FORESTS 2018 QUESTIONS ONE: EXAMPLE ANSWERS/DISCUSSION


These are open-ended questions.  This will be the norm for this class.  I won't be asking for definitions, giving multiple-choice, etc..  But you should use the terminology and concepts we develop in class, and your answers will be evaluated, in part, on how well you do that (although evaluation will be primarily on the basis of a) how clearly and fully you address what's asked and b) how well you build your arguments.

Don't wander;  suitable answers will almost always be possible in a paragraph of a few sentences and often less -- so it's important to thnk carefully about what, specifically, is being asked, and to keep your own answers sharply focused.

1. A question for working with hypothesis generation and testing: Review notes on the ecology of Lyme disease in the notes from the first class. Researchers studying the Lyme disease epidemic have offered several models (hypotheses) proposing ecological linkages in a network of causes and effects that might regulate Lyme disease -- either promoting or limiting its spread. They have compiled data to assess or test some of these hypotheses, but most have been only partially tested and some, while plausible, remain essentially untested. Briefly (a paragraph or so -- no more!), propose further research focused on better understanding some part of this network of interactions. 
    a) Identify and state a research question that's related to some part of this system, framing it as a hypothesis you might be able to test (this could be something that's directly suggested -- but not pursued -- in material discussed/presented, or something that's not specifically mentioned, but is suggested to you by that material).
    b) Suggest
at least one testable prediction (could be more than one!) of that hypothesis (“If this hypothesis is true, then we should see....”).  
    c) Suggest a basic, plausible approach to
testing that prediction; this could be pretty general and rough (“I would collect data of this sort and see if I see the pattern predicted....”). 
    Refer to data-graphics or studies cited in the class notes as appropriate to support your research question and design. 
Be concise; focus on the essential points. 

Endless possibilities here, of course; the main things to focus on  are:  1) refining your hypothesis so that it states with a clearly defined cause-effect relationship you're suggesting -- a mechanism (i.e., not just "x should be related to y", or 'this happen if I did that', but incorporate some explanation as to why!), 2) and suggesting a prediction that is a necessary consequence of your hypothesis  and that offers a lever for testing it rigorously.  That last point is the trick; a goodprediction for this purpose is one that would be true if your hypothesis is correct but NOT if  some alternative mechanism applied.  In a number of cases, you offered predictions that were consistent with your hypothesized cause-effect model -- but could also result from other processes as well (this is a 'control' issue; think about how you can be confident that IF you get the results you expected/predicted, that they're due to the reasons you've hypothesized).. Then 3), when you suggest ways of testing this prediction, think in terms of critical comparisons (where results would be different) if hypoth true or not.  Also, always try to think in terms of  active manipulations/experiments (not just comparisons of 'natural' situations) that would make your approach more powerful.  A lot of suggestions involved simply monitoring what happens and looking for correlations between things (like "more Lyme disease in years with more acorn production") but how can you be confident that the 'more Lyme' is because of the more acorns (maybe big mast years for oak follow wet summers, and it's actually weather that drives tick populations...).  Yourapproach could  be much strengthened by doing some active 'tweaking' (like adding acorns or removing acorns and seeing if hypothesis' predictions still happen...).   HERE, the IMPORTANT thing is to focus on how to go from an observation or question to a USABLE HYPOTHESIS and then to think rigorously and concisely about how to address/test it.

2. Thinking about island biogeography and diversity: Recall the general relationships between colonization rate, extinction rate, island distance from a source of migrants (= isolation), and island size in the "equilibrium biodiversity" in the MacArthur-Wilson model of island biogeography (refer to online notes for summary; remind yourself particularly of the graphical version of the model). As we discussed briefly, other types of habitats occurring in isolated patches, surrounded by very different habitats might be island-like enough to be described by the same model. Fresh-water ponds, for example, amount to ‘islands’ of distinct habitat isolated, to some extent, in a terrestrial landscape.  Thus, it might make sense to try to apply the MacArthur-Wilson equilibrium model to predicting differences in diversity among ponds ponds for aquatic groups. However, translation of the ‘distance effect’ on colonization rate may require careful thought. How might you restructure that part of the model to make it applicable for bodies of fresh water?  Are there differences in the way you need to think about 'isolation'?  What factors might influence colonization rate, and how might you express them in  colonization curves as in the original M-W model?  Ponds host organisms of many types (algae, birds, fish, plants, many groups of insects, frogs..); some spend their entire life-cycle in the water (e.g., fish, many small invertebrates, algae), while others may be only partially aquatic (like pond turtles), or be able to move between ponds at some point in their life-cycle (like frogs, dragonflies, that are strictly aquatic as larvae, but can disperse as adults). How might your thinking about factors constraining or promoting the 'colonization curve' affect these groups differently (choose two or three to contrast)?  How might this affect your predictions about diversity patterns for those groups (i.e., how would diversity patterns relate differently to degree of isolation)? 

Main thing her is to recognize that, while 'colonization rate' in the M-W model is approached as primarily related to distance from mainland (or, more generally, any source of new colonists), that's really just a way of representing degree of isolation; more remote islands have lower colonization rates, thus lower equilibrium diversity.  If you're looking at other isolated habitat patches as 'island-like' the question becomes, 'can I describe degree of isolation in terms of some other variables?'.  If you can, then, in principle, the same model structure can be used -- you just 'code' isolation differently.  Oceanic islands and ponds are both types of habitats isolated in a background of very different habitat; applying the colonization model to ponds, however, offers a couple of challenges: there's usually no simple 'mainland' equivalent for freshwater bodies (unless you happen to be near a very large lake or something).  That's one thing you could talk about.  Also, the ways in which aquatic creatures can disperse across land vary a lot, and some simply can't.  The first problem might be addressed by coming up with some measure of proximity of any larger body of water (or number of nearby ponds).  The second problem may be trickier but more interesting.  For groups that have parts of life-cycle where they can fly (like dragonflies) or are fully terrestrial (like toads), ponds may be, in effect, not particularly isolated at all -- not islands in the M-W sense -- and so diversity might be little affected by distance from other aquatic habitat, and the M-W model wouldn't really apply!  However, for organisms that require on aquatic (or very wet) habitat during all of their life-cycle (fish, some amphibians, algae, ...), isolation should certainly affect colonization rates and so diversity.   But simple distance between ponds might not be the best measure of  isolation; two nearby ponds with a big, dry ridge between might be more isolated than two that are much further apart, but connected by a stream.  You might come up with some  measure of 'connectedness' as functionally equivalent to 'remoteness' of an island from mainland -- are there wetlands or streams that connect the 'target' pond to others? How long/large are they? How persistent/reliable?   Some of you considered treating movement between ponds  by fishermen, boats, even birds or moose as possible mechanisms of dispersal for organisms that can't fly or walk across land themselves; this is certainly plausible and one of you suggested using  frequency of boat launchings as another index of isolation, comparable to distance in the standard M-W model...  (but note: this question asks about a) the applicabiity of the M-W model and b) specifically of the colonization-as-function-of -isolation part of the model.  It's not about extinction rate or island/pond size or other things not related to these two points.  Some of you talked a lot about other stuff, and what you said might have been reasonable -- but not really pertinent to building a good answer to this question!)