For this exam, you will render this Rmarkdown document into a PDF as you have done with your labs. You can write your answers below each question and show your work using the code chunk. Note that small coding errors will cause your render to fail and Rmarkdown is not as easy to debug as a typical R script that can be run one line at a time. We recommend (a) rendering your exam before you start, and (b) rendering after each answer, and/or (c) using a separate script to debug your code and then transferring the working code into the Rmarkdown code chunk.
Read each question carefully and make sure that you answer completely. SHOW ALL YOUR WORK to get partial credit for incorrect answers. Use graphs and illustrations any time, but it will be clear if you need to use them. USE APPROPRIATE AXIS LABELS on your graphs. If you aren’t sure about a question, give it your best shot and explain what you did and why.
# risk of quasi-extinction is the probabilty that a population will decline to a critically low level population that is too small to recover. We can’t assume that extinction = 1 animal because there needs more than 1 individual to reproduce and even if a few individuals remain, they might not be able to reproduce successfully leading to extinction.
# The Lincoln-Peterson analysis estimates population size by comparing the proportion of marked individuals in the second sample to the total number of individuals captured in the second sample. It assumes that the population is closed (there are no births or deaths occuring and no immigration and emigration), that the process is randon and there is an equal ability to catch marked and unmarked animals.
# the difference between continuous and discrete rates of population increase lies in how population growth is measured over time. Discrete are calculated over specific time intervals while continuous is calculated as an instantaneous rate of growth at any given moment.
# a p-value helps determine the statistical significance of the population trends. A p-value of less than 0.05 suggests that it is significant. The R2 value indicates the strength and fit of the relationship.
# the intrinsic rate of increase is different from r because it shows the predicted changes while r is based on observed changes.
P0<- 300 #inital population
decline_rate <- 0.12
P1 <- P0 * (1-decline_rate)
P1
## [1] 264
# population of 264 animals after 1 year
P7 <- P0*(1-decline_rate)^7
P7
## [1] 122.6027
# population of 122.6 animals after 7 years
#lambda is population growth over time.
pop <- c(140, 170, 190, 228, 215, 251)
lambda <- numeric(length(pop) - 1)
for (i in 2:length(pop))
{lambda[i - 1] <- pop[i] / pop[i - 1]}
lambda
## [1] 1.2142857 1.1176471 1.2000000 0.9429825 1.1674419
mean(lambda)
## [1] 1.128471
#it tells us that the population is increasing over time because the value is greater than 1
r <- 0.07
t <- 0:3
N0 <- 100
N1 <- N0 * exp(r*1)
N1
## [1] 107.2508
N2 <- N0 * exp(r*2)
N2
## [1] 115.0274
N3 <- N0 * exp(r*3)
N3
## [1] 123.3678
N20 <- N0 * exp(r*20)
N20
## [1] 405.52
N0 <- 100
r <- 0.07
triple <- log(N0*3/N0)/log(1+r)
triple
## [1] 16.23757
#16 years
knitr::include_graphics("Exam1_Q9a.png")
#information you can obtain from the negative slope of this plot is that the population is decreasing over time.
