Part 1 - Introduction

The data analysed in this project was collected from the top 15 National Parks in America (NPS, 2025). Using this data, we analysed the proportion of nativeness of all bird species in each park versus all mammal species. Biologically, this is important as it allows for a comparison between classes of animals with different mobility levels, as birds tend to be more migratory than mammals. This gives an indication as to whether there are many migratory species using the national parks as a stopover, which would likely be the case if the proportion of nativeness in birds is higher than the proportion of nativeness in mammals. In the face of climate change, the conservation of these bird species is crucial as those that have not yet evolved to a changing environment are experiencing a population decline (Møller et al., 2008). Here, the explanatory variable is whether the species is a bird or a mammal, and the response variable is the calculated proportion of nativeness for each national park. Figure 1. Harlequin ducks that have migrated to Yellowstone National Park to breed (Winkler, 2024).

Figure 1. Harlequin ducks that have migrated to Yellowstone National Park to breed (Winkler, 2024).
Figure 1. Harlequin ducks that have migrated to Yellowstone National Park to breed (Winkler, 2024).

Part 2 - Main Research Question

Is there a difference between bird and mammal nativeness proportions in the 15 top most-visited national parks?

Part 3 - Exploring the Data (Descriptive Statistics)

NPS_df %>%
  group_by(CategoryName) %>%
  summarise(
    mean_proportion = mean(Proportion_Native),
    sd_proportion = sd(Proportion_Native),
    n = sum(Proportion_Native),
    se_proportion = sd_proportion / sqrt(n)
  )
## # A tibble: 2 × 5
##   CategoryName mean_proportion sd_proportion     n se_proportion
##   <chr>                  <dbl>         <dbl> <dbl>         <dbl>
## 1 Bird                   0.939        0.0916  14.1        0.0244
## 2 Mammal                 0.937        0.0568  14.1        0.0152
ggplot(NPS_df, aes(x = Proportion_Native, fill=CategoryName)) +   
    geom_histogram() + 
    facet_wrap(~ CategoryName, ncol = 1) +scale_x_continuous(labels = scales::percent_format()) + labs( title = "Nativeness Proportion of Birds and Mammals Distribution", x ="Nativeness Proportion") +scale_color_brewer(palette = "Set1")
## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(NPS_df, aes(sample = Proportion_Native, color=CategoryName)) +
  stat_qq() +
  stat_qq_line() +
  facet_wrap(~CategoryName) + scale_fill_brewer(palette = "Set1") +labs(title="Q-Q Plot of Nativeness Proportions of Birds and Mammals")

ggplot(NPS_df, aes(x=CategoryName, y=Proportion_Native, fill = CategoryName)) +geom_boxplot()+geom_jitter()+theme_classic()+scale_y_continuous(labels = scales::percent_format())+labs(title = "Nativeness Proportions of Birds and Mammals", x="Animal Class", y= "Nativeness Proportion")+theme_pubr() +
  scale_fill_brewer(palette = "Set1") +
  theme(legend.position = "none")

Part 4 - Statistical Tests (Inferential Statistics)

HNull: There is no difference in proportion of nativeness between birds and mammals. Halt: There is a difference in proportion of nativeness between birds and mammals. Our data is not normally distributed due to extreme outliers. We tried to transform the data by performing a log transformation but due to the extreme outliers we were unsuccessful. We checked for equal variance and found that the data has equal variance. Our data fits the requirements of a Mann Whitney U-test.

shapiro.test(NPS_df$Proportion_Native)
## 
##  Shapiro-Wilk normality test
## 
## data:  NPS_df$Proportion_Native
## W = 0.6685, p-value = 5.602e-07
leveneTest(Proportion_Native~CategoryName, data = NPS_df)
## Warning in leveneTest.default(y = y, group = group, ...): group coerced to
## factor.
## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  1  0.0327 0.8579
##       28
NPS_df$LogProportion <-log(NPS_df$Proportion_Native)
ggplot(NPS_df, aes(x = LogProportion, fill=CategoryName)) +   
    geom_histogram() + 
    facet_wrap(~ CategoryName, ncol = 1)+
  scale_fill_brewer(palette = "Set1") +
  labs(
    title = "Distribution of Log-Transformed Native Proportions",
    x = "Log(Proportion Native)") +
  theme_pubr()
## `stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(NPS_df, aes(sample = LogProportion, color=CategoryName)) +
  stat_qq() +
  stat_qq_line() +
  facet_wrap(~CategoryName) +scale_fill_brewer(palette = "Set1")+labs(title = "Q-Q Plot of Log-Transformed Native Proportions")+theme_pubr()

wilcox.test(Proportion_Native~CategoryName, data=NPS_df)
## Warning in wilcox.test.default(x = DATA[[1L]], y = DATA[[2L]], ...): cannot
## compute exact p-value with ties
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  Proportion_Native by CategoryName
## W = 144.5, p-value = 0.1913
## alternative hypothesis: true location shift is not equal to 0

There is not a difference in bird and mammal nativeness proportions in the top 15 visited National Parks (W=144.5(df=1),p=0.1913).

Part 4 - Discussion

There is no significant difference for the proportion of nativeness between birds and mammals (W=1.44.5(df=1), p=0.1913). We therefore fail to reject the null hypothesis. This analysis is important as it is a likely indication that there are low numbers of migratory birds in the top 15 National Parks in the US, which is important for conservationists of these species, so they can know where to direct their efforts. A major limitation of this analysis is that there are some outliers in the data where the proportion of nativeness is significantly higher in birds, where conservation efforts should definitely be directed. For example, Arcadia National Park has a lower proportion of nativeness (66%) of bird species than any other national park, and it is lower than the proportion of mammal

species (77%). This is because Arcadia National Park is located on the Atlantic migration flyway (Pittsfield, 1980). To truly use this data for its intended purpose, the categorical data (species present at each national park) should be examined to truly see where migratory birds are located. This would also need to be monitored at various times of the year, during different stages of migration.

Part 5 - Conclusion

Overall, there is no significant difference between the proportion of nativeness of birds and mammals in the top 15 national parks in the US. Due to this, we fail to reject the null hypothesis. Although conservation of migratory bird species in these areas is still crucial, in general it is of no greater importance than in other areas along migration flyways.

Part 6 - References

Møller, A.P., Rubolini, D., Lehikoinen, E. (2008). Populations of migratory bird species that did not show a phenological response to climate change are declining, Proc. Natl. Acad. Sci. U.S.A. 105 (42) 16195-16200. https://doi.org/10.1073/pnas.0803825105

National Park Service. (2025). NPSpecies: Information on species in national parks. U.S. Department of the Interior. https://irma.nps.gov/NPSpecies/

Pittsfield, M. A. (1980). Atlantic Flyway Review: Region I. North American Bird Bander, 5(2). https://digitalcommons.usf.edu/nabb/vol6/iss2/9

Winkler, C. (2024). Wonders of Bird Migration. Buffalo Bill Centre of the West. https://centerofthewest.org/2024/03/22/wonders-of-bird-migration/