ESA - more mobile blogging

This week is ESA (the Ecological Society of America) meeting in Austin, TX. Where 3,000 ecologists get together to talk about anything you can think of that goes under the umbrella of "ecology" (which is a lot!). I am giving a talk on my research Friday morning (the last session, where an unfortunate number of friends are giving overlapping talks). I am going to try some mobile blogging again, adding to this post throughout the day.

Me and my lab-mate Dan got in yesterday (Monday) afternoon. Pretty successful so far. Saw some mediocre talks go before my friend's very good talk, saw a couple people I've met at EEID meetings, and saw a former Madison classmate (we did a project on leaf-cutter ants for our tropical botany class) who will be defending his PhD from Princeton next week.

First session on trait-based approaches to disease ecology was pretty interesting. The best thing to come out of it was a discuss with a few of the other women working with tick-borne disease. Have a meeting with one of them who is worked on things really similar to me, immune function and how that relates to ticks and disease, since she will miss my talk on Friday. Its so nice to have the people who have the possibility of being academic rivals are actually really nice!

Doing the right thing

http://xkcd.com/552/

I’m not a very good liar, never have been (much to the chagrin of my teenage self). I have a big problem when people outwardly lie or try to deceive people (don’t even get me started on our current political situation). I sometime wonder if this is why I like science to think I make a good scientist, because ultimately we are looking for what’s true about the world. We look at data and see what it tells us about our research questions. A true scientist can have his preconceived views come crashing in around him at any moment if the evidence is there to overturn what he previously thought. There could be books written on “the best science at the time” and the significant repercussions of acting on what we thought was true then, and that we know isn’t true now. This can be really scary to people, knowing that you probably don’t have all the answers or that what you think one day can be completely challenged on the next. While most of the time this makes me feel good about my work, that it is honest and forthright, but there are days when I wish I wasn’t so truthful.

Sometimes it is really annoying to be a person who has to do thing, in this case concerning the data I am working on writing up as a manuscript to be submitted to a journal. Because I am dealing with field data, which is inherently messy, and a diverse community of multiple host and tick species, there are a lot of ways to analyze the data. There are many opinions on what the best way is. Which stats are best suited for the data, how to divide up and compare groups, how to/if transforming some of the data will improve the analysis. My head has been spinning the last couple weeks with all the data that I have available, how to describe it in a way that makes sense, and how to emphasize the patterns we think are the most important.

Unfortunately, the best solution on how to approach some of these issues would diminish the impact of our results a little bit. For instance, the data I have from the immune response assays I did with serum from the rodents needed to be transformed to deal with a handful of individuals who has somewhat strange patterns in the results of their test (I am going to withhold some of the details because this is unpublished work). There were two camps of how to transform the data, and I tried both out. Analyzing one kind of transformed data led to some near significant differences between groups of hosts. The other way showed far less statistically significant differences, and after digesting the pros and cons of each method seemed to be the more correct way to do things. This is pretty annoying because the differences I was seeing in these results we’re super significant and I thought this may diminish the impact of the story I was trying to tell. But in the long run I will feel better about presenting the results I think are the most accurate than cherry-picking the results that I “like” best. Still annoying, though.

This constant tug of war with wanting to find the answers and have mind-blowing results, and scientific integrity is a tough one. You’re not going to always get what you expect and you have to remember that negative or unexpected results are still results and they always teach you something. Whether it’s telling you that the pattern you expected isn’t actually what is happening, shows you that you need to improve your methods, or helps you design your next experiment. Because all scientists have probably experienced this in some way is probably why there is always such an uproar when its found out that someone faked their results. Top researchers have lost their jobs, or at least their credibility for doing dodgy science (for example: http://www.scientificamerican.com/article.cfm?id=fudge-factor). While I am definitely not at that extreme a point with my work, I can understand how tempting it is to present the results you think are true, even if they are not actually true. But that isn’t what being a scientist is about, and I think I’m ok with that.

P.S. The comic found here was shared by some friends on Facebook. It’s funny and true, it can makes the average grad student happy and sad at the same time (hehe)

Some reflections on looking outside your scientific box

Overall it’s been an interesting week participating in the ABS conference here in Bloomington. An opportunity to see talks and posters on subjects I never would otherwise, like bird song, aggression and face coloration in paper wasps, and competition between seasonally sympatric wren species. I went to a talk just because it was on Beldings ground squirrels (they are very cute!).  It was also interesting to go to the handful of talks on parasites and immune function and see how they are presented at a behavior, as opposed to a disease or epidemiology, meeting. These were interesting, but often with less rigor or complexity than I have come to expect from the studies presented at the disease meetings.