#the shape of the plot would not be so linear because it would be showing a constant rate of decrease
SuckerCatch <- data.frame(Year=1995:2005,catch=c(29,71,16,5,40,42,8,2,11,1,6))
reg_log=lm(log(SuckerCatch$catch)~SuckerCatch$Year)
summary(reg_log)
##
## Call:
## lm(formula = log(SuckerCatch$catch) ~ SuckerCatch$Year)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.3275 -0.8013 -0.0525 0.7962 1.3376
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 538.66373 199.56816 2.699 0.0244 *
## SuckerCatch$Year -0.26813 0.09978 -2.687 0.0249 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.047 on 9 degrees of freedom
## Multiple R-squared: 0.4452, Adjusted R-squared: 0.3835
## F-statistic: 7.221 on 1 and 9 DF, p-value: 0.02491
#Call:
# lm(formula = log(SuckerCatch$catch) ~ SuckerCatch$Year)
#
# Residuals:
# Min 1Q Median 3Q Max
# -1.3275 -0.8013 -0.0525 0.7962 1.3376
#
# Coefficients:
# Estimate Std. Error t value Pr(>|t|)
# (Intercept) 538.66373 199.56816 2.699 0.0244 *
# SuckerCatch$Year -0.26813 0.09978 -2.687 0.0249 *
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# Residual standard error: 1.047 on 9 degrees of freedom
# Multiple R-squared: 0.4452, Adjusted R-squared: 0.3835
# F-statistic: 7.221 on 1 and 9 DF, p-value: 0.02491
coef(reg_log)[2]
## SuckerCatch$Year
## -0.2681318
# SuckerCatch$Year
# -0.2681318
summary(reg_log)$coefficients[2,4]
## [1] 0.02491362
# 0.02491362
confint(reg_log)
## 2.5 % 97.5 %
## (Intercept) 87.2091862 990.11826873
## SuckerCatch$Year -0.4938588 -0.04240484
# 2.5 % 97.5 %
# (Intercept) 87.2091862 990.11826873
# SuckerCatch$Year -0.4938588 -0.04240484
# the population is decreasing based on the coefficent (-0.26813). Because the p-value is less than 0.05 (0.0249) there is strong evidence to suggest that the trends are significant. The confidence interval also supports the trend of decreasing because it is from -0.4938588 and -0.04240484 meaning we are 95% confident that the population is decreasing from -0.4938588 and -0.04240484
lambda2 <- exp(-0.26813)
lambda2 #0.7648083
## [1] 0.7648083
percent_change <- (1-lambda2)*100
percent_change # 23.52 %
## [1] 23.51917
# Yes, based on this information this population should be given a special conservation status because the data is statistically significant in showing that the population is decreasing over time. It would also be beneficial to run an extiction risk time series to determine extinction risks and how the population should be listed.
#population abundances are associated with high overall growth rates with N=0, they are associated with lower overall growth rates when N=K
knitr::include_graphics("Exam1_Fig1.png")
# no it wouldn’t be a good method because this method assumes that as density increases growth rate decreases. The figure shows that as density increase, breeding success (therefore growth rate) is also increasing.
# The breeding success would eventually decreasing because the population woudl reach carrying capacity and there wouldn’t be enough room or resources for the population to grow.
# Breeding success generally decreases with colony density increase because there is higher competition for resources and mates. The term for this is negative density dependence.
The dataset includes the number of individuals caught each day (Catch), the number of those that are recaptures (Recaptures), and number of individuals newly marked (Newmarks). Hint: Does the Schnabel estimate use the NEW marked each day or some other metric? Make sure your results make sense given the data as a sanity check.
As a reminder:
n = sum(Mt*Ct) / sum(Rt)
Where: n = an estimate of the true number (N) of individuals in the population at the time of initial marking. Mt = the number of individuals in the population that have been marked at the time the t sample was taken. Ct = the number of individuals captured at the sampling that occurred at time t. Rt = the number of recaptured, marked individuals taken at time t.
eulachon <- data.frame(Month=1:5,Catch=c(29,34,28,20,25),Recaptures=c(0,14,10,8,9),Newmarks=c(29,20,18,12,16))
Ct <- eulachon$Catch
Rt <- eulachon$Recaptures
NM <- eulachon$Newmarks
Mt <- c(0, cumsum(NM)[1:5])
Mt
## [1] 0 29 49 67 79 95
est_N <- sum(Mt*Ct)/sum(Rt)
## Warning in Mt * Ct: longer object length is not a multiple of shorter object
## length
est_N #205.561
## [1] 205.561
Risk and IUCN status criteria. The International Union for the Conservation of Nature (IUCN) uses the following general criteria for “listing” species: http://www.iucnredlist.org/technical-documents/categories-and-criteria/2001-categories-criteria
“>=50% chance of extinction within 10 years or three generations, whichever is the longer (up to a maximum of 100 years) = list as ‘Critically Endangered’”
“>=20% chance of extinction within 20 years or five generations, whichever is the longer (up to a maximum of 100 years)= list as ‘Endangered’”
“>= 10% chance of extinction in 100 years = list as ‘Vulnerable’”
“<10% chance of extinction in 100 years = no listing”
Species or populations are always listed at the highest level for which they meet the criteria.
knitr::include_graphics("Exam1_Q12a_TS.png")
# the population should be listed as critically endangered because within 3 generations the population would be below 10%
knitr::include_graphics("Exam1_Q12a_QET.png")
# this population would not be listed becaus the population increased over time
knitr::include_graphics("Exam1_Q12b_TS.png")
# this population would be listed as vunerable
knitr::include_graphics("Exam1_Q12b_QET.png")
# this population would be listed as critically endangered