Article Figures & Data
Figures
- Figure 1.
Locations of 3 cities in Massachusetts (upper left) and the specific site at which training and evaluation occurred (insets).
- Figure 2.
Examples of trees from (A) Northampton (Carpinus caroliniana), and (B) Springfield (Nyssa sylvatica), Massachusetts.
Tables
- Table 1.
Morphological data for trees used in the evaluation of suggested pruning actions, including, tree number, stem diameter (cm) 1.5 m above ground (diameter at breast height [DBH]), height (m), and number of branches for each tree used in the pruning prescription assessment in (a) Northampton (American hornbeam, Carpinus caroliniana), (b) Springfield (black gum, Nyssa sylvatica), and (c) Greenfield (Princeton elm, Ulmus americana ‘Princeton’).
Location Tree No. DBH Height Number of branches (a) 1 3.8 3.7 22 2 3.6 3.1 29 3 4.6 3.7 20 4 3.8 3.1 19 5 3.8 4.3 27 6 5.1 4.0 33 7 2.8 3.7 21 Mean (std. dev.) 3.9 (0.7) 3.6 (0.45) 24.4 (5.3) (b) 1 5.1 3.7 32 2 8.9 4.0 48 3 7.2 5.2 40 4 5.7 5.2 50 5 4.4 2.4 19 6 5.1 4.6 27 7 4.4 3.1 24 Mean (std. dev.) 5.8 (1.6) 4.0 (0.11) 34.3 (12.0) (c) 1 8.9 4.6 22 2 8.1 4.3 30 3 7.4 4.6 25 Mean (std. dev.) 8.1 (0.8) 4.5 (0.18) 25.7 (4.0) - Table 2.
A list of the knowledge domains, evaluation methods, response variables, and significance tests used to analyze data.
Knowledge domain Evaluation method Response variable Significance test Tree physiology and response to pruning Written exam Score (0-100) Beta regression Good tree structure Evaluation of suggested pruning actions* Score (0-100) Beta regression Good tree structure Evaluation of suggested pruning actions Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Evaluation of removal cut Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Removal cut did not damage branch collar Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Removal cut did not damage branch bark ridge Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Removal cut did not leave a stub Acceptable / Unacceptable Fisher’s exact test Making good pruning cuts Evaluation of reduction cut Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Angle of reduction cut not too steep Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Reduction cut did not damage branch bark ridge Acceptable/Unacceptable Fisher’s exact test Making good pruning cuts Angle of reduction cut not too shallow (did not leave a stub) Acceptable/Unacceptable Fisher’s exact test ↵* Data was analyzed separately by city; for other evaluation methods, data was pooled from 3 cities.
- Table 3.
Participants’ (n = 47) responses to the survey intended to gauge their experience prior to training.
Type of experience with plants Count Conservation 3 Gardening/Yardwork 34 Houseplants 2 Landscaping 2 None 2 Volunteer tree planting 4 Self-rated level of experience with plants Count None or almost no experience 9 A little experience 8 Some experience 16 A lot of experience 12 Extensive Experience 2 - Table 4.
The 5 best beta regression models (measured by Akaike’s Information Criterion corrected for small sample sizes [AICc]) to predict a participant’s score on the written exam from independent variables (type of training, city of training, participant’s previous experience with plants); models were based on pooled data from three sites. Model parameters include the number of predictors (K), AICc, comparison of model AICc with AICc of the best model (ΔAICc), goodness of fit (ModelLik), and log likelihood (LL).
Model K AICc ΔAICc ModelLik LL Training 3 −121.02 0 1 63.80 Intercept only (null) 2 −118.14 2.88 0.24 61.21 City 4 −115.15 5.87 0.05 62.06 Training + Experience 7 −112.05 8.97 0.01 64.50 Experience 6 −108.72 12.30 0.00 61.44 - Table 5.
The 5 best beta regression models (measured by Akaike’s Information Criterion corrected for small sample sizes [AICc]) to predict a participant’s score on the pruning scenario from independent variables (type of training, number of primary branches, tree height); separate models were constructed for (a) Northampton and (b) Springfield. Model parameters include the number of predictors (K), AICc, comparison of model AICc with AICc of the best model (ΔAICc), goodness of fit (ModelLik), and log likelihood (LL).
Model K AICc ΔAICc ModelLik LL (a) Training 3 −26.25 0.00 1.00 16.63 Training + Branches 4 −24.20 2.05 0.36 16.97 Intercept only (null) 2 −21.53 4.73 0.09 13.00 Height 3 −19.26 6.99 0.03 13.13 Branches 3 −19.18 7.08 0.03 13.09 (b) Training 3 −62.38 0.00 1.00 34.74 Training + Branches 4 −61.83 0.56 0.76 35.87 Intercept only (null) 2 −60.03 2.35 0.31 32.28 Branches 3 −59.82 2.57 0.28 33.45 Height 3 −57.76 4.62 0.10 32.43 - Table 6.
Contingency tables for evaluations of removal and reduction cuts, including whether participants who received classroom or hands-on training (a) did not damage the branch bark ridge, (b) did not damage the branch collar on removal cuts or cut at too steep an angle on reduction cuts, (c) did not leave a stub on removal cuts or cut too shallow on reduction cuts, and (d) made a satisfactory cut overall; P-values are from Fisher’s Exact Test; data are pooled from all 3 cities.
Removal Reduction Training Acceptable Unacceptable Total Acceptable Unacceptable Total (a) Classroom 19 1 20 18 2 20 Hands-on 23 3 26 24 2 26 Total 42 4 46 42 4 46 P-value 0.435 0.783 (b) Classroom 15 5 20 16 4 20 Hands-on 24 2 26 24 2 26 Total 39 7 46 40 6 46 P-value 0.105 0.219 (c) Classroom 17 3 20 14 6 20 Hands-on 25 1 26 26 0 26 Total 42 4 46 40 6 46 P-value 0.183 0.003 (d) Classroom 16 4 20 15 5 20 Hands-on 23 3 26 23 3 26 Total 39 7 46 38 8 46 P-value 0.428 0.232 - Table 7.
Logistic regression models to predict whether participants made a reduction cut at the correct angle rather than too shallow (leaving a stub) from independent variables (type of training, participant’s previous experience with plants). Models were based on pooled data from 3 cities. Model parameters include the number of predictors (K), Akaike’s Information Criterion corrected for small sample sizes (AICc), comparison of model AICc with AICc of the best model (ΔAICc), goodness of fit (ModelLik), and log likelihood (LL).
Model K AICc ΔAICc ModelLik LL Training 2 28.71 0.00 1.00 −12.22 Intercept only (null) 1 37.71 9.00 0.01 −17.81 Experience 5 37.97 9.25 0.01 −13.23
Supplemental Material
PowerPoint lecture used to train one group of participants.
- Supplemental Powerpoint - PowerPoint lecture used to train one group of participants.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.