Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

Dianne A Stroman; Urs P Kreuter; Carissa L Wonkka

doi:10.1371/journal.pone.0238688

. 2020 Sep 8;15(9):e0238688. doi: 10.1371/journal.pone.0238688

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

Dianne A Stroman ^1,^*,^#, Urs P Kreuter ^2,^#, Carissa L Wonkka ^3,^#

Editor: Francisco X Aguilar⁴

PMCID: PMC7478812 PMID: 32898171

Abstract

Grassland environments face a number of threats including land use change, changing climate and encroachment of woody plants. In the Southern Plains of the United States, woody plant encroachment threatens traditional agricultural grazing economies in addition to grassland dependent wildlife species. Numerous studies have examined the physical drivers of conversion from grassland to woodland but social drivers may be equally important to understanding the causes of and prescriptions for environmental degradation. In this paper, we report the results of a survey of landowners in the Southern Plains of Texas and Oklahoma in which we asked participants to estimate the current amount of woody plant cover on their land, their preferred amount of woody plant cover and about their perspectives regarding the use of prescribed fire for managing woody plants. Prescribed fire is ecologically and economically one of the most effective tools for maintaining grasslands but many landowners do not use this tool due to lack of knowledge, lack of resources and concerns over safety and legal liability. We found that while most of our respondents did express a desire for less woody plant cover on their land, woody plant preference did not affect landowner’s use of prescribed fire. However, belonging to a prescribed burn association and owning larger properties were correlated with increased use of prescribed fire. Woody plant cover preference was significantly influenced by landownership motivations, with hunters and other recreational motivated landowners preferring more trees and ranchers preferring fewer. This is important because throughout most of our study area, there has been a steady shift from agricultural production to amenity or recreational landownership, a trend that may undermine efforts to restore or maintain open grasslands. Future outreach efforts to promote prescribed fire to maintain grasslands should more actively support prescribed burn associations, which is an effective vehicle for increasing prescribed fire use by private landowners.

Introduction

Throughout many historically grassland-dominant ecoregions, there has been a shift towards increasing concentration of woody plant species. Within our study area, the Southern Plains of the USA, increases in woody cover have ranged from 10% in the 1930’s to over 30% today, a trend that is particularly prevalent in areas with more development [1]. This conversion from grassland to woodlands often has cascading effects on local floral and faunal communities and can alter abiotic factors such as soils and local hydrology [2,3]. Additionally, grasslands that have converted to woodlands have lower livestock production potential and are more susceptible to wildfires [4]. The reasons behind this vegetation shift include climate change, elimination of fire from the landscape, overgrazing by livestock and shifts in land use [1,5].

The increase in woody plant cover has encouraged the development of a variety of programs at the local, state and federal levels designed to restore grasslands. Examples include the Brush Busters program, Prescribed Burn Associations (PBAs), Texas Parks and Wildlife Department’s Habitat for Upland Birds, and the Natural Resources Conservation Service EQIP program [6,7]. These programs provide financial support, training, education and other incentives to combat woody plant invasion on grasslands. While such programs play an important role for landowners who need help maintaining productive grasslands, the overall problem of woody plant expansion in grasslands continues to increase.

In this paper, we report the results of a survey of landowners in the Southern Plains regions of Texas and Oklahoma in which we asked the survey participants to estimate the amount of woody plant cover on their rural land, and about their perspectives regarding the use of prescribed fire as a woody plant management tool. Accurately estimating woody cover density is difficult from ground level because visual foreshortening of dispersed trees seen from that perspective may suggest higher woody plant density than overhead measurements of canopy cover would indicate. Overestimation of the scope of a problem may lead some people to assume it is too difficult or too risky to deal with, leading to inaction [8]. On the other hand, underestimation of a problem can also lead to inaction because the issue does not seem urgent. We also asked landowners what the preferred amount of woody cover is for their land. Knowing landowners’ preference for woody plant cover is important because if landowners prefer more woody plants, then they are unlikely to reduce them or seek outside assistance designed to restore grasslands.

We also examine whether woody plant cover preferences influenced landowner use of prescribed fire. While prescribed fire has been found to be ecologically and economically one of the most effective tools for maintaining grasslands [4,9,10], many landowners do not use this land management tool due to lack of knowledge and resources to apply it safely and concerns over legal liability [11,12]. Here, we explore if landowners’ preferences for woody plant cover may be an additional social driver of landscape scale transition from grasslands to woodlands.

We address three key questions. 1) Do landowner perceptions about woody plant expansion correspond with estimated changes in their area? 2) Do landowners within the Southern Plains desire less woody cover on their land? 3) Do expressed woody cover preferences influence the adoption of prescribed fire on private land?

Methodology

Study area

The study reported here was conducted in the Southern Plains region of Texas and Oklahoma (Fig 1), where the dominant invasive woody plants are native mesquite (Prosopis glandulosa), Ashe juniper (Juniperus ashei), redberry juniper (J. pinchotii), eastern redcedar (J. virginiana) and exotic Chinese tallow (Triadica sebiferium). The 16 counties we selected for the study were chosen because they incorporated an active PBA and because the role of prescribed fire was central to the overall study. This allowed us to include both general landowners and PBA members in our survey sample. In Texas, we focused on an eight county region within the Edwards Plateau eco-region in the central part of the state. Those eight counties were: San Saba, Llano, Mason, Gillespie, Kimble, Menard, Sutton and Schleicher. In Oklahoma, our sampled counties were clustered in the eastern and northcentral parts of the Rolling Plains eco-region within the state. The eight counties in Oklahoma included: Beckham, Comanche, Dewey, Ellis, Roger Mills, Tillman, Pawnee and Payne.

Mail survey

This study included human study participants. It was approved by the Texas A&M University IRB board. Consent came in the form of participants choosing to fill out and return a survey questionnaire. All participant data was analyzed anonymously. The survey included 1,918 landowners in Texas and Oklahoma who owned a minimum of 40.5 ha (100 acres). We applied this minimum size requirement with the assumption that landowners with smaller sized properties are less likely to apply prescribed fire, regardless of their perspectives of woody plant density. This survey sample was derived using a stratified sample selection approach. The first stratum consisted of general landowners included in the tax databases of each of the 16 selected counties. From each of these databases, 100 landowners who owned at least 40.5 ha were randomly selected for a first stratum sub-sample of 1,600 landowners. The second stratum consisted all members of PBA in the 16 selected counties. The population of PBA members in each of the two states was small enough that our PBA sample consisted the whole subgroup. This resulted second stratum sub-sample of 318 PBA members (126 in Texas and 192 in Oklahoma). To ensure that the PBA members were not a biased subset of the combined sample of 1,918 landowners, we compared responses from the two subsamples regarding perceptions of woody plants and prescribed fire and about the proportion of income derived from the property.

