Several variables were significant predictors of how often NRCS personnel recommend prescribed fires to land managers for controlling brush. There were differences among all levels of the response to how often NRCS personnel recommend prescribed fire (Table 1; never= 0, rarely= 1, occasionally = 2, frequently = 3). Holding all other variables constant, how often NRCS personnel recommend prescribed fire to landowners is 3.12 times more likely to go up a level if interactions regarding prescribed fire with landowners are reported as positive rather than negative (see odds ratio from Table 1., 4.12-1 \(p=0.035\) and (Fig. 1). There was complete separation in levels of responses for how often NRCS personnel are asked to deal with brush management and how often they recommend prescribed fire, so odds ratios are not interpretable, but it is a significant explanatory variable. Rate of priority for prescribed fire education in the job is also a significant explanatory variable. How often NRCS personnel recommend prescribed fire for controlling brush is 0.58 times more likely to increase by a level with every one unit increase in their self-reported rate for priority of prescribed fire education in their job (\(p=0.010\)).

Table 1: Output for ordinal logistic regression modeling how often NRCS personnel recommend prescribed fire
  Prescribed Fire Recommendations
Predictors Odds Ratios Conf. Int (95%) p-Value df
0|1 1672.96 1011.99 – 2765.64 0.008 113.00
1|2 20994.17 13488.72 – 32675.83 0.001 113.00
2|3 460196.72 29843.33 – 7096426.25 <0.001 113.00
Rank of livestock
production 		1.08 0.64 – 1.77 0.763 113.00
Rank of Wildlife
operation 		0.80 0.51 – 1.24 0.325 113.00
Interaction with
landowners re. brush
management (Neutral) 		1.18 0.08 – 18.45 0.904 113.00
Interaction with
landowners re. brush
management (Positive) 		0.18 0.02 – 1.83 0.145 113.00
Interaction with
landowners re. RX
(Neutral) 		1.74 0.61 – 5.12 0.310 113.00
Interaction with
landowners re. RX
(Positive) 		4.12 1.13 – 15.41 0.035 113.00
Years of formal education 0.94 0.71 – 1.26 0.682 113.00
How often asked to deal
with brush mgt. - linear 		31.68 0.23 – NA 0.020 113.00
How often asked to deal
with brush mgt. -
quadratic 		430.51 <0.001 113.00
How often asked to deal
with brush mgt. - cubic 		0.00 <0.001 113.00
How often recommend
mechanical and chemical
trt 		2.03 1.26 – 3.34 0.005 113.00
Level of knowledge re. RX
fire - others 		0.88 0.71 – 1.08 0.219 113.00
Level of comfort with RX
fire - others 		1.05 0.88 – 1.26 0.605 113.00
Gender (Female) 0.67 0.27 – 1.68 0.397 113.00
Level of knowledge re. RX
fire - yourself 		1.24 0.61 – 2.52 0.559 113.00
Level of comfort with RX
fire - yourself 		1.96 1.01 – 3.88 0.052 113.00
Rate for priority of RX
fire education in job 		1.58 1.13 – 2.23 0.010 113.00
Understand difference RX
vs. wildfire 		1.44 0.82 – 2.53 0.206 113.00
Awareness re. laws/regs
re. burn bans 		1.08 0.62 – 1.90 0.790 113.00
Consider landowner
liability in RX
recommendation 		0.95 0.57 – 1.58 0.844 113.00
Observations 136
R2 Nagelkerke 0.849
Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for how oftern NRCS personnel recommend prescribed fire for brush management (x axis) by their rating for priority for prescribed fire education in their position (y axis). Each panel represents the response distribution for different categores of how they described their interactions with landowners about prescribed fire (positive, neutral, negative, or not applicable).

(#fig:Fig.1)Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for how oftern NRCS personnel recommend prescribed fire for brush management (x axis) by their rating for priority for prescribed fire education in their position (y axis). Each panel represents the response distribution for different categores of how they described their interactions with landowners about prescribed fire (positive, neutral, negative, or not applicable).

