits declined so severely. A pandemic effect is possible among other factors, but are all speculation and would require further insight.
Figure 9. Boxplot of number of households with fishing permits for each community for all years data were collected.
Figure 10. Number of households with fishing permits from all communities for the period from 1992 to 2022.
Figure 11. Number of households with fishing permits from all communities for the period from 1992 to 2020 only.
The government also put salmon fishing restrictions in place for the majority of years from 1992 to 2022 when harvest information was recorded (Table 2). These restrictions were established based on population dynamics for each of the species and were intended to conserve populations.
Table 2. Comments for each year restrictions were placed on salmon harvests.
Since the total number of fishing households for 2021 and 2022 are significant outliers from the rest of the data, we removed these years from analysis. Figure 12 shows harvest trends for each species of Salmon from 1992 to 2020 only.
Harvest Levels by Species
Figure 12. Line graph showing total number of salmon harvested per year (1992 to 2020 only) by each community for each of the 5 species studied. Y-axis is shown as a logarithmic scale. Trend lines are shown based on community mean harvest levels.
There is significant variation year to year for harvest levels which led to us convert total harvest counts into catch per household to account for some of the variability that we anticipated was due to changes in number of households participating in subsistence harvests each year.
A scatter plot with regression lines depicting the relationship between mean harvest and Mean Annual Temperature (MAT) provides a visual representation of potential correlations. The plot reveals distinct patterns for each salmon species, offering insights into the influence of temperature on harvest levels. The regression lines help identify trends, showing whether there is a consistent increase or decrease in mean harvest as MAT changes
Figure 13. Scatter Plot of MAT vs Mean Harvest