I keep saying this is a very “hard-core behavior” meeting, and I think that still is the best way to describe it. This is where the folk who do focal observations, activity pattern analysis, song pattern analysis, play behavior, and sex role-reversal research present their work. These are things I have not thought about for a long time. There is very little applied or mechanistic work, so less kind of comprehensive or interdisciplinary work looking at the evolutionary mechanisms and patterns in behavior. There are of course talks on those subjects, but they don’t seem to be in the majority. Being a person to dips her scientific quill into the ecology, eco-immunology, and behavior wells, which seems to be the case of a lot of disease ecologists, it’s a bit strange to step back to see work that only work in one of these fields. This research is very interesting and worth-while but I don’t think it would be as satisfying for me. Maybe it is more fun to get to see talks on cool things animals do instead of studying it all the time, I need ecology and interconnectedness for my real work.

So one of the themes of this week has been thinking about how other biologists do their science. I went to a genomics lab meeting on Monday, one of the students who rotated in my lab is doing some work on genome sequences from lonestar ticks and presented her work. She has been looking at genetic sequences from this tick species, comparing it to other available sequenced genomes and trying to find matches for possible genes. When Mandy was showing some candidate gene matches and talked about how they could be related to the biology of the tick, mainly blood-feeding, one thing the PI (principle investigator) said during the meeting was, “Great, we can really emphasize the biology when we present these findings.” Because so much of genomic research is descriptive and mechanistic at a very basic, functional level (Mel can scold me if I am making too much of a generalization here), I guess its not always the case that the researcher brings their story back to the biology or natural history of the study species. You rarely have to remind someone to “emphasize the biology” in a community ecology study, for instance.

During the poster session last night, another IU grad student friend had similar sentiments about the meeting not being very mechanism heavy, and she said you didn’t hear the term “evolution” in the presentations at this meeting. This isn’t an issue of people not “believing” in evolution, more that the evolutionary biology (and sometimes the ecology) of the behaviors they study isn’t always the first priority for their research. This is not the case at more integrative behavior meetings that she is used to going to. It also made us think about the personality of the IU biology department, which is quite interdisciplinary, and in the animal behavior section in particular very strong in terms of evolutionary mechanisms. It was also clear that the people studying disease and parasites in animals do not come to these meetings, or the growing interest in parasites that seems to be everywhere I look hadn’t made it to this community. I had maybe 5 people come to my poster last night, and most of those people were friends or professors who already know about what I do. This is a change from the EEID meeting in Santa Barbara where I was talking to people about my poster for 3 hours. I also got very few questions on my methods or details of the work that I was hammered with at EEID. The behavioral ecology poster next to mine, which was very interesting, had a crowd around it the whole session.

I’m glad everyone is interested in different things. If we all liked the same kind of research it would be pretty boring. We need experts and people that integrate different things, this is what helps us get real understanding. The big thing is that I think scientists from different fields need to be ok talking to each other, thinking about how they can work together and what connections there are between our work. This is where science can be really exciting and innovative.  

Thursday - ABS meeting

Plenary speaker this morning, theme was immune function and animal behavior. Talk was a bit too general for me, which was shared by a fellow eco-immunology student. I guess thats necessary for a talk of this kind, but was hoping to get more out of it than I did. Some interesting examples of direct trade-offs between stress and response to bacterial infection in crickets, though.

High competition in females comes with the term "role reversed species". Funny to hear that in relation to animals. Of course, any female can tell you that females are very competitive, just in a different way.

Female topi deer have intense competition for males in leks. Highly preferred males "become physically exhausted" and "sperm depleted" because all the females have synchronized estrus, lasting for about a day. Quite intense!

Talk from Caroline Drea on masculinized female spotted hyenas. I did a report on this phenomenon for my physiological ecology class. Google this, its pretty crazy.

Apparent this also happens in lemurs. Lots of pictures of lemurs "junk" (the actual term used in the talk).

Presenting at the second poster session of the meeting. It was very crowded, but I ended up getting a spot on the outside row, very nice. Only about 5 people came to my poster, 4 of them were my friends or professors. Very different from the response I have gotten at the last meetings, talked for 3 hours about the same poster at EEID. Oh well, I enjoyed talking to the few people that came by.
Pretty tired now, was at the meeting for 12 hours today. One more talk tomorrow and I'm done.

Some mobile blogging from the ABS meeting

Sign language interpreter at a talk on vocalizations in chickadees. We were wondering how many of the scientific terms actually have their own signs. I bet there are more than we think, but I bet there are a lot of things that challenge them.

Great hypothesis testing from tropical birds. It has taken 10 years to collect enough data to test a series of hypotheses on why a tropical species of wren sing duets. Have to get used to seeing small sample sizes and non-parametric stats for these behavior talks!

Dustin rockin' out the junco short-range songs. These are a group of bird songs that are not studied very often and their role is not very well understood. They sound really crazy, like robot noises. His talk ended with a great animation to a recording made of a male "dive-bombing" a female.

So frustrating when the AV set up detract from the quality of a talk. Can barely hear a word from this speaker, microphone pointing the wrong direction. Slides full of too small images. I think this must actually be an advertisement for a behavioral data collection program.