In order to assess landowners’ perceptions and preferences concerning woody plant cover and the use of prescribed fire on their land, we asked the survey participants a series of questions as part of a broader mail survey about woody plant expansion and the role of prescribed fire in the Southern Plains (S1 File). During initial development, the survey questionnaire was provided to selected stakeholders for feedback and content recommendations, after which it was revised before mailing.

The survey was administered using a five-phase mail protocol [13] consisting of: initial letter to inform landowners about the study (day 1); survey questionnaire with postage paid return envelope (day 7); reminder/thank-you postcard (day 21); replacement questionnaire to non-respondents (day 42); and a final reminder/thank you postcard to non-respondents (day 56). The survey was conducted in October and November 2015 and completed questionnaires were accepted for 150 days from the survey start date.

A non-response bias analysis was not conducted because alternative contact information (e.g., telephone number) required for such an analysis [13] was not obtainable. In the absence of non-response data, we compared data from early and later respondents (i.e., first and second wave responders) as a means of inferring non-response bias from differences between these two groups [14]. Respondents were included in one of two groups based on receipt of their questionnaire during or after the first 3 weeks from the mailing date of the questionnaire (i.e., first and second wave responders). We used Mann-Whitney rank tests to compare the variable values of these two groups for five key variables: age of respondent; reliance on income from their property; perception of and preference for woody plant cover on their land, and whether or not they thought woody plants were a problem on their land.

Measuring woody cover perceptions and preferences

Survey participants’ perceptions of the amount of woody plant cover on their property was assessed by asking if specific types of woody plants (cedar/juniper–Juniperus spp. and mesquite–Prosopis glandulosa) had increased on their property in the last ten years. Using a categorical visual question prompt (Fig 2), participants were also asked to estimate the current proportion of trees to grassland and their preferred amount of woody cover for their land.

Fig 2 — Question prompts were: 1) If you were to look down on your land from above, which picture do you feel would best represent the proportion of trees/shrubs (black boxes) to grassland (gray boxes) found on your property? 2) Which proportion of trees/shrubs (black boxes) to grassland (gray boxes) would you prefer on your property?

Using landowners’ responses to whether or not woody plant cover had increased on their land, we compared average response rates on a county by county basis to published estimates of changes in woody cover between 2004–2014 [15]. While this did not allow us to assess whether woody cover on individual properties had actually increased during that 10-year period, it did allow us to determine if landowners’ perceptions of changes in woody plant cover in a specific county corresponded with cover change data for that county.

To gauge the difference between individual landowner’s current cover perceptions and their stated woody cover preferences, we subtracted the numeric value for their answers to the two questions referred to in the visual prompt in Fig 2. By subtracting the respondent’s current amount of woody cover estimate from their cover preference a number between -3 and 3 was obtained. Negative numbers indicate respondents’ desire for less woody plant cover then they currently have, zero indicates respondents’ satisfaction with their current woody plant cover, and positive values indicate that respondents prefer more woody plant cover.

Data and regression analyses

Data were initially analyzed using frequency distribution metrics. We used null values for unanswered questions, resulting in some variation of n values for each variable. We based this on the assumption that data were missing completely at random (MCAR) as we could not determine respondent’s reasons for not answering any particular question. Based on this assumption, we did not impute MCAR data.

In addition to the visual question prompt asking respondents about their preferred amount of woody cover (Fig 2), we predicted that a variety of variables would influence woody cover preferences. Using both demographic and landowner specific variables (Table 1) we used ordinal logistic regression models to estimate woody cover preferences. Woody cover preference was used as the dependent variable in our model. Independent variables in our models included demographic controls (age, gender, state of residence) and variables we predicted would influence woody cover preference (residency on property–full time, occasional, absentee; property size; annual income from property; PBA membership; and landownership motivations).

Table 1. Dependent and independent variables used in cover preference regression model.

Dependent Variables	Variable Descriptions
Woody cover preference	Categorical variable response to question, “Which proportion of trees/shrubs to grassland would you prefer on your property?”: approximately 5%, approximately 25%, approximately 50%, approximately 75%
Independent Variables
PBA membership	Binary single item variable (0 = No, 1 = Yes)
State of property location	Binary single item variable (0 = Oklahoma, 1 = Texas)
Gender	Binary single item variable (0 = female, 1 = male)
Age (years) in 2015	Continuous single item variable
Education level	Ordinal variable: High school (reference category), some post-secondary/bachelor degree, graduate/professional degree.
Property size	Ordinal variable: 40.5–202 ha (100–500 acres = small, reference category), 203–404 ha (501–1000 acres = medium), > 404 ha (>1000 acres = large)
Period of residence on property	Ordinal variable: absentee resident (reference category), occasional resident, and full time resident
Income from property	Ordinal response for, “In 2014, approximately what percent of your total annual income was generated from activities on your rural property?” 0% (reference category), 1–25%, 26–50%, 51–75%, 76–100%.
Lifestyle/recreation as ownership motivator	PCA derived latent variable (answer scale ranging in value from 1 = not at all important to 7 = very important)
Ranching as ownership motivator	PCA derived latent variable (answer scale ranging in value from 1 = not at all important to 7 = very important)
Heritage as ownership motivator	PCA derived latent variable (answer scale ranging in value from 1 = not at all important to 7 = very important)
Hunting as ownership motivator	PCA derived latent variable (answer scale ranging in value from 1 = not at all important to 7 = very important)
Farming as ownership motivator	PCA derived latent variable (answer scale ranging in value from 1 = not at all important to 7 = very important)

Open in a new tab

Table 2 reports on the results of the principal components analysis (PCA) used to reduce respondents’ landownership motivations into five latent variables for inclusion in the regression model. These latent variables represent the ownership motivation categories listed in Table 1. We used Cronbach’s alpha (α) > 0.6 to identify internal reliability of each latent variable [16,17]. These latent variables were used as independent variables in subsequent regression models examining woody plant cover preferences and use of prescribed fire. All statistical analyses were completed using STATA 14 software [18].

Table 2. Orthogonal varimax rotated factor loading results of PCA analysis of landownership motivations with Cronbach’s α measuring internal scale reliability.