Several variables were also significant predictors of NRCS self-rated level of knowledge regarding prescribed fire. Holding all other variables constant, knowledge of fire increases by 0.02 - 0.12 (\(p=0.009\)) for each one unit increase in perceived knowledge of others related to prescribed fire (the combined knowledge scores of the general public, landowners, county commissioners, and district court judges; Fig. 2) Holding all other variables constant, NRCS personnel’s level of knowledge of prescribed fire increases by 0.40 - 0.68 (\(p<0.001\); Table 2) for every one unit increase in their level of comfort with prescribed fire. While the cubic term for how often NRCS personnel are asked to deal with brush management is a significant variable, A likelihood ratio test of the full model compared to the model reduced by that variable shows that overall this is not a significant predictor (\(\chi^2=1.56\), \(p=0.114\)).

Table 2: Linear regression output modeling how NRCS personnel self-rate their level of knowledge about prescribed fire
  Knowledge of Prescribed Fire
Predictors Estimates Conf. Int (95%) p-Value df
(Intercept) -1.11 -2.60 – 0.38 0.142 116.00
Rank of livestock
production 		0.05 -0.08 – 0.18 0.471 116.00
Rank of Wildlife
operation 		0.07 -0.03 – 0.18 0.174 116.00
Interaction with
landowners re. brush
management (Neutral) 		-0.18 -0.90 – 0.55 0.631 116.00
Interaction with
landowners re. brush
management (Positive) 		-0.10 -0.73 – 0.52 0.741 116.00
Interaction with
landowners re. RX
(Neutral) 		0.04 -0.23 – 0.31 0.744 116.00
Interaction with
landowners re. RX
(Positive) 		0.02 -0.31 – 0.35 0.920 116.00
Years of formal education 0.05 -0.02 – 0.13 0.187 116.00
How often asked to deal
with brush mgt. - linear 		0.07 -0.76 – 0.89 0.876 116.00
How often asked to deal
with brush mgt. -
quadratic 		-0.82 -1.66 – 0.03 0.057 116.00
How often asked to deal
with brush mgt. - cubic 		1.01 0.16 – 1.85 0.020 116.00
How often recommend
mechanical and chemical
trt 		-0.01 -0.13 – 0.11 0.856 116.00
Level of knowledge re. RX
fire - others 		0.07 0.02 – 0.12 0.009 116.00
Level of comfort with RX
fire - others 		-0.03 -0.08 – 0.01 0.178 116.00
Gender (Female) -0.04 -0.28 – 0.20 0.747 116.00
Level of comfort with RX
fire - yourself 		0.54 0.40 – 0.68 <0.001 116.00
Rate for priority of RX
fire education in job 		0.00 -0.08 – 0.09 0.911 116.00
Understand difference RX
vs. wildfire 		0.05 -0.10 – 0.20 0.487 116.00
Awareness re. laws/regs
re. burn bans 		0.02 -0.12 – 0.16 0.733 116.00
Consider landowner
liability in RX
recommendation 		0.05 -0.08 – 0.18 0.451 116.00
Observations 136
R2 / R2 adjusted 0.528 / 0.451
Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for NRCS personnel rated their knowledge regarding prescribed fire (x axis) by their rating for knowledge of others (general public, landowners, county commissioners, and district court judges; y axis).

(#fig:Fig.2)Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for NRCS personnel rated their knowledge regarding prescribed fire (x axis) by their rating for knowledge of others (general public, landowners, county commissioners, and district court judges; y axis).

