This is just going to be a quick post about frame analysis. First: what is it? Frame analysis is an important step in determining how much the mass of a hive has changed over a period of time. During the hive analysis (where we weigh every part of the hive) we also take pictures of the front and back of each frame (I'll insert one of those pictures below).
 |
| This frame has no brood in it, but it has plenty of cells full of honey. |
When we weigh a frame, we cannot tell how much of that weight comes from adult bees, honey, or brood. We conduct the frame photo analysis to separate the brood mass from the mass of other beehive components. Determining the mass of a colony's brood lets us determine how healthy it is by measuring the amount of brood it is producing. Now, how is frame photo analysis done? There are several ways to do it, depending on how the brood is positioned on the frame.
 |
| All the capped cells in this photo contain brood. There is no honey on this frame. |
The first step is always to determine the size of the frame in pixels. We do this by selecting the entire frame with the selection tool, and then letting the program (called ImageJ) determine the area (in pixels) of the entire frame. Then, depending on the frame's configuration, we can do one of two things. For the frame above, I simply had to trace all the brood with a "freehand selection" tool and count all the empty cells in the middle with a "multi-point tool" (which counts points). Here's what I ended up with:
 |
| Notice the faint yellow lines tracing the brood. This is how we determine brood area. |
However, some frames are a lot harder to handle because the brood is scattered widely across the frame (what we call "shotgun brood") or simply because the shape is impossible to trace.
 |
| A prime example of a frame with "shotgun brood". |
 |
| This frame's brood was too hard to trace, so I had to count every cell with the counter tool. |
After finding the area of the brood (in pixels), we input it into Excel. The spreadsheet then calculates the percent of pixels traced from the entire frame. Since all frames are of a standard size, this allows us to estimate how much of a frame's mass is composed of brood. In cases where I am forced to count every cell with the "multi-point" tool, I also input the amount of cells counted into an Excel spreadsheet. However, this spreadsheet takes the average mass of a brood cell and multiplies it by how many cells were counted. Taking all of these steps into consideration, frame photo analysis is probably the most time-consuming part of my data analysis. While most of the data we collected simply require an Excel spreadsheet to be analyzed, this data requires hours (no lie) of painstaking tracing and counting. Even worse, if I accidentally deselect the "freehand selection" tool, I have to start all over again. I can only hope that I'll be finished in time for my presentation. Anyways, that's all for this post. I will probably post again soon about my presentation, or about data analysis if anything new happens.
