Dear evolutionists,

here are Nathan Lovejoy's answers to your questions. They are divided up into seven large groups, within which similar questions were collected. Nathan can be reached by e-mail or you can look for him at the Museum of Vertebrate Zoology (3101 Valley Life Sciences Building) if you would like to learn more about his research. We hope you enjoy this first round of the remote experts exercise!

I. Replication of the pattern seen in stingrays in other taxa, and other examples of marine incursion.

Q1. Alan Shabel, Jan Nakamura, Patrick Smith, Robert Young.

What is the next step in the effort to falsify the marine incursion hypothesis? Would one of the marine-derived, freshwater taxa in the Amazon basin (eg, dolphins or crabs or anchovies) be more amenable to analysis than the others, for theoretical, biomolecular, ethological or practical reasons? Is there another biogeographic region (eg, Africa or Asia) where an incursion hypothesis could be tested against an invasion hypothesis?

A1. The next step is to include more groups. Yes, some taxa are definitely more amenable to analysis than others, for a variety of reasons. For example, fishes of the family Sciaenidae (called drum or croakers because of their sound-producing abilities) might be good candidates for analysis, except that the marine drum along the coasts of South America are extremely diverse ( fewer than 20 genera in the Pacific), complicating attempts to get sufficient specimens for analysis. Other groups (flatfishes) tend to be rather rare in South American rivers. The dolphins are difficult because getting tissue samples is very complicated (they are protected by endangered species laws). I will probably work on anchovies next, because they are easy to catch in both freshwater and marine habitats, and their diversity is manageable.

Q2. Allison Smith, Paul Rojas, Amy Maxmen, Scott Lim, Matt Marcus, Cecilia de Robertis, Danielle Lee.

Has any research been done on the other marine derived groups (e.g. flatfishes, pufferfishes)? It seems expected that they should follow the same pattern, but there's no reference to analysis of said groups.

A2. Explicit phylogenetic hypotheses for most of these other groups are not yet available. I just finished work on needlefishes (for my PhD), and believe that they also fit the Miocene incursion model. I am still analyzing this data.

Q3. Jen Steele, Alex Tolmasoff and Rich. Shefferson

Have other organisms in this system been examined to determine if this speciation event is shared among other taxa and is the likely result of this marine incursion? see A2

Q4. Andrew Lowenthal, Dafna Hopenstand and Monika Lipinski.

Has miocene incursion been born out for other genera in Amazon besides freshwater stingrays?

see A2. There are lots more groups to look at, and I would like to do this in the future.

Have other similar incursions lead to speciation in other areas of the world?

This is a really interesting question. I noticed a sign the other day at the Cal Academy of Sciences that noted that many freshwater North American molluscs are the product of ancient marine incursions. I've been trying to find references concerning this topic. This is the only other example I know of, but I need to look more thoroughly at the invertebrate literature.

Q5. Sabrina Minter, Laurel Suter and Marion Wittmann

A. Have any studies since been done on the other taxa you mentioned (dolphins, fish, etc) that would support or refute your hypothesis about the single unique origins of endemic freshwater stingrays?

see A2

B. Could you please provide another example of a marine incursion and compare it to the one described in your paper?

There were massive marine incusions in North America during the Cretaceous (I believe). The Sundance Sea essentially split the continent in half. This was a much bigger incursion than the on mentioned in my paper.

II. Specific aspects of marine incursion hypothesis, progressive desalinization

Q1. Jan Nakamura, Patrick Smith, Robert Young, Alan Shabel

How certain are scientist that there was a mass movement of marine waters into the upper Amazon region during the Early Miocene epoch? What evidence do they have?

A1. Fairly certain. The evidence consists of "marine" sedimentary deposits-- i.e., the rocks of a certain age look like they were deposited and formed under marine conditions. Also, scientists have found fossils of marine organisms (such as mangrove pollen and marine fishes) in the upper Amazon. More recently, some authors have looked at strontium isotopic ratios in fossil mollusc shells and again found evidence for marine influences.

Q2. Mike German, Jodie Horn, Jose Dinney, Gautam Guruprasada, Anne Marie France, Maggie Kuo.

The paper speaks of "progressively desalinized habitats" as being a key factor in the evolution of freshwater stingrays. Are ancestral forms of potamotrygonids found in relatively brackish water in comparison to the habitats of the more derived taxa? In other words, is there any evidence of an evolutionary cline with respect to salinity, or are all presently known species in the family equally adapted to the same freshwater conditions?

Yes, there is evidence of such a cline, but not within the freshwater rays. The Potamotrygonidae are all exclusively freshwater taxa, and are only found in the rivers of South America. To my knowledge comparisons of salinity tolerance within this group have not been done (but that might be interesting). However, the closest relatives of this family (Himantura pacifica and H. schmardae) are euryhaline (i.e., they tolerate a broad range of salinities), while more distant relatives tend to be strictly marine. So we see a progression from strictly marine to euryhaline to freshwater-tolerant.

Q3. Mandy Heddle, Cherie Le Doux, Jessica Riquetti, Emily Marquez, Sarah Rieboldt, Wyatt Korff.

