Betsy Lyons
Crown Structure and Spatial Distribution of Epiphytes in western hemlock trees. Forest Canopy Workshop, April 2002

Mix-n-Match Scenarios

After looking over the datasets, I came up with 3 potential mix-n-match scenarios. However, I haven't proceeded so far as to work out all the details that would actually be necessary (e.g. specific data conversions, analysis of sample size and selection of specific statistical analyses). I have specific questions regarding some of the datasets that would be useful to discuss in more detail at the workshop.

1. Comparative analysis of structural data
Some of the structural data collected at different sites could be compared to look for spatial patterns. For example, how does the distribution of branches (by height or size) within individual tree crowns or within height ranges differ between forest types (e.g. old-growth conifers vs. hardwood; young vs. old)? Although this type of data could be useful for developing protocols and tools for combining and/or comparing data, it might not be that useful for answering new questions.

2. Extrapolation across spatial scales
Using structural data from other studies conducted at the same site or a similar site could be useful for generating hypotheses for future studies, or depending on reliability of data, for modeling. My dataset contains information on the distribution of epiphytes among three height classes of hemlock trees at the WRCCRF. Structural data on the distribution of hemlock trees throughout the entire research site is also available. Would it be possible to combine these datasets and extrapolate to the stand level to estimate the relative abundance of epiphytes throughout the entire site? or to estimate the distribution and abundance of epiphytes within a given height range?or for a given size class of branches?

The usefulness of this type of analysis would depend largely on statistics used and the sample size of each dataset. Extensive data conversion could be necessary to ensure that data was being compared appropriately (e.g. by relative height within trees or absolute height above ground).

3. Correlations among structural and biological data
Data from Shaw's study of hemlock dwarf-mistletoe and my data on epiphyte distribution and crown structure (collected at the same site) could be combined to look for correlations. One hypothesis that could be examined is whether the distribution of dwarf
mistletoe affects the distribution and abundance of epiphytes or vice versa. Do branches or trees heavily infested with mistletoe have a greater or lesser amount of epiphytes than un-infected trees?

However, if correlations were found it would still be difficult to ascertain whether the effects were due to direct competition for space between epiphytes and mistletoe, or changes in the physical structure of branches, or other related factors such as changes in tree chemistry or physiology.