June 26, 2009

Covet thy neighbor’s light: Influence of neighboring species on tree growth

Article reviewed: Tree growth at stand and individual scales in two dual-species mixture experiments in southern Washington State, USA

By H.E. Erickson, C.A. Harrington, and D.D. Marshall. Published in the journal Canadian Journal of Forest Research, Volume 39 pp. 1119-1132

The plot line: The authors of this study planted seedlings in an open environment. Some seedlings were planted next to seedlings of the same species, and some were planted next to seedlings of a different species. Species combinations included Douglas-fir with western white pine, and Douglas-fir with western hemlock. After 24 years of growth, they found that when Douglas-fir was surrounded by western hemlock, it tended to have grown larger. When surrounded by other Douglas-fir, it grew less. They project the possibility that yields would have been greater in the stands that had mixtures of species had the experimental design better accounted for high mortality, but this was not observed from their measurements.

Relevant quote: “For forest managers worldwide, the decision to reforest with one or several tree species is an important one.”

Relevance to landowners:

Comparisons of forest growth between different species combinations of trees are intriguing for landowners and managers. Whether we are managing for timber yield, wildlife habitat, carbon, water, or any other objective, species composition will likely influence the degree to which objectives are met. The basic concept in terms of influencing productivity is something like this: Different species use different resources with different efficiencies. For example, a Douglas-fir may be very efficient at capturing light and using it to grow tall. A hemlock, on the other hand, may be less efficient at light capture but might be more efficient at capturing water or nutrients for root growth. So having the two of them together can theoretically lead to a greater total efficiency in use of all resources. This is known as “niche complementarity” (in ecologyspeak, it is intra-specific competition being greater than inter-specific competition). In this study, they did not observe greater productivity of mixed-species at the stand scale, but they did observe differences in competitiveness between species and the potential for an increase in yield.

Relevance to managers:

There could be several reasons for preferring a mixed-species stand over a single-species stand:

1.     Potential increases productivity (with certain species combinations)

2.     Increased resistance to insect or pathogen outbreaks

3.     Greater canopy stratification and hence diversification in habitat

4.     Hedge-betting against large fluctuations in timber prices between species

The general expectation, supported from this and other studies, is that species with distinct differences in physiological makeup (i.e. shade tolerant and shade intolerant species) are more likely to have niche complementarity, and hence increased productivity. In my book, we are still a LONG way away from knowing how to manage mixed species stands with the same precision that we are now able to manage single-species stands. Many more experiments with multiple species combinations, site productivity treatments, and density treatments are needed.

These studies are certainly relevant for plantations managed for timber, but also have relevance for areas that are planted after severe fires. They also have basic relevance for restoration treatments that alter species composition, and for climate change effects on species composition.

Critique (I always have one, no matter how good the article is) for the pedants:

There were problems with the experimental design that led to biases in the results. When trees died, they sometimes replanted with younger trees. Or they let natural regeneration eventually replace the dead tree. This meant that some trees had neighbors of the same age, and others had neighbors of quite different ages. If a certain species was more likely to die than another in a certain area (very likely), then there would have been a bias. That said, problems like this are hard to avoid with long-term studies (but they could have indeed done some things to mitigate this problem, like plant more seedlings at each spot and then thin them before competition started; or throw out the trees that had a neighbor die).

Their suggestion of mixed stands of Douglas-fir and western hemlock having greater productivity than pure Douglas-fir was not observed from the data. It is based on a projection and an assumption that is not well-explained. I don’t understand how they “corrected for” the mortality issue mentioned above.