Landownership motivations	Lifestyle/ recreation α = 0.8574	Ranching α = 0.8526	Heritage α = 0.9206	Hunting α = 0.6289	Farming α = 0.6693	Mean response score^a
Place to relax	0.8883	0.0195	0.0572	0.0747	0.0117	6.18
Enjoy the outdoors	0.8132	0.1247	0.1615	0.0798	-0.0706	6.51
Hunting/fishing (recreational)	0.8020	-0.0949	0.0590	0.2146	0.0697	5.89
Non-hunting/fishing recreation	0.8386	0.0450	-0.0060	0.2309	0.0412	5.89
Operate a farm/ranch	0.0702	0.8709	0.1704	0.0710	0.0980	6.15
Maintain family farming/ranching tradition	0.0004	0.6943	0.5299	0.0843	0.1167	5.92
Produce grazing livestock	0.0185	0.8561	0.1962	0.0040	0.0942	6.06
Earn a profit	-0.0270	0.6702	0.1280	0.0774	0.3765	5.68
Keep land in family	0.0669	0.2411	0.917	0.0169	0.0608	6.43
Leave land for family	0.1050	0.1664	0.918	0.0655	0.0738	6.42
Operate a hunting enterprise	0.0177	0.2631	0.0877	0.8071	-0.0769	3.90
Manage large wildlife (deer)	0.2923	-0.0128	-0.0010	0.8657	-0.0793	4.82
Manage game birds	0.2770	-0.0773	0.0799	0.7387	0.1939	4.62
Produce hay/forage	0.0599	0.3569	0.1370	-0.2315	0.6425	4.63
Cultivate crops	-0.0216	0.1995	0.1282	-0.0159	0.7628	3.61
Obtain income from minerals	-0.0988	0.0775	0.1841	0.1023	0.6747	4.02
Have a financial investment	0.2230	0.1096	-0.1402	0.0359	0.6175	5.49
Eigenvalue	4.6	3.4	1.7	1.5	1.1	-
% explained variance	17.9	16.5	12.8	12.6	12.2	-

Open in a new tab

Shaded results indicate variables that load on a particular factor.

^a Mean response scores based on Likert scale 1 = not at all important to 7 = very important.

In order to test our prediction that landowners who prefer less woody cover on their property will be more likely to use prescribed fire, we developed a regression model using previous use of prescribed fire as the dependent variable. Independent variables for this model included: demographic controls (age, gender, state of residence) and other variables we predicted would influence use of prescribed fire (woody cover preferences; residency on property–full time, occasional, absentee; property size; PBA membership; landownership motivations).

Results

Response rates and non-response bias assessment

Of the 1,918 survey questionnaires mailed, 65 were returned as undeliverable, lowering our sample size to 1,853. A total of 680 useable questionnaires were returned representing an overall 36.7% useable response rate. However, response rates differed by state and by category of survey participants–general landowners versus PBA member landowners (Table 3). Overall, the response rate was higher in Texas than in Oklahoma and PBA members were much more likely than general landowners to participate in this study.

Table 3. Survey questionnaire response rates.

Response Rates by group	Responses received	% of total respondents	Intra-group response rate
Texas general landowners	272	40.0	(272/800) = 34%
Texas PBA members	112	16.5	(112/126) = 89%
Oklahoma general landowners	192	28.2	(192/800) = 24%
Oklahoma PBA members	104	15.3	(104/192) = 54%
Total	680	100%

Open in a new tab

The Mann-Whitney rank tests conducted to determine if there was potential response bias among PBA members compared to general respondents found no statistically significant differences between these two groups of respondents with respect to the proportion of annual income derived from the property or about existing and preferred woody plant density (Table 4). However, PBA members were more likely to respond that woody plants are a problem on their land (non -members also agreed with that statement but not as strongly), and they were much more likely to favor the application of prescribed fire to their own land (both were p < 0.0001).

Table 4. Mean response scores from PBA members and non-member survey questionnaire responses testing non-random participation bias.

Survey question	PBA mean (n = 216)	Non-PBA mean (n = 464)	% diff. in means	Mann-Whitney sig.
% annual income earned from property ^a	2.9	2.6	11.5	0.0177
Current proportion of woody plant to grassland on property ^b	2.7	2.6	3.8	0.3289
Preferred proportion of woody plants to grassland on property ^b	1.7	1.9	11.7	0.0771
Woody plants are a problem on my land ^a	6.3	5.7	9.5	<0.0001
In favor of applying prescribed fire on my land ^a	6.5	4.9	32.7	<0.0001

Open in a new tab

^c Mean response score based on “income from property” variable categories listed in Table 1.

^b Mean response score based on “woody cover preference” variable categories listed in Table 1.

^a Mean response scores based on Likert scale 1 = strongly disagree to 7 = strongly agree.

Additionally, Mann-Whiney tests used to compare early and later respondent data identified no statistically significant differences between the two groups with respect to age of respondent, reliance on income from their property, perception of and preference for woody plant cover on their land, and whether or not they thought woody plants were a problem on their land (p = 0.0707–0.7721).

Landowner characteristics

Respondents were predominantly older (average age of 66 years), male (81%) and well educated (85% reported at least some college) (Table 5). Over half of the respondents (57%) lived in Texas and over half (54%) lived on their rural land full time. Approximately equal proportions of respondents owned properties that were 45.5–404 ha (100–1,000 ha) and > 405 ha (1,000 acres) in size. The large majority (84%) of the respondents reported earning some of their annual income from their rural land, and over a quarter (28%) reported they earned more than half of their income from activities on their rural land. About a third of respondents (32%) reported being a member of a PBA.

Table 5. Survey respondents’ demographic and property characteristics.

Demographic variable	Survey respondents
Age	Range = 30–93, mean = 66 years, std. dev. = 11
Gender
Male	81%
Female	19%
Education
Some high school/high school graduate	15%
Some college/college graduate	50%
Some graduate school/graduate degree	35%
State of residence
Texas	57%
Oklahoma	43%
Residency on rural property
Full time resident	54%
Part-time/weekend resident	19%
Non-resident	27%
Property size
40.5–202 ha (100–500 acres–small)	29%
203–404 ha (501–1000 acres–medium)	20%
> 406 ha (>1000 acres–large)	51%
Annual income from property
0%	16%
1–25%	40%
26–50%	16%
51–75%	14%
76–100%	14%
PBA member (self-identified)
Yes	32%
No	68%

Open in a new tab

Woody plant cover perceptions and preferences

When we asked participants if cedar or juniper had increased on their property in the last ten years, 68% responded that it had (36% strongly agreed) and 52% of respondents agreed that mesquite had increased on their land in the last ten years (24% strongly agreed). However, we also asked participants to estimate their current and preferred amount of woody plant cover based on a visual schematic (Fig 2). Fig 3 presents our findings about how the survey respondents’ perceptions about woody cover on their land compared with their preferred amount of woody cover. In Texas, 80% of respondents estimated they had 50% or more woody plant cover on their land, whereas 67% of the Oklahoma respondents reported they had 25% or less woody plant cover. In both states, the large majority of respondents preferred less woody cover than they estimated they had on their land. Overall, 78% of Texas respondents and 94% of Oklahoma wanted woody cover on their land to be 25% or less, but the predominantly preferred woody cover was 25% in Texas (64% of respondents) and 5% in Oklahoma (64% of the respondents). Therefore, in both states, respondents preferred less woody plant cover but the Oklahoma respondents expressed a stronger preference for more open grassland.

