October 9, 2012
One of the guiding principles of the Extension Service is to be a source of research-based information. Research-based? Meaning that the information we provide is not supposed to be based on rumor or anecdotes, but is supported by science.
University researchers are obviously an important source of our research-based information. Nonetheless I believe that “research” and “science” come in many forms and on many scales. Many woodland owners like to experiment on their own forests to come up with management techniques that work for them.
I’ve found this especially to be the case when it comes to preventing deer and elk damage to western redcedar seedlings: from painting seedlings blue to scare tactics, I think I may have heard it all. Are these experiments “research”? Maybe – it depends on how they are set up and measured.
Recently a forest owner wrote to ask about one such browse deterrent method whereby a cedar and a spruce seedling are planted together (see photo). (The hypothesis: the animals are deterred by the sharp spruce needles; the spruce thereby protects the cedar; eventually, when the cedar has grown above browsing height, the spruce is carefully cut away.) The individual wanted some specific guidance on how to do this, and wanted to see a demonstration site. Although we know that people have tried this method, to our knowledge none of these plantings were carried out in a scientifically valid way. We can provide a hypothesis on how things might turn out, but to date we do not have research-based information to provide. Instead, we can only rely on anecdotal evidence.
I’m a strong advocate for woodland owners contributing to our collective knowledge of woodland management by trying out different techniques on their own properties. However, there are several important design factors to keep in mind if you want to call your experiments “research”:
- Have a control. Suppose you planted 100 cedar/spruce in the same hole, but did not plant any cedar without spruce. If the cedar are not browsed, it is not possible to know whether the spruce had any effect. It might just be that the deer were not hungry that year. In a controlled experiment, you leave a portion of the area untreated, or without the variable whose effect you are trying to test.
- Have a large enough sample size. Suppose you only plant five spruce/cedar combinations, and of them, two cedars are browsed and three are unbrowsed. It is hard to draw a conclusion from five seedlings. Was the treatment 60% effective, or did the two browsed trees happen to be unluckily planted right along a deer trail? If you had planted 50 spruce/cedar combos, and only two were browsed, then it is easier to say that the technique is effective.
- Replicate. What works on a north slope in Columbia County may not be effective on a south slope in the Willamette Valley; what works in a dry year may not work in a wet year. By repeating the entire experiment in more than one year or on more than one site, you can draw conclusions that have more power. This is probably the hardest one for small woodland owners to pull off individually. However, collectively there are a lot of experimenters out there. What if we could compile the results from everyone’s scientifically valid experiments? Then we might have some real research-based information (and some real value to all you frustrated cedar growers).