There were also significant variables in the model for how NRCS personnel rate their level of comfort with prescribed fire (Table 3). Holding all other variables constant, NRCS level of comfort with prescribed fire decreases by 0.14 with every one-unit increase in the rank of livestock production’s relative abundance as a land use in the area of operation (\(p=0.41\); Fig. 3, top panel). How often NRCS personnel are asked to deal with brush management is a significant variable also, with a decrease in comfort with increasing time spent on brush management (\(\chi^2=6.32\), \(p<0.001\)). Both the quadratic and the cubic terms are significant (Table ??). Level of comfort decreases by 0.2-0.13 (\(p=0.11\)) with every one-unit increase in others’ knowledge of prescribed fire. While comfort increases by 0.01-0.10 (\(p=0.23\)) with every one-unit increase in others’ level of comfort with prescribed fire Fig. 3, middle panel. Similarly, comfort with prescribed fire increases by 0.46-0.79 (\(p<0.001\)) with every one-unit increase in their own knowledge of prescribed fire (Fig. 3, bottom panel) and it increases by 0.13-0.41 (\(p<0.001\)) with every one-unit increase in their awareness of laws and burn ban regulations.
Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for how comfortable NRCS personnel are with prescribed fire (x axis) by their rating for the relative importance of livestock production as a land use in the region they service (y axis, top panel), their rating for how comfortale others are with prescribed fire (y axis, middle panel), and their own knowledge regarding prescribed fire (y axis, bottom panel)

(#fig:Fig.3 )Boxplots showing median and interquartile range (boxes) and the range (whiskers) of responses for how comfortable NRCS personnel are with prescribed fire (x axis) by their rating for the relative importance of livestock production as a land use in the region they service (y axis, top panel), their rating for how comfortale others are with prescribed fire (y axis, middle panel), and their own knowledge regarding prescribed fire (y axis, bottom panel)

Table 3: Linear regression output modeling how NRCS rate their level of comfort with prescribed fire
  Comfort with Prescribed Fire
Predictors Estimates Conf. Int (95%) p-Value df
(Intercept) 2.10 0.53 – 3.67 0.009 116.00
Rank of livestock
production 		-0.14 -0.28 – -0.01 0.041 116.00
Rank of Wildlife
operation 		-0.05 -0.17 – 0.06 0.382 116.00
Interaction with
landowners re. brush
management (Neutral) 		-0.16 -0.94 – 0.63 0.693 116.00
Interaction with
landowners re. brush
management (Positive) 		-0.12 -0.79 – 0.55 0.717 116.00
Interaction with
landowners re. RX
(Neutral) 		0.02 -0.27 – 0.32 0.867 116.00
Interaction with
landowners re. RX
(Positive) 		0.14 -0.21 – 0.50 0.426 116.00
Years of formal education -0.04 -0.12 – 0.05 0.381 116.00
How often asked to deal
with brush mgt. - linear 		0.29 -0.59 – 1.17 0.519 116.00
How often asked to deal
with brush mgt. -
quadratic 		1.65 0.78 – 2.52 <0.001 116.00
How often asked to deal
with brush mgt. - cubic 		-1.92 -2.78 – -1.06 <0.001 116.00
How often recommend
mechanical and chemical
trt 		-0.05 -0.18 – 0.08 0.446 116.00
Level of knowledge re. RX
fire - others 		-0.07 -0.13 – -0.02 0.011 116.00
Level of comfort with RX
fire - others 		0.06 0.01 – 0.10 0.023 116.00
Gender (Female) -0.23 -0.48 – 0.02 0.076 116.00
Level of knowledge re. RX
fire - yourself 		0.63 0.46 – 0.79 <0.001 116.00
Rate for priority of RX
fire education in job 		0.06 -0.03 – 0.15 0.194 116.00
Understand difference RX
vs. wildfire 		0.04 -0.12 – 0.20 0.598 116.00
Awareness re. laws/regs
re. burn bans 		0.27 0.13 – 0.41 <0.001 116.00
Consider landowner
liability in RX
recommendation 		-0.11 -0.24 – 0.03 0.123 116.00
Observations 136
R2 / R2 adjusted 0.654 / 0.597