Fig 4 presents the degree of separation between each respondent’s preferred and estimated woody plant coverage. A zero value indicates those respondents whose preferred and estimated coverage were the same, i.e., they are satisfied with their current amount of woody plants, whereas negative values increasingly indicate respondents’ desire for less cover then they currently have. Approximately 69% of respondents indicated that they preferred less woody cover with over half desiring a 25% reduction in woody cover.

We were also interested in how respondents’ perceptions of woody plant cover change matched with satellite data measuring cover in the sampled counties. We were able to obtain Landsat data measuring woody plant cover increases or decreases between 2004–2014 in our sampled counties, some of which are available in the published literature [14]. While the Landsat data, presented in Table 6, reflects entire counties and not the respondents’ specific properties, it does serve to highlight some interesting dichotomies. The majority (>50%) of respondents in all but Tillman county in Oklahoma perceived that woody plant cover increased on their land during the preceding 10 years, and in nine of the 16 counties in the study more than 75% of the respondents believed this to be the case (Table 6). However, in 2018 Hartfield and van Leeuwen reported that, based on Landsat data, only nine of these counties including seven in Oklahoma reflected woody plant increases while six of the counties in Texas experienced a 1–10% decrease during that time period [15].

Table 6. Landowner estimates of woody cover increases and documented woody cover increases by county between 2004–2014.

State/County	Approximate woody plant cover % 2004 2014	% change in woody cover within county 2004–2014^a	% of respondents who perceived of woody plant increases in their county ^b
Texas
Schleicher	17.3% 20.0%	2.8%	78%
San Saba	25.1% 25.2%	0.1%	79%
Menard	24.3% 23.5%	-0.8%	69%
Sutton	26.0% 23.0%	-2.9%	88%
Gillespie	35.9% 32.6%	-3.4%	76%
Llano	27.9% 23.3%	-4.6%	71%
Mason	28.1% 22.6%	-5.5%	54%
Kimble	34.8% 24.4%	-10.4%	87%
Oklahoma
Pawnee	24.7% 36.3%	11.6%	80%
Payne	25.3% 36.5%	11.3%	78%
Beckham	5.4% 9.0%	3.6%	69%
Dewey	11.0% 14.6%	3.6%	87%
Roger Mills	5.4% 8.8%	3.4%	77%
Comanche	10.7% 13.7%	3.0%	73%
Tillman	3.7% 4.9%	1.2%	35%
Ellis	5.1% 4.4%	-0.7%	70%

Open in a new tab

^a Bolded values indicate increasing woody cover in that county between 2004–2014 [15]

^b % of respondents within county indicating increased woody cover in last 10 years

Regression analyses

The ordinal logistic regression model for factors influencing respondents’ preference for woody plant cover is presented in Table 7. The most common drivers of cover preference pertain to landownership motivations. Respondents who owned their land for lifestyle or recreational purposes and those who owned their land for hunting purposes were significantly more likely to prefer more woody plant cover, while those who owned their land for ranching purposes were more likely to prefer less woody plant cover and more grassland for their livestock. In addition, the regression model emphasized that Texas respondents preferred a greater amount of woody cover then Oklahoma respondents. Older respondents were also slightly more likely to prefer less woody cover on their land.

Table 7. Ordinal logistic regression model for factors influencing woody plant cover preferences of respondents in Texas and Oklahoma.

(Bold items indicate P <0.05).

Independent Variables	Woody cover preference model: Pseudo R² = 0.206; P<0.001
Independent Variables	β coeff.	Std. err. β coeff.	p-value	% Δ odds
Lifestyle/recreation motivation	0.179	0.095	0.050	19.7
Rancher motivation	-0.442	0.111	<0.001	-34.5
Heritage motivation	-0.151	0.089	0.091	-14.1
Hunter motivation	0.492	0.112	<0.001	63.7
Farmer motivation	-0.178	0.108	0.101	-16.3
Occasional resident ^a	0.306	0.274	0.265	35.8
Full time resident ^a	-0.024	0.231	0.919	-2.3
State	1.761	0.245	<0.001	482.0
PBA member	-0.186	0.204	0.360	-17.0
Income = 1–25% ^b	0.205	0.299	0.493	22.8
Income = 26–50% ^b	-0.387	0.390	0.322	-32.1
Income = 51–75% ^b	-0.114	0.413	0.783	-10.8
Income = 76–100% ^b	-0.572	0.433	0.186	-43.6
Large sized properties ^c	-0.196	0.252	0.439	-17.8
Medium sized properties ^c	0.292	0.271	0.283	33.9
Gender	-0.404	0.243	0.096	-33.2
Age in 2015	-0.018	0.008	0.029	-1.9

Open in a new tab

^a Absentee resident is reference category

^b 0% of annual income derived from rural property is reference category

^c small property (40.5–202 ha, 100–500 acres) is reference category

Finally, we wanted to determine if woody cover preference was correlated with willingness to use prescribed fire as a relatively low cost, effective tool for reducing woody plants (Table 8) [9]. The regression analysis results found no link between cover preference and use of prescribed fire. Respondents from Oklahoma were about 61% more likely than Texas respondents to have used prescribed fire on their land indicating a stronger pro-fire culture in Oklahoma, and respondents in both states who owned large sized properties (>406 ha, 1000 acres), were also significantly more likely to use fire than landowners with smaller properties. Unsurprisingly, PBA members were significantly more likely (774%) to have used prescribed fire on their property than non-members. Finally, older landowners were significantly less likely to have used prescribed fire.

Table 8. Logistic regression results on factors influencing use of prescribed fire on private lands.

(Bolded results indicate P <0.05).

