document attached
DGS2018_MetapopulationExercise 1 Landscape Genetics Conceptual Exercise (Nusha Keyghobadi) Concepts in metapopulation genetics In this series of conceptual exercises, we will consider the effects of metapopulation structure, and extinction-colonization dynamics, on the genetics of populations. We will also briefly explore the concept of patch connectivity within a metapopulation framework. Part 1: Effects of metapopulation structure and extinction-colonization dynamics Consider a group of ponds that are the breeding habitat for the freshwater snail (Drepanotrema depressissimum). In the diagrams below, each pond (habitat patch) is represented by a circle. Movement of individuals (and gene flow) between patches is represented by arrows. Double-headed arrows indicate movements and gene flow in both directions. And the thickness of the arrows represents the probability and amount of movement or gene flow (thicker arrow = more individuals/genes moving). Consider the three different scenarios depicted below. Comparison 1 Scenario A Scenario B Scenario C Question 1. Which scenario corresponds best to: (i) a classic metapopulation, (ii) isolated populations, and (iii) a patchy population? Question 2. What are some other terms that might be used to describe the system depicted in Scenario B? (either from the ecological or population genetics literature) 1 3 3 2 4 4 1 2 3 1 4 2 Becky Cruz week four concepts in metapopulation genetics 2 Question 3. Rank the three scenarios (from highest to lowest) in terms of the properties listed in the table below. ‘System’ refers to the group of four habitat patches (and their resident populations) in each scenario Highest Intermediate Lowest Global FST (FST for the entire system; neutral loci) Global NE (effective population size of the entire system) Mean genetic diversity within individual patches (neutral loci) Comparison 2 Now consider the two scenarios below. In scenario A, the habitat patches are continually occupied and local populations never go extinct. In scenario B, local populations sometimes go extinct leaving behind an unoccupied habitat patch (illustrated by dark coloured habitat patches), which may eventually be recolonized (single-headed arrows represent the potential for colonization of empty patches). Individuals that colonize an empty patch usually all come from the same patch. Scenario A Scenario B Question 4. Rank the two scenarios in terms of the properties listed in the table below. Higher Lower Global FST Global NE Mean genetic diversity within patches 3 1 4 2 3 2 3 Comparison 3 Now we have two new scenarios (below). In both cases, the systems experience local extinctions and recolonizations. Individuals that colonize an empty patch usually all come from the same patch. However, in Scenario B, local populations are more likely to go extinct (indicated by the darker colour of unoccupied patches) and empty, unoccupied patches are less likely to be recolonized (indicated by the narrower arrows in Scenario B). Scenario A Scenario B Question 5. Rank the two scenarios in terms of the properties listed in the table below. Higher Lower Global FST Global NE Mean genetic diversity within patches 3 2 3 2 4 Comparison 4 In the two scenarios below, both systems also experience local extinctions (dark coloured patch) and recolonizations (arrows). Here, the overall probabilities of local extinction and recolonization are identical in the two scenarios. But, the scenarios differ in the sources of individuals that recolonize empty habitat patches. In Scenario A, colonists come from across the entire system. In Scenario B, colonists all come from a single patch. Scenario A Scenario B Question 6. Which of the scenarios above best fits the ‘propagule pool’ model and which best fits the ‘migrant pool’ model? Question 7. Rank the two scenarios, in general, in terms of the properties listed in the table below. Higher Lower Global FST Global NE Mean genetic diversity within patches Question 8. Compared to a scenario with no extinctions and recolonizations, in which of the two scenarios above is it possible to have a lower global FST (i.e., lower levels of differentiation among local populations)? 3 2 3 2 5 Comparison 5 Now, both systems experience extinctions and recolonizations at the same rate, and both systems are characterized by the same patterns of recolonization (i.e., founding individuals come from several different patches). The number of founder individuals (k) that recolonize an empty patch (relative to the average population size in occupied patches, N) is much higher in Scenario A than in Scenario B. Large founding groups Small founding groups (k is large) (k is small) Scenario A Scenario B Question 9. Rank the two scenarios, in general, in terms of the properties listed in the table below. Higher Lower Global FST Global NE Mean genetic diversity within patches Question 10. Among the hypothetical scenarios above, we changed variables such as: (i) the probability of extinction of local populations, (ii) the probability of colonization of empty patches, (iii) the probability and rates of movement between occupied patches, and (iv) the sources of founders in colonization events. Think about two or three ecological/landscape factors that might lead to changes in each of these variables in real populations of the snail Drepanotrema depressissimum (there might be overlap in the 3 2 3 2 6 factors that affect these different variables, but try to think of as many different things as you can). Part 2: Patch connectivity within a metapopulation framework Compared to the scenarios illustrated in Part 1, below is a more realistic illustration of what a metapopulation might actually look like – with varying distances among patches and varying patch sizes. Scenario 1 The Incidence Function Model (IFM) is a patch-occupancy model for metapopulations that allows for this more realistic representation of the system, with patches that vary in size and isolation. One metric used in the IFM is patch connectivity (inverse of isolation). One way to define the connectivity (Si) of any patch in the system (i) is as: ?! = exp −??!"! !! ?!?! where the summation is over all other patches (j) in the network, dij is the distance between patches i and j, α is a parameter that describes how dispersal is affected by increasing distance between patches, pj is the probability of patch j being occupied, and Aj is the area of patch j. Basically, the connectivity of any patch depends on the extent to which it can receive migrants from all the other patches in the system. This is a function of its distance to all the other patches in the system (assuming dispersal decreases as distance increases), and on the occupancy state and size of all those other patches (since only an occupied patch can send out emigrants, and assuming that patch size is positively correlated with the size of the local population). 7 Question 11. Compare Scenario 1 (above) to Scenarios 2 and 3 (below). Consider patches ‘A’ and ‘F’. How does the connectivity of each of these patches change among scenarios? Why? Scenario 2 Scenario 3 Question 12. How do you think genetic diversity of local populations will be related to patch connectivity? For Scenario 1, in which patch do you predict you would see the lowest genetic diversity? And the highest? Question 13. How do you think the genetic differentiation of local populations (from other local populations in the system, on average; e.g., measured using site-specific FST) will be related to patch connectivity? For Scenario 1, in which patch do you predict you would see the most highly differentiated population? Question 14. The equation for connectivity given above only considers patch sizes and distances. Think of at least two ways you might modify the equation to incorporate more information about the landscape and its effects on immigration (this can include information about the patches themselves, or about the intervening landscape).