BIOGEOGRAPHY ETC. 2018
FIRST Questions Set

DUE: October 1

NOTE that these are all open-ended sorts of questions.  Don't go overboard.  A medium-length paragraph should allow for sufficient answer on each.  Try to assess what's most important in each case, then construct an argument focusing on what's asked!  (They're mostly not 'look-upable' either; do your best working with the ideas and concepts we have discussed/developed.  Part of the point is to apply approaches and ideas in novel contexts...)

TWO QUESTIONS FOR WORKING WITH BACKGROUND IDEAS about ecology and natural selection

 1. Ecologists typically describe species as either 'specialists' or 'generalists'.  Specialists tend to have relatively specific resource or environment requirements (narrow fundamental niches), but adaptations that make them very effective in competing for these resources or very effective at dealing with environmental constraints.  Generalists have broad fundamental niches -- can use diverse resources, tolerate a wide range of habitat conditions -- but are not particularly effectively adapted to specific conditions (so likely to be competitively excluded by specialists from situations within the specialist's specialty).  Many biologists have suggested that Darwinian selection ought to produce progressively more sophisticated specialization (Dawkins says more specialized 'survival machines' (phenotypes) that serve as vehicles for replicators (genes) that build them).   Of course, more specialized organisms are more vulnerable, over the long term, to extinction as environment changes, but that's not a problem that's generally 'visible' to selection (i.e., risk of eventual extinction doesn't affect individual fitness differences now).  Propose a scenario where selection might favor a more generalist phenotype/genotype over a more specialized one -- i.e., where a more generalist individual might have higher fitness.  What sorts of traits ought to 'go with' a generalist phenotype?  Developmental plasticity -- where the same genotype can produce different phenotypes, but phenotype is 'locked in' over development -- can be regarded as a form of generalization (same genotype can do different things...); does your model suggest insight into adaptive costs and benefits of plasticity? Explain.

Most of you suggest that generalists should be selectively favored in more variable habitat, which is a plausible start -- but not quite sufficient.  For example, if variation is predictable (e.g., strict seasonal change), the organism could simply evolve it's own seasonal phenological change, in effect being a specialized for that particular kind of variation.  This is not so plausible if the variation is RAPID and/or UNPREDICTABLE, so individual fitness of a more 'generalist' creature is likely to be higher if it's more generalist in a rapidly and unpredictably varying habitat.  There's also an area-based line of argument; specialists can only survive if the habitat they're specialized is present and extensive/abundant enough to support a viable population.  SO, if a habitat is spatially variable enough, generalists might be favored (because they can travel between 'patches' of different sorts of habitat and exploit them all) over a specialist whose particular needs are met in patches so scattered/isolated that they can't efficiently use them. Generalist traits can include lots of stuff; ability to use a wide array of food types/sizes, tolerance of wide range of environmental conditions, etc.  Finally, developmental plasticity, where  organism can respond to environmental context early in life, but, at some point is 'locked in', has some real trade-offs/costs in a continually varying environment -- but might be quite effective adaptation when individuals disperse into habitats where they may then spend rest of life...

2. We have discussed how competitive interactions among organisms may shape their adaptations to reduce overlap in resource use. One result of this is that the niches (range of resources used) of organisms within the same ‘ecological guild’ tend to be separated, or ‘spaced out’ along a gradient of resource quality or type (e.g., size of seeds eaten by granivorous mice and ants). You can ‘place’ more organisms along such a gradient or axis if niches are narrower; conversely, broader fundamental niches will intensify competition and tend to eliminate some species. Imagine a guild of ant species using a range of seed sizes; each species can exploit a certain range of seed sizes, depending on mandible size -- a heritable trait -- and mandible size varies within species. The total range of seed sizes available does not change over time, but other properties of the environment might change. Suggest at least one way you might change environmental circumstances to ‘permit’ more species to persist in coexistence – i.e., ways to allow tighter ‘packing’ of niches without extinction of populations; you can invoke likely selective responses of species (i.e., evolutionary change in mandible size distributions within species).

A few possibilities here. 1) Increasing variety of resources (increasing size of 'resource space') might be the most straightforward change (e.g., more species of plants in an area woud likely allow increased diversity of plant-eating insects specializing on different host species); 2) Increasing AMOUNT of resources, even without increasing variety of resources, might allow tighter 'niche packing' for less obvious reasons; species persistence is more likely the larger its population, so increasing total resource base would likely allow rare specialists that might go extinct in sparser environment to persist; 3) Increeasing stability of environment could have similar effect by decreasing likelihood of extinction of species present.  In both 2 and 3, you might expect selection to drive competing species in environment towards increased specialization (which usually means they're superior competitors for a narrower range of resources -- and that's not a big selective costs when world is stable OR resource-rich...)