In the article you say that "these conditions would have been ideal for the isolation of marine fishes in progressively desalinized habitats". This implies to me that earlier speciation events (e.g. P.aiereba) resulted in species that were more tolerant to fresh water than their ancestor but less tolerant than their descendant species. Do you find any evidence for this? Is there a sharp divide between fresh water habitats and marine ones in the amazon basin as it now exists,which would explain a lack of intermediate forms? I think it would be an interesting project to look at the abilities of these different species to tolerate different degrees of salinity.

See question above. I agree, it would be interesting to compare relative degrees of salinity tolerance in different taxa. Remember though, the freshwater rays have been living in rivers for more than 10 million years, thus we might expect all of them to be equally well adapted to freshwater conditions.

Q4. Blake Suttle and Sethi Thomas.

Is there any evidence (fossil, molecular, or otherwise) to suggest that during the latter part of the marine incursion (~15-16mya), individuals in desalinized regions that had undergone some differentiation from their marine ancestors may have migrated back into the marine environment?

Not to my knowledge. We would have to figure out how such scenario would look in terms of a phylogenetic tree (or the other sources of data you mention). It might be very difficult to determine whether such an event happened.

III. The molecular clock

Q1. Wyatt Korf, Mandy Heddle, Cherie Le Doux, Jessica Riquetti, Emily Marquez, Sarah Rieboldt.

Dr. Lovejoy: while reading your Nature paper Marine incursion into South America a number of questions arose that I hope you might be able to clarify/expand upon.
You assert that the divergence time between potamotrygonids and their closest marine relatives occurred during the Miocene based on cytochrome b molecular clock techniques. My question is this: How accurate are molecular clock techniques? Are rates of cytochrome b evolution constant or predictable? I know that cytochrome b sequencing is only one of 3(?) techniques for molecular phylogenetic reconstructions (is it cytochrome a and ??), how well do you think this timescale would hold if one were to utilize these other sequences?

Q2. Rafael Oliveira, Scott Nichols, Chet Moritz e Malia Rivera

The main conclusion of the paper relies heavily on the assumption of a molecular clock, and has even been calibrated on a relatively recent known geographic event which separated one species, P. yepezi (which lies w/in the potamotrygonids) some 8 million years ago. But is this calibration enough to validate rate constancy? Could they perhaps performed some sort of test to see if cyt b actually evolves in a clock like manner before drawing conclusions about the date of origin of the potamotrygonids?
I guess we're just wondering if there is reason to believe that there is a constant rate among lineages, to put it in simpler terms.

Q3. Allison Smith, Paul Rojas, Amy Maxmen, Scott Lim, Matt Marcus, Cecilia de Robertis, Danielle Lee.

How accurate is the rate calculated for cytochrome b evolution, given that it is only based on one historical event?

Q4. Dan Jung, Diana Juan, Jen Highland and Amy Jess.

B. You refute the Andes uplift hypothesis by stating that potamotrygonids originated more recently than the uplift. You base your rate on cytochrome b analysis, however, do you feel that the analysis is a good reflection of evolutionary rates in potamotrygonids? Are evolutionary rates conserved? What were your samples for the cytochrome b analysis? Is mitochondrial DNA a good reflection of overall DNA rates, since it only reflects the maternal lineage?

Q5. Jen Steele, Alex Tolmasoff and Rich. Shefferson

A. In Lovejoy et al. (1998), you mention that you "calibrated a rate for cytochrome b evolution in potamotrygonids based on the amount o sequence divergence between species separated by a known geological event..." However, your use of this approach to then infer the evolutionary history of the stingrays in your study implies that a.the degree of divergence in this gene is equal across all stingrays in the study, and b. the degree of divergence is not underestimated, which I suspect may occur given that other genetic phenomena may be masked in a diploid organism. For example, if you discover a certain rate of divergence in your calibration species, but the genetic changes that that particular organism underwent are different than in other stingrays (as would most certainly be the case), then even if they evolve "at the same rate," your calculations may be underestimated and suggest the wrong evolutionary history. Although your results certainly fit the historical/geological framework you present very nicely, I am curious about the uncertainties I just mentioned and how they affect your overall conclusions. A very limited amount of scientific inference can be made with historical correlation, after all.

Molecular clock answer:

IV. Methodology: Inferring the tree, choice of outgroups, molecules versus morphology

Q1. Jan Nakamura, Patrick Smith, Robert Young, Alan Shabel

How were the outgroups in the study selected, and exactly how many were used and what were they?

The complete data set included all stingray genera that had been previously hypothesized to be closely related to the freshwater rays. As more distant relatives, we also included representatives from the large pelagic stingrays (manta and eagle rays), a guitarfish, and even a shark.

Q2. Wyatt Korf, Mandy Heddle, Cherie Le Doux, Jessica Riquetti, Emily Marquez, Sarah Rieboldt.

It seems that since the application of molecular techniques to the study of evolution, there has been a de-emphasis on using morphological characters to reconstruct phylogenies. Doesn't selection act upon the phenotypic traits of organisms...so having said that, shouldn't one look for differences at the phenotypic level instead of the the genotypic level (as many mutations may never be expressed). What should one do when morphological and molecular phylogenies clash?