Independent Variables	Use of prescribed fire model: Pseudo R² = 0.209; P<0.001
	β coeff.	Std. err. β coeff.	p-value	% Δ odds
Woody cover preference	0.167	0.165	0.313	18.2
Lifestyle/recreation motivation	0.133	0.106	0.212	14.3
Rancher motivation	0.005	0.134	0.973	0.5
Heritage motivation	-0.035	0.109	0.752	-3.4
Hunter motivation	0.004	0.121	0.975	0.4
Farmer motivation	-0.032	0.123	0.796	-3.1
State	-0.963	0.278	0.001	-61.8
PBA member	2.169	0.238	<0.001	774.6
Income = 1–25% ^a	0.314	0.360	0.383	36.9
Income = 26–50% ^a	0.558	0.442	0.207	74.8
Income = 51–75% ^a	0.422	0.468	0.368	52.6
Income = 76–100% ^a	0.757	0.471	0.109	113.1
Large sized properties ^b	0.657	0.284	0.021	92.8
Medium sized properties ^b	0.195	0.315	0.536	21.6
Gender	0.153	0.270	0.569	16.6
Age in 2015	-0.025	0.009	0.008	-2.5

Open in a new tab

^a 0% of annual income derived from rural property is reference category

^b small property (40.5–202 ha, 100–500 acres) is reference category

Discussion

The survey participants in the Edwards Plateau and the Rolling Plains study areas in Texas and Oklahoma overwhelmingly reported increases in woody plants on their property. However, based on the additional remote sensing data from the sampled counties, only 9 of the 16 counties actually had overall woody plant increases, while six counties experienced decreases in overall woody plant cover during the 10 years preceding this survey.

The disparity between respondents’ perspectives about woody plant cover and the results of the land cover change analysis is striking. It is possible that individual landowners are experiencing woody plant increases but it is unlikely that significant woody plant expansion has occurred on most properties during that 10-year time period. There are several potential reasons behind the seeming disconnect between perception and documented land cover changes. The land cover change estimates may have been influenced by the effects of a severe drought between 2000 and 2011 in that dead standing trees, which were more abundant in the Edwards Plateau, may have been undetected from space whereas landowners could easily see them from ground level [19–21]. Landowners are also able to see small saplings that indicate woody plant increase but these may also not be detectable from space. Additionally, Roberts et al. (2018) discovered that nearly all policies aimed at containing juniper tend to promote rather than contain the spread of the species due to mismatches between policy and ecology and contradictory programs within natural resources agencies that simultaneously promote and control invasive species. [22] Therefore, reasons why ground level perceptions of woody plant expansion may be inconsistent with estimates derived from remotely-sensed data need to be explored in order to formulate meaningful and scientifically accurate policy guidelines for managing woody plant expansion effectively [4]. Developing simple tools that landowners can use to quantify woody plant cover on their property may also be useful for accurately monitoring changes over time. If landowners do not have an accurate representation of the extent of woody plant expansion on their land, they may not perceive woody plants as a problem.

Our study also found significant inter-state differences in woody cover preferences, with Texas respondents generally preferring higher cover (25%) compared to their Oklahoma counterparts (5%). In part, the preference for more woody cover among Texas respondents may be due to the marked increase in Central Texas of landowners who are moving away from traditional farming and ranching into recreational or wildlife-centric land uses [23,24]. By contrast, the lower woody cover preference in Oklahoma may be attributable to the continued importance of the ranching industry, particularly cattle ranching [25]. One question our research did not address is what type and patterns of woody plant cover do landowners prefer. Previous research found that many landowners within the Southern Plains demonstrate a preference for a moderately heterogeneous grassland landscape whereas those who were more production focused tended not to prefer highly heterogeneous landscapes [26]. This complexity of preferences among landowners may also extend to areas of woody plant encroachment or mixed grassland/woodland landscapes. For example, some landowners may want to maintain preferred tree species, such as mature oaks (Quercus sp.), while other may prefer trees that provide shade, cover, supplemental food or windbreaks for livestock or wildlife species [27,28].

Our results indicated that landownership motivations are significant drivers of woody cover preference, a finding corroborated by others [23,24]. Those who owned their property for recreation or hunting preferred more woody plant cover while more traditional livestock-oriented ranchers preferred less. This finding highlights the need for continuing research into changing land use trends. If regions experience a shift from larger livestock-oriented ranches to smaller properties that are owned primarily for lifestyle and recreational purposes, then woody cover expansion is likely to continue, in part, because the latter landowners tend to prefer more woody plants on their land [1,24]. This has serious ramifications for the maintenance or restoration of open grasslands and their associated biodiversity. It may also escalate wildfire risk in rural areas with increasing woody cover and development, which is typical of ever larger portions of the Edwards Plateau in Texas. Increasing risk of wildfires is of particular concern in areas where cedar and juniper species are dominant, as they are considered volatile fuel sources [4,29].

Given our finding that most of our respondents preferred lower woody plant density on their land than they currently have and the fact that prescribed fire has been shown to be an ecologically effective and economical efficient management tool to contain woody plant expansion [4,9], we anticipated there would be a negative correlation between preferred woody plant cover and use of prescribed fire. However, while our study did find that Oklahoma respondents who preferred lower woody plant cover appeared to be more accepting of prescribed fire than the Texas respondents, our study did not find any statistically significant correlation between woody cover preference and the use of prescribed fire on private lands. One reason for this is likely that many landowners still do not consider fire a safe and effective tool for managing rangelands [11]. Furthermore, encouraging the use of fire on private lands has proven to be complicated for a variety of reasons including landowners perceived risk from fire both as a physical hazard and from a legal liability standpoint [11,12,30,31] and a lack of knowledge or confidence in application of fire [6]. This suggests the urgent need for outreach programs aimed at educating landowners about the ecological and economic efficacy of periodic prescribed fire for containing and reducing woody plant cover and that also provide training and support in the use of prescribed fire to reduce concerns about the legal liability of using this management tool.

However, previous research has also shown that while educational programs can increase landowner’s knowledge and tolerance of prescribed fire programs, that may not translate into adoption of the practice [32]. By contrast, it has been found that PBA’s are very effective in engaging landowners in the use of prescribed fire, because they provide their members with fire safety training and equipment and they promote safe fire application experience among members who assist each other on burn days [4,33–35]. Other research has also supported the use of social networks in order to achieve land management objectives, especially in increasing landowners access to labor resources [36]. Given that our research also found that one of the differences between PBA members and non-PBA members was that PBA members were more likely to view woody plants on their land as a problem, PBA membership may be the catalyst for the greater use of prescribed fire to combat woody plant expansion. Therefore, in order to use prescribed fire to reduce woody fuel accumulations and thereby mitigate the increasing risk of wildlife, one important avenue for future research is how to best support and encourage the growth of PBA’s.

