When endeavoring to develop an understanding of the nature of something it is generally useful to create a model of some type to aid in surfacing that understanding.
People often complain that developing models is too complicated, takes too long, is too esoteric, doesn't provide understanding, etc. You name it, they have an excuse. Here is model development distilled down to a single word... And? This is based on an old Sufi saying, "Because you understand one, you think you understand two, because one and one is two. But first you have to understand AND."
[Crop Eating Insects] lead to [Crop Damage] which is destroying my livelihood because I am a farmer. And? what do I do about it?
I [Apply Pesticides] because I know pesticides will kill the [Crop Eating Insects]. Here I have a Control Insects Balancing Loop (B1).
If I spray to Control Insects (B1) why is it that I have to spray more every year? And? what else is happening here?
[Crop Eating Insects] have [Insect Births] which adds to [Crop Eating Insects] and we have a vicious Population Growth Reinforcing Loop (R2).
If I [Apply Pesticides] and kill [Crop Eating Insects] then there should be fewer [Insect Births] though there are more insects each year. And? what else is happening here?
Insects have proven to be very adaptable creatures. When I [Apply Pesticides] it promotes [Pesticide Immunity] which results in more [Crop Eating Insects] because fewer are killed by pesticides. This leads to more [Crop Damage] encouraging me to [Apply Pesticides]. Here we have a vicious Pesticide Immunity Reinforcing Loop (R3) moving us in a direction we don't want to go.
And? what else might be happening here? Are there other things that [Apply Pesticides] influences?
When I [Apply Pesticides] it also kills [Controlling Insects] which devour [Crop Eating Insect]. This results in more [Crop Eating Insects] which leads to more [Crop Damage] encouraging me to [Apply Pesticides]. Here we have another vicious Predator Destruction Reinforcing Loop (R4) also taking us in a direction we really don't want to go.
If we promoted growth of the [Controlling Insects] population there would be fewer [Crop Eating Insects] and we could reduce [Apply Pesticides].
And? it would be good to consider what the [Controlling Insects] would find to eat after they consumed most all of the [Crop Eating Insects]. You can be rather sure they'll be looking for something to eat.
And? how long can you [apply pesticides] before you contaminate the water supply or ruin the ground so it won't support crop growth? Is this beginning to look at lot like the Bird Feeder model at the beginning?
The moral of the story is that the real implications of your actions may be counterintuitive and lead to the exact opposite of what you're trying to accomplish. Continuing to ask And? what else is influencing, or is being influenced by, your actions may provide meaningful insights. And? have you gotten the message?
When you run out of questions to ask then explain your model to someone else. You can be rather sure they'll have some questions you didn't think of.
If everything is connected to everything else then when do you stop? You stop when you are quite certain you've identified all the influences relevant to what you're trying to understand. Often relevant is difficult to determine. Then you have to go through trial and learning to figure it out.
Actually determining the loops and labeling them provides a more meaningful relationship map.