The numbers tell the story
After just 18 months, the Lake Kaniere Community trapping programme is beginning to show measurable results, both in the number of pests removed and in the growing body of data we are collecting.
It is starting to produce some useful headline numbers:
148 traps (many of which are in double-trap boxes, so roughly 70 boxes)
352 pests eliminated
Behind these two headline numbers sits a wealth of information recorded by our volunteer trappers using the Trap.NZ app. The app records:
The exact location of each trap
Details of the trap, including whether it is single or double, the make and model, service history, and maintenance issues
For each trap-checking event, the app also records:
Exact date and time of the check
Species caught, along with gender and age
Bait that attracted the catch
Condition of the bait
New bait added
Entering a new trap into the app takes only a couple of minutes and is a one-off task. Recording the results of a subsequent trap check takes about 30 seconds.
The amount of data collected is now becoming useful in helping us understand seasonal patterns and the effectiveness of different trap locations, trap types, and bait choices. Over time, we can use that knowledge to improve trap placement, adjust bait selection seasonally, and optimise how often traps are checked.
There is nothing pioneering about our collection and use of this data. We are still newcomers to the craft and science of successful trapping at scale and to the use of tools like the Trap.NZ app. From the start, we have been keen to learn from other more experienced trapping groups about subjects such as optimal trap placement, trap types, bait selection, and seasonal variation. However, every locality is different, and I think we will learn a great deal from analysing our own growing pool of data.
To that end, one of our group recently fed our trapping data into an AI tool with the goal of identifying traps that might be moved to improve results. The AI tool came back, in about a nanosecond, with a three-page analysis that included observations about the year-on-year doubling of our catch rate and identified two traps that might be worth relocating.
Why does any of this matter?
The first answer is that it gives visibility to our results. Based on estimates of the impact predators have on native bird populations, those 352 eliminated pests may have prevented the deaths of thousands of birds, potentially as many as 10,000.
The second answer is that, if we want to sustain and improve these results, we need to make every volunteer hour and every donated dollar as effective as possible.
The value of AI is not that it magically solves our trapping challenges. The value is that it can sift through thousands of data points in seconds and highlight patterns and trends that are worth investigating.
As our dataset grows, so does the opportunity to learn from it. If you are interested in data analysis, conservation, or simply helping us make better decisions, I would love to hear from you. Please give me a call on 021 332 387.
In the meantime, here are a few images from our Trap.NZ data:
Comments