Happy Wednesday, readers! It’s hard to believe that this will be our last Wednesday here at the Garden, but it is—things are winding down. Well, sort of winding down. We still have a lot of work to do! For me and Jackie, we’ve finally gotten to start the data analysis (also known as the fun part).

But first, we had to do all of the x-raying! We mentioned it briefly in some other posts, but here’s what the results actually ended up looking like:

We then had to look through the images and count which achenes were full, empty, or partially-full (shown in red, blue, and green, respectively). We use these x-rays to get a sense of seed-set, or how well the Echinacea heads were fertilized. All of the previous steps have been looking at achenes, which are actually the fruit of the flower. All the heads produce achenes, but only some of those achenes have seeds that will grow into more Echinacea. In this way, the x-raying can be considered the most important part; it’s measuring reproduction most directly. Also interesting to note: the x-raying protocol is very careful to minimize the x-ray exposure of achenes being studied. That way, any seeds produced are more likely to be viable, and still grow later!

With that out of the way, we’re on to the fun of analysis, which doesn’t actually look that fun (unless you like computers):

Whoops, I accidentally got my finger in the shot.

We’ll be spending a lot of time manipulating our data in R (as shown) and feeding it into models. The idea is to test whether achene location on the head, plant isolation, and flowering timing relate to seed-set. Tune in on Friday for results!

With our second of three weeks coming to a close, the externs are working hard to get their projects finished. While Belle is still busy with the pollinator database, Audrey and Jackie have been desperately trying to finish processing all the randomly sampled heads from the remnant prairie populations. This processing isn’t quick–that’s why it’s taken up the majority of our externship. As mentioned previously, processing has five steps: cleaning or dissecting the head to get all the achenes out, scanning the achenes, counting the achenes, randomizing a sample of the achenes to x-ray, and finally x-raying the achenes for the presence of seeds. Now that we’re finally done with all of the cleaning, we’ve been focusing on scanning, counting, and randomizing so that we can get x-raying next week. Here’s a behind-the-scenes look at how we’ve been doing these steps:

The scanning is just how it sounds: we very carefully pour the achenes out onto a standard office scanner and get an image of them. This image is what we use for counting. It may seem superfluous to scan the achenes and count from the image when we could count the achenes themselves, but when heads often have more than 200 achenes, it’s tough to accurately count by hand. Computers are much better of keeping track of what number they’re on, so we let them do the work. Once the image is scanned, we just have to click on each achene we see in the image and it’s marked with a dot. The computer keeps track of the number of dots. That way, you know you haven’t missed any achenes and that your number is accurate. Here’s Jackie, in action counting:

After scanning, I’ve been taking the achenes and randomizing them. In other words, I’ve been taking all the achenes from the middle of the head and taken a random sample of 1/6 of all the middle achenes. This is done pretty simply: you take the achenes and pour them out onto a  circle divided into wedges and labeled with letters. Then, from a list random letters, you determine which wedges you’re taking achenes from. For the picture below, for example, the achenes chosen for x-raying were from wedges G and H:

If you think these steps sound a little tedious, you’re right. But, we’re hoping all this processing leads to some really interesting data to analyze next week!