Hi, Everyone - I am hoping folks in the iNat community can advise me on how best to approach a small project to track the number and location of ant-lion larvae that dig pits in a patch of soil beside my driveway. I discovered larvae living there last summer. This year I’d like to record their numbers through the spring and summer by taking still images .
My thought is to photograph the patch in the same way each day with my iPhone camera, then overlay each image with a digital grid of 1-inch squares, aligning the grid using stones for registration, then mark the location of each pit on the grid. Is this a reasonable approach? Is a digital grid that can be used this way already available somewhere or would I have to make one? (I have Elements 11.)
I’m in Columbus, OH. On March 19, I noticed two pits, the first for the year. March 21, a couple more were present. The next day, they were obliterated by windblown fir needles. The photo shows the needle-covered area; the red outline marks the area where most pits occur, partially sheltered by a small burning bush. Does anyone have suggestions that will help me track the pits in some reasonably simple way?
Each leg is adjustable both in angle and height, so it can sit level on uneven terrain. One marked leg goes on a fixed datum (in my case a nail or spike hammered into the ground or between rocks), the long axis is aligned to a set orientation (or a second datum can be used to line up with another leg) and the height of the grid is adjusted to be a set height above the ground
I marked fixed intervals down the sides, and long the top, numbered them, and used rubber bands to create a grid. If you wanted to you could drill holes through and use thread, dental floss, fishing line, or cord instead.
For the survey a photo is taken from a fixed height above the center of the grid (later I colored the center square/crosshair for easier targeting).
This provides a fixed and easily replicable method for recording the type of data you’re looking for.
I came up with this system back in 2010 and the national park I was working for at the time is still using it.
Just did some digging in my old documents and found the plans for it. This was made on the basis of the largest I could easily pack, carry, and reassemble in the forest and on the cliffs I was working, so these dimensions are not necessarily what anyone else should use. This is mainly so you can see the parts list (note, In the future I might spend the extra time to drill holes and make the grid from dental floss as that’s extremely durable stuff).
Although I do love @earthknight’s solution, for monitoring your side garden area an easier and more casual approach might be appropriate. Since you aren’t sampling, you don’t necessarily need a grid. I have two options:
First Option: You could lay out a bit of plastic garden mesh and peg it to the ground. It would be inobtrusive and very cheap: a couple square metres for a few dollars. That would allow for referencing of positions and landmarking. I’d recommend anchoring it with a couple pegs. (Bonus effect: dissuades cats & dogs from making a mess…)
Second Option: Place 3 or more plastic garden pegs to roughly bound the area. Ideally the tops should be circular. Paint the top of each peg a different colour. Provided that at least 3 pegs are visible in a given photo, it would be possible to (1) find coordinates of a given antlion and (2) scale and transform the images to compare.
Thank you for your reply, @earthknight. Very cool plans and photo. This would be fun to assemble and use. Naive crafting question: Did you use a tap to cut threads for the set screws? Is gray PVC easy to find, e.g., at Lowe’s or Home Depot? I plan to give this a try, drilling holes and using fine wire for the grid.
Thank you, @murphyslab! Your second option is particularly interesting. I fear that I lack the confidence that I would accurately triangulate the locations of the pits from day to day. Thus, I’m drawn to a grid. However, I’ve noticed pits in a second area close to the first and will try your option 2 there in an effort to expand my observation skills! Thank you again for these additional approaches.
Well, if you do try the 2nd option, there is a plugin for ImageJ that can transform images based on landmark correspondences here: https://imagej.net/Landmark_Correspondences
I got everything at the local hardware store that was near the headquarters of the park I was working at. The color of the PVC doesn’t matter, it’s the diameters that do. If you find PVC of different dimensions that fits together well, then that can be used instead as well. If you want to make a metal or wood folding one the plans can easily be adapted to that.
The size as well can be customized to whatever suits your use best. I recommend that the interior dimensions be done in units of 10 cm though, and that the gridline spacing be no less than 2 cm per unit. I used 2.5 cm per unit as that’s what the park was already using for their groundcover surveys and I wanted to make sure that my system was back-compatible.
For the parts that required threading (where the eye-bolts go in to hold the legs in position) drill the hole just a little undersized (basically undersized by the thickness of the threads, or even less) and gently, but firmly screw the eye-bolts in. They’ll self-tap into the plastic. You might need to use a screwdriver run through the eye of the bolt as a handle for leverage. Screw in a bit, back out, screw in a bit, repeat.
For data recording any regular system works fine, but given the grid layout I used an excel table to record presence/absence. The nice thing about having a photo-associated data record is that you can go back and mine it for additional detail, and it makes it pretty easy to calculate areas, either by estimation, or by placing a scale-card on the ground where it’s in view of the camera and using software to do the exact calculation.
As murphyslab said, Image-J gives you some options for this sort of thing. Image-J is a pretty powerful open-source bit of image processing software meant more for technical use than for photo editing. It’s unforgiving and has a bit of a learning curve, so practice with it a bit and keep back-up copies of the original photos.
Fun project! I agree with the ImageJ suggestions. I usually add a ruler in the photo. I have used it to measure zebra mussel trails in sand. And also to measure skulls from photographs. Great software.
Here is how my grid came out. I followed the directions from @earthknight quite closely, except that I used 1-inch number 6 screws and fine florist wire to make the grid. I’ve used the grid so far with no legs, the tee junctions holding it off the ground enough to clear the pits. This works OK, but I’m going to try 30 cm legs and a 15 cm height so I can observe the pits better under the grid. I’ll mark the location of grid using two garden stakes that slip inside the PVC legs (or into the tees). Thanks again, @earthknight. :)
PS: I do plan to try triangulation and using ImageJ.