We also found age to be a significant factor both in woody cover preference and use of prescribed fire. Our average respondent age was 66 years old, highlighting the fact that many rural landowners tend to be older, which is a common finding in rural landowner surveys in our study area and throughout the US. Our results showed that older landowners may be slightly less likely to use prescribed fire but other research has indicated that the ages of the people who actually work on a farm or ranch is a better predictor of applied land management than the age of the primary decision maker [6,37–39].

While we feel that this study provides important preliminary evidence relating to landowners’ views about woody plant expansion and prescribed fire in the Southern Plains region of the United States, there are some limitations that should be addressed in future research. First, the purposive inclusion of all PBA members in our study area versus use of a random sample of non-member landowners may have produced biased results, a concern we attempted to remedy by comparing PBA member and non-member data relating to key study questions. While those tests found no difference in terms of perception or preference of woody plants on their land, PBA members were more likely to view woody plants as a problem and, not surprising, have a more positive view of prescribed fire. However, the directionality of this difference is unclear; i.e., whether PBA membership elevated the perception of woody plants as a problem or if landowners who viewed woody plants as a problem were more likely to become a PBA member in order to apply fire using PBA membership benefits including access to labor, fire management equipment and fire training. Future research should endeavor to identify changes in landowner perspectives about woody plant density and prescribed fire before and after joining a PBA. Second, we could not follow up with non-respondents and, therefore, were unable to rigorously ascertain if there was any non-response bias in our data. Others have recommended comparing data from early versus late respondents for testing non-response bias [14]. Our own comparative tests did not find any statistically significant differences between the two groups. Nevertheless, given that we found no evidence that this approach can accurately identify non-response bias, we opted to limit the discussion of our results to the respondents rather than extrapolate out findings to the general landowner population from which the sub-sample was drawn. Future research of landowner perspectives of woody plant densities and prescribed fire should endeavor to obtain a substantially higher survey response rates and some comparable data from non-respondents.

Conclusion

This study sought to address three research questions. 1) Do landowner perceptions about woody plant expansion correspond with estimated changes in their area? 2) Do landowners within the Southern Plains desire less woody cover on their land? 3) Do expressed woody cover preferences influence the adoption of prescribed fire on private land? Our results indicated that there was a disconnect between the perception of woody plant expansion on private lands and actual, local measurements of woody plant cover. Specifically, many landowners indicated that woody plants had increased on their own properties during the preceding 10 years whereas county level measurements showed little change in overall woody cover. Within the Texas and Oklahoma study areas, the survey respondents generally preferred about 25% less woody plant cover on their land than they estimated they have. Given the efficacy of prescribed fire as a woody plant management tool, we expected a positive correlation between degree of dissatisfaction with woody plant cover and willingness to use this management tool to combat woody plant expansion but this expectation was not confirmed in our study. These findings highlight the complexity of determining the relationship between landowner management goals, perceptions of woody plant cover and prescribed fire. This suggests the need for effective outreach programs to make the connection between woody plant reduction and prescribed fire for diverse and shifting types of landowners in the Southern Plains and to increase the number of prescribed burn associations across the ecoregion to encourage more landowners to use this management tool to restore grasslands and reduce wildfire risks on their land.

Supporting information

S1 File. Mail survey questionnaire.

(PDF)

Click here for additional data file.^{(584KB, pdf)}

S2 File. Anonymized data file.

(XLS)

Click here for additional data file.^{(289.5KB, xls)}

Acknowledgments

The authors want to thank all 1853 landowners who participated in this study.

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

UK received finding from the National Science Foundation Dynamics of Coupled Human and Natural Systems Program (Contact # DEB-1413900). The sponsors did not play a role in the study design, data collection and analysis, decision to publish or preparation of this manuscript. the funder URL can be found at: https://www.nsf.gov/pubs/2018/nsf18503/nsf18503.htm.

