By Max Mitchell, Nashoba Regional High School Summer Intern, Summer 2018

While photographing plants on the farm, I turned over a squash leaf, seeking pests to photograph. To my pleasant surprise, a large frog hopped out. I got as close as possible and took many pictures before the frog’s strong legs launched it into the adjacent rows of crops.

The northern leopard frog, makes its home in swamps and marshes, the typical habitat for many amphibious creatures. In summer, however, the northern leopard frog has a tendency to leave the water and spend more time in grassy areas, such as fields and lawns. Flats Mentor Farm contains both wetlands and plains and provides an apt habitat for the northern leopard frog.

This frog is fun to look at, and its presence indicates that Flats Mentor Farm is a suitable location for this frog to live. It is a signal that there is an adequate population of crickets, flies, and other bugs for this frog to eat. In addition, it is a sign of balance in the local ecosystem. A careful equilibrium of both predators and prey is vital to any natural space.

There are many signs, other than frogs, that a farm plot is doing well. Plantings, an absence of weeds, and a lack of pest holes are all good tidings. But serendipity cannot be relied upon; we cannot count on every leaf being turned over to reveal a frog. Thus, a process had to be developed and revised to provide the World Farmers’ office with information on plots at the Flats Mentor Farm.

Flats Mentor Farm spans approximately seventy glorious acres of soil, divided into more than three hundred farming plots of varying sizes. Each plot needs to be assessed multiple times throughout the summer, via a process aptly named plot assessment. It seemed simple enough: walk through the farm, take a picture of each plot, and fill out a half-sheet slip providing information on the plot’s weediness, planting status, and overall condition. However, reviewing each of these plots and reporting that data back to the World Farmers offices proved a daunting task.

The first full-farm assessment took us interns over a week to complete, and another week to fully process and organize all the data collected. We performed this assessment using pen, paper, and clipboard, walking the rows of the farm and noting various aspects of each plot. This method had multiple flaws; the most blatant one was that all this data had to be converted from paper to digital, a long and arduous process. We decided that the second plot assessment would be done in a more efficient manner.

The second assessment took two and a half days to complete data collection, and, within the week, we completed data processing using multiple online programs, Google Forms, and DataScope. Google Forms allowed us to enter data and photographs in the field and directly upload them to a spreadsheet. The only flaw with Google Forms was that it did not have offline capabilities; this meant that uploading photos was quite slow at times, and when it wasn’t, the transfer of pictures used insane amounts of mobile data. DataScope allowed us to complete our forms in the same manner as Google Forms, but offline. The flaw with DataScope, however, was that its features were unreliable and limited when operating the free version. DataScope only allows sixty form submissions per account and could not reliably transfer information to a spreadsheet. These restrictions required a few workarounds, but all in all, using about 50% Google Forms and 50% DataScope allowed us to complete the process much more quickly.

Looking forward, there are various options for future plot assessments. One is to find a way to bring a hotspot to the farm, but this alternative would require a phone with both unlimited data and a reliable signal to be a viable option. A second is to pursue other form applications, such as Formotus, a program similar to DataScope. Unfortunately, programs like these cost money, which means either paying exorbitant prices for the application or creating multiple accounts and exploiting free trial systems. While this option can work, it is not the best solution.

A third alternative may be to create our own system which ports our individual form into an offline document that would sync with a spreadsheet, requiring an in-depth knowledge of computer programming and significant time to develop. It could be possible to use an open-source form framework such as TellForm to establish an offline form system, but, unfortunately, TellForm doesn’t support file uploads from the field (yet). Since pictures are a key component of plot assessment, this would render TellForm useless for our purposes. All in all, creating our own data collection software may engender greater strife than other solutions.

After a bit of further research and exploration, I found KoBoToolbox, a free, open source suite of data collection tools, developed for field data collection in challenging environments. KoBoToolbox provided us with an offline form service, including picture taking capabilities and an export to Excel spreadsheets. As of the writing of this post, this program would require the least human interaction in order to achieve the most streamlined plot assessment process yet.

For additional efficiency, any questions capable of being stated as multiple choice were set up in the form in that manner. This is faster than typing in various entries and makes standard all form inputs and syntax.

It’s important to remember that methodology alone does not determine speed and success of a plot assessment. Plot assessment does entail going to the actual farm; thus, weather conditions do affect our ability to assess the plots. For example, during a week when we were slated to perform the third assessment, a monsoon-like rain ensued, forcing a postponement of the assessment until the Flats were more hospitable. Alternatively, sweltering heat waves can slow the process as well, fatiguing the workers performing the assessment. Because of these uncontrollable weather factors, it is vitally important that a streamlined process be developed, so that we can take advantage of any time we have, even if it is only a few days of clear weather.

The impact of weather events also sheds light on another, possibly superior plot assessment technique: using drones. Albeit more expensive and requiring technological expertise, using live drone footage to perform the assessments remotely may work better. Footage could be collected by one person out in the field, snapping aerial pictures of each plot, then assessments could take place in the comfort of the office. Spending less time in the field would reduce the effects of drastic weather, and assessing in a cool, comfortable room would increase the speed of the actual assessment process.

And, as weather conditions change, the foundation of what a plot assessment needs to be changes, too. In late August, torrential downpours left the Flats with some of the most severe flooding in years. With tons of crops lost, and nearly innavigatable roads, the traditional plot assessment was not what was needed at the time. Instead, assessing flood damage was a more reasonable goal.

And so, we set off into the muck, photographing damage as we went and marking the most flooded areas on paper maps. Because in dire times, there isn’t always an opportunity to develop the most efficient method before completing a task.

While out, I saw more frogs than usual: to be expected, given the heavy flooding. But the farm was in a state of severe destruction. One indicator is never enough to draw a conclusion. An assessment must be adapted to needs, and I can guarantee between the writing of this post and the next assessment, changes will be made to the process. Which is why I cannot say that we have solved plot assessment. But as far as I can tell, KoBo Toolbox appears to be the best service for our purposes. Just as the northern leopard frog adapted spots to better survive, so, too, has our assessment methodology adapted new pathways to a more efficient process.