ISLAND BIOGEOGRAPHY AND SPECIATION:

3. If two sibling lineages are separated geographically, they are likely to diverge evolutionarily.  However, there is no inherent reason why the two populations would develop, while allopatric, a reproductive isolating mechanism (RIM) except by accident (e.g., if they become so different they don't recognize each other as potential mates, or if they've changed morphologically so that mating is impossible).  Suppose that  two such sibling populations have diverged in allopatry due to different types of environmental selection so that they now have significant adaptive phenotypic differences.  Eventually, contact is re-established but there is no RIM - they are still capable of interbreeding (maybe think polar bear and grizzly bear).  Is there reason to anticipate that an RIM might develop secondarily? (AFTER contact is re-established), as a consequence of selection? What would be required for this to occur? Explain (use fitness terminology correctly).

Assuming the two separated sibling lineages experience somewhat different selective regimes (almost a certainty; environments are unlikely to be identical, and different adaptive mutations may accumulate in each even if they were), the two lineages are likely to diverge in ways having to do with their specific adaptations to environment -- to develop different ways of making a living.   Upon secondary contact, individuals from the two lineages are able to interbreed with fertile offspring (given in question -- there is no RIM at this point) -- BUT, if the two parents have significantly different ways of making a living (distinct, integrated sets of adaptations -- different fundamental niches), the hybrid offspring may have a mish-mash of traits that won't allow them to be as effective in exploiting resources or getting mates or whatever as either parental type (a hybrid between a screwdriver and a hammer may not be very good at driving nails OR screws).  As most of you noted, such hybrid offspring would have reduced fitness compared to 'purebreds'; if their fitness were REALLY low (near zero), this could amount to a post-zygotic RIM.  However, more fruitful to think of FITNESS OF PARENTS.  IF hybrids are poorly adapted, individuals that have ANY HERITABLE TENDENCY to mate with 'like' individuals (say a new mutation with that effect) will have offspring more likely to survive and themselves reproduce -- higher fitness -- because their offspring retain the integrated set of adaptations that define one successful life-history. Any heritable trait that confers higher fitness consistently will quickly spread through the population to fixation; in this case, fixation of the trait would be same as development of an RIM  (a PRE-zygotic one)

4. Like any basic model, the MacArthur-Wilson equilibrium model of island biogeography can be tricky to apply in real-world situations.  This study looked diversity of anole lizards on Caribbean islands with a focus on how isolation ("distance" in M-W basic model) affects diversity by influencing the rate at which new species are added to the biota of an island: " Helmus, M. R., D. L. Mahler, and J. B. Losos. 2014. "Island biogeography of the Anthropocene". Nature 513:543–546."  Authors encountered three challenges in fitting the model by traditional methods (that is, by assessing the rate of colonization as a function of distance from a mainland/continental source pool in South America).
    * First, other islands can serve as a source of colonists; they may not all come from the single original source
    * Second, human-assisted dispersal: anoles have a habit of hitching rides on ships that pass between the mainland and islands and among islands
    * Third, new species can 'arrive' on an island by speciation - one original population splitting into two or more species within a single island
WITHOUT first reading the article (which you're welcome to do, subsequently, if you want), come up with ways YOU might deal with at least TWO of these problems by either 'controlling' for them or somehow incorporating them into the 'colonization function' of the model (perhaps by re-coding how you assess isolation).

A. You could 'control' if you could somehow identify colonists from other islands (genetic similarities?) and not 'count' them in figuring out the colonization 'curve'. But this would eliminate a lot of info (and these colonists might, in fact, inhibit colonization of same species from mainland), so more powerful approah might be to create an 'isolation' function that incorporates the distances and species pool sizes for all potential soruces of colonists, including other islands (some studies simplify this by using distance to 'nearest island larger than target island...').
B. The study built magnitude of human traffic/trade to island (using number of annual ship landings) -- 'economic isolation' --  into their isolation index/curve.  Some of you suggested something similar.....
C. Speciation is, in effect, a 'supplement' to immigration in bringing an island's species richness up to the 'equilibrium' for the island; this it is part of the 'colonization'  curve.  However, it is constrained by time -- it only matters over longer periods, so immigration swamps speciation unless isolation is very high -- and by island size -- islands have to be big enough to afford potential allopatry for within-island speciation.


AND ONE MORE:

5. Choose one of the two papers we read and discussed in class (Naka and Brumfield, concerning speciation patterns in Amazonia, and Lomolino, concerning island evolution).  Frame a further research idea based on that study -- a question unanswered, a further implication of findings (explicitly realized by authors or not), or anything else that is somehow rooted in your reading of the article.  State the question and explain how it arises from the reading. See if you can articulate a hypothesis or two to answer the question.  Offer a BRIEF (3-4 sentences maybe) suggestion of a basic research approach to address your question/test your hypothesis.  What kinds of data would you need to collect and what kinds of patterns would your thinking suggest should be seen in those data?