Lab 2
Coral Levi
Questions from Part III
Question 2 (from page 4).
The mean of survivors in the start of the experiment was higher on the experimental islands than on the control islands. However, at both the 6 month and 12 month marker the mean survivors on the control islands were much higher than those on the experimental islands.
Question 4.
The anoles are more likely to survive on the control islands than on the experiment islands.This is most likely due to the lack of predation. Since the predator wasn’t on the control islands, the anoles had less pressure on their population size. The factors that affected the population size on the control islands were also present on the experimental islands.
Question 5.
Other factors that can affect population size on both control and experimental islands are based on limited resources, be it limited food, space, etc.
Question 6.
Similar to the mean of total survivors, the means of ground survivors was originally higher on the experiment islands and then reduced dramatically. After 6 and 12 months the mean of ground survivors on the control islands was almost six times larger than the mean on the experiment islands.
Question 7.
Complete the code below, do not answer the parts of question 7.
Load data
#Collect data, save file as "anole.csv"
# Set working directory or go find your data sheet.
getwd()## [1] "C:/Users/levic/Desktop/Evolution"#setwd("C:/Users/levic/Desktop/Evolution")
getwd()## [1] "C:/Users/levic/Desktop/Evolution"#make sure "anole.csv" is in your working directory, and that this
#rmd file is in there too!
#Enter the data
anoles <- read.csv("anole2.csv")
attach(anoles)
names(anoles)## [1] "time" "island" "treatment"
## [4] "total.survivors" "ground.survivors"#Should have columns for time, island, treatment, total.survivors, ground.survivors
#NOTE: YOUR TABLE SHOULD LOOK A LOT LIKE THIS:
# time island treatment total.survivors ground.survivors
#1 0 A control 34 20
#2 6 A control 31 10
#3 12 A control 27 15
#4 0 B experimental 30 12
# ... MORE ROWS ...
# ... MORE ROWS ...
# ... MORE ROWS ...Generate summary statistics
#Find the means (table 1)
tapply(total.survivors, list(time, treatment), mean)## control experiment
## 0 29.00 44.0
## 6 18.75 11.0
## 12 15.00 5.5#Calculate survival rates. Insert your own code to find the rates.
control6mo <- 18.75/29.00 #mean 6mo/mean 0mo
control12mo <- 15.00/29.00 #mean 12mo/mean 0mo
experiment6mo <-11.00/44.0 #mean 6mo/mean 0mo
experiment12mo <- 5.5/44 #mean 12mo/mean 0mo
#Calculate the proportion of survivors on the ground
proportion <- ground.survivors/total.survivors
#Find the means (table 2)
tapply(proportion, list(time, treatment), mean)## control experiment
## 0 0.5038371 0.5015610
## 6 0.4029895 0.1174451
## 12 0.4625000 0.0750000Create models
#set up subsets of the data
time_cnt <- time[treatment=="control"]
time_trt <- time[treatment=="experiment"]
prop_cnt <- proportion[treatment=="control"]
prop_trt <- proportion[treatment=="experiment"]
#Plot the data
plot(time_cnt, prop_cnt, xlab="Time",
ylab="Mean Proprotion of Ground Anoles",
pch=19, col="blue", cex=1.5, ylim=c(0,.6))
points(time_trt, prop_trt,pch=19,
col="red",cex=1.5)
#Draw a legend
legend("topright", c("control","experiment"),
pch=19, col=c("blue","red"))
#Run linear regressions
#technically we should use logistic regressions because the dependent variable
#is a proportion, but a normal linear regression will do for illustration
#fit linear model for data from control
controlmod <- lm(prop_cnt ~ time_cnt)
abline(controlmod,col="blue",lwd=2)
#complete the model below with data from the experiment
expermod <- lm(prop_trt ~ time_trt) #LINEAR MODEL FOR DATA FROM EXPERIMENT
abline(expermod,col="red",lwd=2)
#look at the results of the regressions--is the slope significant for either treatment?
