Variation of repeated DBH measurements by students in the Ecological Research as Education Network’s (EREN) Pilot Permanent Forest Plot Project (PFPP)

Background/Question/Methods

The Ecological Research as Education (EREN) Network is a group of 123 faculty at 93 Primarily Undergraduate Institutions (PUIs) collaborating on multiple site faculty-student research to answer regional to continental-scale ecological questions.   EREN faculty strive to maximize student engagement in science while generating publication quality data.  Faculty and students are now participating in one of the EREN pilot research projects, the Permanent Forest Plot Project (PFPP), by using a web-accessible common protocol to collect site, plot, and tree data for 20 x 20 m forest plots.  The PFPP will address questions related to tree biomass, carbon accumulation, invasive species, and disturbance across a range of sites and ecoregions with a shared online database.  To test the repeatability of tree diameter-at-breast-height (DBH) measurements by students using the PFPP protocol, we collected three sets of paired DBH data as part of classes or independent study: (1) Ohio Wesleyan University (OH), 12 September 2011 (5 students and 1 faculty) and 17 October 2011 (2 students) (OWU1); (2) Meredith College (NC), 1-15 September 2011 (2 students) and 17 November 2011 (14 students) (MC1); and (3) Meredith College (NC), 25 January 2012, 25 students split into 2 groups (MC2).

Results/Conclusions

We found that paired tree measurements differed significantly in two (MC1 and MC2) of the three datasets; the non-significant difference in OWU1 was due possibly to a small sample size.  Average absolute DBH differences and percent differences between the pairs were: (1) OWU1: 0.66 cm ± .33, 4.2% (n= 20 trees); (2) MC1: 0.26 cm ± 0.33, 3.3% (n= 62 trees); and (3) MC2: 0.19 cm ± 0.42, 2.4% (n= 68 trees).  Two Fraxinus americana trees in OWU1 had the largest DBH differences (5.3 and 4.6 cm smaller on second measurement).  When we removed them for the OWU1 analysis, the DBH differences (0.18 cm ± 0.03, 3.0% [n= 18 trees]) were similar to the MC1 and MC2 results.  DBH measurements taken by an independent student researcher in Fall 2007 and 2011 (MC1) for a larger tree plot dataset (n= 902 trees) showed 0.80 cm ± 0.88 DBH change in four years, or approximately 0.20 cm annually.  Given that our differences in DBH paired measurements are similar to the estimated average annual tree growth, we suggest that investigators not assume that student-measured DBH changes of ≤ 0.2 cm equate to actual tree growth or loss.