References

1.Berg MD, Sorice MG, Wilcox BP, Angerer JP, Rhodes EC, et al. (2015) Demographic Changes Drive Woody Plant Cover Trends—An Example from the Great Plains. Rangeland Ecology & Management 68: 315–321. [Google Scholar]
2.Harr RN, Morton LW, Rusk SR, Engle DM, Miller JR, et al. (2014) Landowners perceptions of risk in grassland management: woody plant encroachment and prescribed fire. Ecology and Society 19. [Google Scholar]
3.Huxman TE, Wilcox BP, Breshears DD, Scott RL, Snyder KA, et al. (2005) Ecohydrological implications of woody plant encroachment. Ecology 86: 308–319. [Google Scholar]
4.Twidwell D, Rogers WE, Fuhlendorf SD, Wonkka CL, Engle DM, et al. (2013) The rising Great Plains fire campaign: citizens' response to woody plant encroachment. Frontiers in Ecology and the Environment 11: e64–e71. [Google Scholar]
5.Archer S, Schimel DS, Holland EA (1995) Mechanisms of shrubland expansion: land use, climate or CO2? Climatic Change 29: 91–99. [Google Scholar]
6.Kreuter UP, Woodard JB, Taylor CA, Teague WR (2008) Perceptions of Texas landowners regarding fire and its use. Rangeland Ecology & Management 61: 456–464. [Google Scholar]
7.Lee AC, Conner JR, Mjelde JW, Richardson JW, Stuth JW (2001) Regional Cost Share Necessary for Rancher Participation in Brush Control. Journal of Agricultural and Resource Economics 26: 478–490. [Google Scholar]
8.Kahneman D (2011) Thinking, fast and slow 1st ed Kahneman Daniel. New York: Farrar, Straus and Giroux. [Google Scholar]
9.Van Liew D, Conner JR, Kreuter UP, Teague WR (2012) An economic assessment of prescribed extreme fire and alternative methods for managing invasive brush species in Texas: A modeling approach. The Open Agricultural Journal 6: 17–26. [Google Scholar]
10.Weir JR, Kreuter UP, Wonkka CL, Stroman DA, Russell M, et al. (2019) Liability and prescribed fire: Perception and reality. Rangeland Ecology & Management 72: 533–538. [Google Scholar]
11.Morton LW, Regen E, Engle DM, Miller JR, Harr RN (2010) Perceptions of Landowners Concerning Conservation, Grazing, Fire and Eastern Redcedar Management in Tallgrass Prairie. Rangeland Ecology & Management 63: 645–654. [Google Scholar]
12.Toledo D, Kreuter UP, Sorice MG, Taylor CA (2012) To burn or not to burn: Ecological restoration, liability concerns and the role of prescribed burning associations. Rangelands: 18–23. [Google Scholar]
13.Dillman DA (2000) Mail and internet surveys: The tailored design method. New York, NY: John Wiley and Sons. [Google Scholar]
14.Armstrong JS, Overton TS (1977) Estimating Nonresponse Bias in Mail Surveys. Journal of Marketing Research 14: 396–402. [Google Scholar]
15.Hartfield K, van Leeuwen W (2018) Woody Cover Estimates in Oklahoma and Texas Using a Multi-Sensor Calibration and Validation Approach. Remote Sensing 10: 632. [Google Scholar]
16.Schmitt N (1996) Uses and abuses of coefficient alpha. 8: 350–353. [Google Scholar]
17.Cortina JM (1993) What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology 78: 98–104. [Google Scholar]
18.StataCorp (2015) Stata Statistical Software: Release 14. 14.0 ed. College Station.
19.Twidwell D, Wonkka CL, Taylor CA Jr, Zou CB, Twidwell JJ, et al. (2014) Drought-induced woody plant mortality in an encroached semi-arid savanna depends on topoedaphic factors and land management. Applied Vegetation Science 17: 42–52. [Google Scholar]
20.Crouchet SE, Jensen J, Schwartz BF, Schwinning S (2019) Tree Mortality After a Hot Drought: Distinguishing Density-Dependent and -Independent Drivers and Why It Matters. Frontiers in Forests and Global Change 2. [Google Scholar]
21.Moore GW, Edgar CB, Vogel JG, Washington-Allen RA, March Rosaleen G, et al. (2016) Tree mortality from an exceptional drought spanning mesic to semiarid ecoregions. Ecological Applications 26: 602–611. 10.1890/15-0330 [DOI] [PubMed] [Google Scholar]
22.Roberts CP, Uden DR, Allen CR, Twidwell D (2018) Doublethink and scale mismatch polarize policies for an invasive tree. PLoS ONE 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Hurst KF, Ramsdell CP, Sorice MG (2017) A life course approach to understanding social drivers of rangeland conversion. Ecology and Society 22. [Google Scholar]
24.Sorice MG, Kreuter UP, Wilcox BP, Fox Iii WE (2014) Changing landowners, changing ecosystem? Land-ownership motivations as drivers of land management practices. Journal of Environmental Management 133: 144–152. 10.1016/j.jenvman.2013.11.029 [DOI] [PubMed] [Google Scholar]
25.Bertone R (2019) Oklahoma's top 10 agricultural products. Farm flavor. [Google Scholar]
26.Becerra TA, Engle DM, Elmore RD, Fuhlendorf SD (2013) Contrasting Preference for Grassland Landscapes Among Population Groups in the Central and Southern Great Plains. Rangeland Ecology & Management 66: 529–538. [Google Scholar]
27.Juzwik J, Appel DN, MacDonald WL, Burks S (2011) Challenges and Successes in Managing Oak Wilt in the United States. Plant disease 2011 v.95 no.8: pp. 888–900. 10.1094/PDIS-12-10-0944 [DOI] [PubMed] [Google Scholar]
28.Olenick KL, Kreuter UP, Conner JR (2005) Texas landowner perceptions regarding ecosystem services and cost-sharing land management programs. Ecological Economics 53: 247–260. [Google Scholar]
29.Brockway DG, Outcalt KW, Estes BL, Rummer RB (2009) Vegetation response to midstorey mulching and prescribed burning for wildfire hazard reduction and longleaf pine (Pinus palustris Mill.) ecosystem restoration. Forestry: An International Journal of Forest Research 82: 299–314. [Google Scholar]
30.Wonkka CL, Rogers WE, Kreuter UP (2015) Legal barriers to effective ecosystem management: exploring linkages between liability, regulations, and prescribed fire. Ecological Applications 25: 2382–2393. 10.1890/14-1791.1 [DOI] [PubMed] [Google Scholar]
31.Eburn M, Cary GJ (2017) You own the fuel, but who owns the fire? International Journal of Wildland Fire 26: 999–1008. [Google Scholar]
32.Loomis JB, Bair LS, González-Cabán A (2001) Prescribed Fire and Public Support: Knowledge Gained, Attitudes Changed in Florida. Journal of Forestry 99: 18–22. [Google Scholar]
33.Toledo D, Kreuter UP, Sorice MG, Taylor CA Jr (2014) The role of prescribed burn associations in the application of prescribed fires in rangeland ecosystems. Journal of Environmental Management 132: 323–328. 10.1016/j.jenvman.2013.11.014 [DOI] [PubMed] [Google Scholar]
34.Weir JR, Twidwell D, Wonkka CL (2016) From Grassroots to National Alliance: The Emerging Trajectory for Landowner Prescribed Burn Associations. Rangelands. [Google Scholar]
35.Kreuter UP, Stroman DA, Wonkka CL, Weir J, Abney AA, et al. (2019) Landowner Perceptions of Legal Liability for Using Prescribed Fire in the Southern Plains, United States. Rangeland Ecology & Management 72: 959–967. [Google Scholar]
36.Sutherland L-A, Burton RJF (2011) Good Farmers, Good Neighbours? The Role of Cultural Capital in Social Capital Development in a Scottish Farming Community. Sociologia Ruralis 51: 238–255. [Google Scholar]
37.Brunson MW, Huntsinger L (2008) Ranching as a Conservation Strategy: Can Old Ranchers Save the New West? Rangeland Ecology & Management 61: 137–147. [Google Scholar]
38.Sorice MG, Kreuter UP, Wilcox BP, Fox WE III (2012) Classifying land-ownership motivations in central Texas, USA: A first step in understanding drivers of large-scale land cover change. Journal of Arid Environments 80: 56–64. [Google Scholar]
39.Jackson-Smith D, Kreuter UP, Krannich RS (2005) Understanding the multidimensionality of property rights orientations: Evidence from Utah and Texas ranchers. Society and Natural Resources 18: 587–610. [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0238688.r001

Decision Letter 0

Francisco X Aguilar

24 Oct 2019

PONE-D-19-20616

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

PLOS ONE

Dear Dr. Stroman,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

ACADEMIC EDITOR: The authors address a topic of potential interest not just in the US but in other regions around the world facing similar challenges and opportunities. Reviewers have raised various issues that shall be properly addressed to merit the publication of this manuscript.

==============================

We would appreciate receiving your revised manuscript by Dec 08 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Francisco X Aguilar

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy as Supporting Information.

We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

We note that Figures 1 in your submission contain [map/satellite] images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a) You may seek permission from the original copyright holder of Figure(s) [#] to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b) If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Review of manuscript PONE-D-19-20616 – Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

Summary

This study reviewed attitudes of landowners in two geographic regions of western Oklahoma and south-central Texas for comparing perceived and preferred woody plant cover and how these preferences relate to use of prescribed fire. Respondents indicated that they had more woody plant cover on their property than desired. Recreational and hunter landowner categories were more likely to prefer woody cover, while the rancher category was less likely, indicating that primary ownership motivations can influence woody cover preference.