summary(controlmod)##
## Call:
## lm(formula = prop_cnt ~ time_cnt)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.13386 -0.07928 -0.02369 0.11209 0.12895
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.477111 0.049038 9.729 2.04e-06 ***
## time_cnt -0.003445 0.006331 -0.544 0.598
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1074 on 10 degrees of freedom
## Multiple R-squared: 0.02876, Adjusted R-squared: -0.06837
## F-statistic: 0.2961 on 1 and 10 DF, p-value: 0.5983summary(expermod)##
## Call:
## lm(formula = prop_trt ~ time_trt)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.154412 -0.092942 -0.001335 0.050661 0.181945
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.444616 0.049281 9.022 4.05e-06 ***
## time_trt -0.035547 0.006362 -5.587 0.000232 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.108 on 10 degrees of freedom
## Multiple R-squared: 0.7574, Adjusted R-squared: 0.7331
## F-statistic: 31.22 on 1 and 10 DF, p-value: 0.0002318Question 8.
Both the control and experiment islands both start with a high proportion of ground anoles. Though there is a decrease in the mean proportion of ground anoles on both sets of islands, there is a greater decrease on the experiment islands. The mean proportion on the control islands decreased minimally after both 6 and 12 months.On the experiment islands there was over 0.2 decrease in the mean proportion of ground anoles after the first six months. In two thirds of the experiment groups there was an additional decrease between 6 and 12 months and a slight increase on for the other third.
Question 9.
The data shows a decrease in the mean proportion of survivors on the ground on the experiment islands. This suggests that the hypothesis is correct. Since the mean proportion of survivors on the ground is so low it suggests that the majority of survivors live in bushes or trees (not on the ground).
Question 10.
Compared to the control islands, on the experimental islands fewer anoles from the intial population survived, and fewer survivors were living primarily on the ground.
Questions from Part IV
Question 11 (from page 7).
Long legs= ground, short legs= bushes and trees Start of the experiment: Expected average hind limb length at the begining of the experiment on both control and experiment islands would be long. Since predators were just introduced on the experiment islands at the begining of the experiment, there was no time for selection to occur to change average hind limb length.*
After 6 Months:Average limb length after 6 months on the control islands would remain the same, long since nothing in the environment was changed. On the experiment islands expected limb length would be shorter. The predators can’t climb so the anoles with shorter legs that climb would have a better survival rate and therefore the number of anoles with shorter hind limb length would be greater than those with longer hind limb length.
After 12 months: At 12 months, the control island anoles would be expected to be the same as the beginning with longer hind limb lengths. On the experiment islands hind limb length would be expected to be short because of their shift from ground dwelling to tree dwelling.
Question 12.
Dr. Lsos and his colleagues discovered that on the experiment islands, hind limb length after six months decreased as the anoles shifted to the bushes and trees to avoid predators. However, after 12 months, anoles with longer legs seemed to survive better.
a. These findings are different from what I expected. My expectation would be to see a greater shift to the bushes and trees leading to an overall decrease in the average hind limb length. Although this was the intial trend, at the 12 month marker the average hind limb length on the experiment islands was longer than the control islands.
b. Dr. Losis explained this trend stating that initially the anoles shifted to the trees to avoid the predators. At 12 months the anoles with longer legs had the highest survival rate. They lived on the ground but were faster than the predator and therefore were more successful at avoiding the predators than the climbing anoles.
Question 13.
a. Dr. Losos measured hind limb length and found there was variation in the trait. Some anoles had long hind limbs while others had short hind limbs.
b. Dr. Losos were unable to support the claim that variation in the trait was heritable because they only measured the trait in one generation. In order to prove the trait is heritable the researchers would need to measure the trait through several generations of parents and offspring.
c. Dr. Losos and his collegues showed that some anoles had a fitness advantage over other anoles through the change in mean survival rates of anoles on the ground in the experiment islands.
d. Dr. Losos showed that natural selection favored certain trait variation at both the 6 and 12 month mark on the experiment islands. At six months they saw that natural selection favored shorter hind limb lengths and living in the tree while at 12 months they saw that selection favored longer hind limb length that increased the anoles speed.
e. Dr. Losos was unable to support the claim that beneficial traits were passed on to future generation because he and his collegues only measured the trait variation in one generation. They would need to measure the trait variation and changes over several generations. Only through this would he be able to support the claim that the population changed as lizards with features better adapted to living on trees evolved.
Question 14.
If Dr. Losos had been able to continue their experiment I expect that they would have seen a change in hind limb length over generations. If one or more of the experiment islands had more bushes and trees than others, I would expect to see limb length on those islands to decrease for better climbing. However, on islands with less bushes and trees hind limb length would probably increase because speed would be more beneficial than climbing for getting away from predators on those islands.The proportion of ground dwelling anoles would decrease on the islands with more bushes or trees and increase on those with fewer bushes and trees. If the islands were identical in topography the expectation is that the populations on all the experimental islands would be similar in genotype and phenotype be in short or long hind limb length.