Second talk about invasive species, didn't really think I would see many of those here.

Saw a poster that was on aggressive behaviors in kittens. This means she got to watch kittens while doing science! Maybe she needs an assistant...

Also saw a poster from the group at Miami University-Ohio that probably has data I helped collect from Bayles Rd when I was a field assistant for them in 2007 :)

What comes now?

Field work is over! Now what? There is something after the field season? Indeed there is, and this part of the research process usually lasts much longer than collecting all our samples. Needless to say, with weather being in the mid-high 90s for the last week and this kind of weather on the horizon, I don’t mind the change of scenery that includes air conditioning! I got a couple requests to talk about this part of the process, what do we do after we have the data and samples from the field season?, so I will do my best to discuss the main part of my work in the lab. To keep things lively and interesting I may also integrate discussions of papers or science news that relate to my research or are otherwise of interest. Also, keep the requests coming for things you would like to see me write about.

I think one of the misconceptions that undergrads have about what its like being an ecologist is that we spend all our time running around in the woods. Unfortunately, that is not the case. When I helped teach an undergrad ecology course, in response to a beginning of the semester survey question on experience with Excel, we had a response along the lines of, he wanted to be an ecologist so learning programs like Excel wasn’t really important. The professor and I thought this was quite funny. When I told this to one of my other grad student friends she said, “If all you did was run around in the woods you would just be a vagabond. You need Excel to go from vagabond to ecologist!” I think that’s a great observation (that you Ms. Mel Toups) because I guess that is a difference between being someone who just likes hiking around outside and someone who is a scientist, and that hiking around in the woods has a larger purpose, involving collecting data, manipulating things in the field, and bringing all that information back into the lab for analysis. So, in my goal to not completely be a vagabond, Excel and other computer related things have become a larger part of my daily life.

The first thing I do when I get a new batch of data is do some exploratory graphing and statistics. This is basically to get an idea of what the data look like, which shows me what the general patterns are and helps me make informed decisions on what the appropriate statistics tests are for this data set.

For example, one of the things I am interested in is what the distribution of ticks on hosts is. In other words, are ticks distributed evenly across all the rodents I caught or are some more parasitized than others? In the 2009 field season, this is what the tick distribution looked like:

The x-axis (horizontal) is the number of ticks per host, the y-axis (vertical) is the number of hosts in each category. This graph shows that there were the most hosts in the zero tick category, most hosts had no ticks, and that few hosts were found with many ticks on them (there was only one host with 33 ticks).

In 2011, this is what the distribution looked like:

There are the same axes on this graph. But you can see that the distribution looks a little different. There appears to be a peak at 1 or 2 ticks, most hosts had 1 or 2 ticks, but there is still a long tail to the distribution, meaning there were still few ticks that had a lot of ticks. Because most classic statistical tests assume your data are “normal” and look like this:

This means that when I do my analysis I have to make sure to tell the program that my data look different, in this case have a “negative binomial” distribution (http://en.wikipedia.org/wiki/Negative_binomial_distribution). We have a great stats consultant on campus that I and others in my lab have talked to before and I will be meeting with her soon to make sure I understand what the appropriate tests are for this data set. I’ll use these statistical tests to see if there are any kinds of hosts (males or females, reproductive or non-reproductive hosts, hosts in certain habitats) that usually have more or less ticks than others. I’ll then extend this to investigating which hosts have tick-borne bacterial infections, but that’s down the road a bit.

So that’s one of the things that I’ve been up to after being back in the lab. I’ll talk about the lab work and tests I’ll be doing in the next post.

Last Day (insert Logan’s Run reference here)

The field season is done! We had a really successful summer, but I am very glad to be finished. My field assistant Alisha and I were trapping for about 2 months, and collected a ton of data.

Let’s do some numbers:

7 weeks, 21 mornings (up at 5:30am, to the site by 7:00am back again to bait at 4:00pm)

2 sites, 4 sampling periods (groups of 3 trap-nights) at each site

48 unique individuals caught, 42 mice, 6 voles

96 total captures

76 blood samples

637 ticks collected from animals

2,680 ticks collected by dragging

1,518 miles put on our car

Unknown numbers of gnat, chigger, mosquito and tick bites

No cases of tick-borne disease

1 great field assistant, who said she had fun doing work that is probably not common for pre-med undergrads (being a farm girl was good preparation, I think!)

1 very patient and understanding husband, who did get two tick bites from what I brought into the house

What comes next is the ID-ing and processing of all the field samples. If all goes well, this will be all the data I need to complete my PhD. There may be a couple weeks of sampling in the fall depending on what the data look like. I will therefore from now on write about what we do after the field work is over, what happens in the lab and in on the computer to make all this information make sense.

This is kind of an exciting and scary time. All kinds of data with the potential to be great, but also with the potential to be too messy, complicated, or not enough to show clear patterns. I am staying optimistic, however. There are a lot of exciting things planned for all these field samples, things can only get more interesting from here, right?