I agree, to some extent the advent of molecular datasets have pushed morphological work to the side of the stage. Ironically, molecular systematists use your very argument against morphological data: since selection acts on the phenotype, we will be misled by morphological characters. I think the best way to reconstruct phylogenies is to use a combination of morphological and molecular characters in one big "total evidence" analysis. I was pleased with the molecular results in this study because they agreed with a morphological phylogeny I previously published.

Q3. Jonathan Woo, Elizabeth Agrilla, Danielle Lee, James Mull, Tonya Van Leuman, Tony Donoghue.

Why was the sequence for cytochrome b chosen, as opposed to other mitochondrial DNA sequences? If another sequence was chosen, would the same conclusions be drawn, or would the conclusions be different, resulting in a different phylogeny being produced?

V. Taxonomic implications

Q1. Mike German, Jodie Horn, Jose Dinney, Gautam Guruprasada, Anne Marie France, Maggie Kuo.

On the cladogram shown in figure 1, why is it that Pleisotrygon iwamae and Potamotrygon orbigyni are depicted as sister taxa? This asserts that they are more closely related to each other that either is to any other taxon on the tree. However, other members of the tree are also in the genus Potamotrygon, which suggests that P. iwamae should be more closely related to them than to its sister taxon. Is this simply a problem of naming, or is it acceptable for members of separate genera to be more closely related to each other than to members of the same genus?

Yes, the position of Plesiotrygon was a surprise, and you're right, this is not acceptable if we want taxonomy to reflect phylogeny. If additional (morphological) work confirms the relationship between P. orbignyi and Plesiotrygon, I would propose changing Plesiotrygon's name to Potamotrygon.

VI. Biogeographical and historical interpretations

Q1.Chet Moritz, Scott Nichols, Rafael Oliveira, Malia Rivera.

It appears (from looking at the figure) that the closest marine species (Himantura) stem from an ancestral taxon which probably diverged in the Pliocene as a result of the rising of the Isthmus of Panama. Lovejoy states .."the marine relative of freshwater stingrays does not occur along the Pacific coast of South America." Considering the Miocene occurrence of these marine incursions, the common ancestor of Himantura and the Potamotrygonids would have been distributed on both sides of what is now the isthmus of panama during these marine incursions. This doesn't invalidate the hypothesis of isolation by marine incursion, but how can Pacific vs Carribean origin be ruled out if trans-isthmanian gene flow was not restricted until the Pliocene?

Q3. Blake Suttle and Sethi Thomas.

Offering modern biogeographical evidence to support the Early Miocene marine incursion hypothesis, the article states, "Moreover, one of the two species that are marine relatives of the freshwater rays is distributed along the northern coast of South America" and later that "the marine relative of freshwater stingrays does not occur along the Pacific coast of South America". Is it conceivable that 23 million years ago, the common ancestor of the marine and freshwater stingrays was distributed across a greater latitude than it's two modern marine descendants, and that changing environmental conditions and physiological constraints coinciding with evolution of the marine lineage have caused a reduction in the range to that of the present? Judging by the relatively recent divergence of the two species of Himantura we assume a similarly recent separation of the Pacific and Atlantic Oceans around southern Central America. The range of the common ancestor must have encompassed parts of both modern oceans, which causes us to wonder if the modern range of one out of two descendants can serve as useful biogeographical evidence of a hypothetical ancestor from over 20 million years ago (the molecular phylogenetic data we find thoroughly convincing, by the way).

Answer to Q1, Q3 above. You make some great points, and clearly understand the biogeographic issues involved. Unfortunately, without fossils, it is almost impossible to reconstruct the past range of the marine "Himantura" clade when it gave rise to the freshwater rays. Our best (and most parsimonious) estimate, therefore, is that the current range approximates the past range. In fact, distributions of fossil marine molluscs from the Miocene show ranges that are coicident with the current distribution of Himantura, lending some support to our hypothesis about its past range. However, the main reason to think that freshwater rays came from a proto-Caribbean source, is that this was the source for the marine incursions that inundated the upper Amazon.

Q2. Alan Shabel, Jan Nakamura, Patrick Smith, Robert Young.

A. First, what evidence is there for Himantura schmardae being more closely related to the Potamotrygonidae clade than H pacifica? (In Figure 1, on the cladogram, both marine species appear to be equidistant from the freshwater species). Second, even if the marine incursion hypothesis is upheld, why are we to expect there to have been only one major radiation of stingrays? How can we be sure that evolution in Amazonia adheres to the principle of parsimony?

Q4. Wendy Park, Paul Moody and Paul Rojas.

How do you know that H. pacifica & H. schmandae share a common ancestor with the freshwater fish?

Because of molecular and morphological characters that these two groups share and other taxa do not.

VII. General questions

Q2. Jen Steele, Alex Tolmasoff and Rich. Shefferson

Can these salt-water vs freshwater fish interbreed?

Nope, they are different species and do not interbreed.