Attitudinal research such as this is a beneficial tool for revealing regional motivations of private landowners and can help guide future management/conservation approaches while furthering the institutional knowledge of the subject. The authors used good survey methodology, including an easy to understand visual graphic for estimation of woody cover. The use of various forms of data analyses helped to illustrate and support findings, and there was good categorical class delineation for the latent variables of primary ownership motivators, including testing for internal reliability. The authors also did a good job of providing relevance for the study by discussing relationship to a larger perspective.

However, there were noticeable areas for improvement before the acceptance of the manuscript. This includes need for discussion about how potential sources of bias were handled, manuscript organization, and a focus on incorporating research data into the Discussion and Conclusion section. Several minor issues will also need to be addressed.

Examples/Evidence

Major Issues

1) Discuss how potential bias was handled

a. Survey recipients included 100 random landowners from each county in the study area, plus PBA members. It seems that including a specific subset of the sample population would potentially influence findings and make it difficult to draw conclusions, especially since PBA members would already understand the value of prescribed fire and therefore are potentially more aware of the presence of woody plants on their property. How was this potential source of bias accounted for during the analyses?

b. Were there any other sources of bias tested for? (e.g. violations of assumptions or outliers)

c. How were unanswered questions handled? (i.e. was there any influence created by non-responses to specific items within questionnaire areas of inquiry)

2) Manuscript organization

a. Information in the Results section should only report results; discussion should be located in the Discussion section (line 240-243)

b. Data should be reported in the Results section (Table 7)

3) Results relationship to Discussion and Conclusions

a. It is important to use the Discussion and Conclusion section to show how the study complements the existing literature and answers the research questions. The authors do a good job of touching on highlights, but more of this study’s research findings should be listed, as well as referencing tables and figures from the Results sections.

Minor Issues

4) Data/analyses transparency

a. What software package was used for data analyses?

b. Insert table to show questionnaire items that composed attitudes concerning woody plants and fire. This would provide context on the number and type of questions defining each construct and would be an opportunity to indicate response %. Also indicate which items were positively and negatively worded for purpose of reverse-scoring in the data analyses.

c. Table 2

i. What type of rotation was used?

ii. List rotated factor loading results for all items

iii. Include % of Variance as a separate row

d. Table 5 and 6

i. Include standard errors for β coeff.

ii. Consider reporting odds ratio and confidence interval instead of % Δ odds, as this will help with interpretation of the results

5) Include supporting references for statements

a. line 284

b. line 296-7 – in that area? statewide?

c. line 315

d. line 241-3 – include reference; how does this relate to Table 4?

e. line 315-6

6) Clarity of statements

a. line 250, 254 – be careful of interpreting % Δ odds as probability; reporting the odds ratio will allow for more accurate interpretation of data

b. line 282-3 – what about the possible influence of PBA member bias?

c. line 286 – expand on this thought

d. line 303-5 – how does this species relate to the study, or include reference for why it should be specifically mentioned

e. line 314 – discuss how/why would it escalate wildfire risk

f. line 344 – what preceding questions? If they are the research questions then they should be restated

7) Completeness of reporting information

a. Be sure to discuss all statistically significant findings – ex. age relationship and woody cover preference (Table 5) – all other significant findings were mentioned, including age for Table 6

8) Proofreading

a. Several typos and grammatical errors exist and need corrected

Reviewer #2: Well conceived and well written paper addressing a topical public policy issue focusing on social drivers of woody cover change in the US Great Plains. The strength of the analysis rests on the stratified random selection of respondents. However, in lines143-148 you note that you do not conduct an assessment of cover change on respondent's properties--a weakness--but you adequately attempt to qualify your results / observations. L120 has an extra 'to', and presumably L129 is missing a 'not'

Reviewer #3: The authors conducted a mail survey in sixteen counties in two states to determine (1) how landowners perceived the current woody cover of their land and whether it had changed over the last 10 years, (2) What their desired woody cover was and how this compared to their perception of actual woody cover, and (3) whether an expressed preference for lower woody cover was linked to a land owner being more likely to have used fire in the past. The responses were analysed to determine whether factors such as landowner demographic characteristics, type of land use, property size and membership of a prescribed burning association (PBA) influenced woody cover preference and use of prescribed burning. The authors make the case that woody cover increase is a significant ecological and economic problem in the study area, and that knowing about different woody cover preferences and factors that influence adoption of prescribed burning will help target appropriate interventions to improve the use of fire to control woody encroachment.

It seems odd to conduct this research on issues surrounding woody cover increase in counties that actually have had decreasing woody cover (Table 7). Was there any baseline data on woody cover and woody cover change in the study area that informed the study and choice of study sites?

The study appears to be largely methodologically sound, but I have three main concerns that need to be addressed. First, the description of the survey and the questionnaire is insufficient to judge the study design and to allow for repeatability. Second, the possible biases introduced the survey methods are incompletely acknowledged and should be dealt with more explicitly. Third, the inclusion of data from a remote sensing study to contextualize the perceptions expressed by landowners in the survey is a good idea, but it was poorly done.

1. Incomplete information on the questionnaire

It would have been useful to have access to the whole (blank) questionnaire. This would allow readers to assess how the questions were phrased, and whether some information that would have been useful to have (e.g. reasons given for using/not using prescribed burning) was collected but not reported, or whether it was not collected. It would also allow for repeatability of the work.

The methods sections states that “we asked the survey participants a series of questions as part of a broader mail survey about woody plants and prescribed fire” (lines 107-108) and the questionnaire was apparently 12 pp long (line 116). This suggests that there was material covered in the questionnaire that is not reported in this manuscript, but which may be useful to include to aid interpretation. Are there substantial aspects of the questionnaire that are not reported in this manuscript? If so, what are the reasons for this? I feel that I am getting an incomplete picture of the questionnaire and its overall aims. Are they the same as reported in this paper or is the work presented here just a subset?

The questionnaire asked respondents to estimate the current woody cover on their land and to indicate their preferred amount of woody cover. Farmers were also asked whether they had perceived an increase in woody cover. It would have been useful to have asked formers to estimate the woody cover on their farms at some earlier data using the sample schematics used to elicit perceptions of present and desirable woody cover.

The authors state that the survey data cannot be freely shared because of ethics/confidentiality concerns, but this restriction should not apply to the actual questionnaire that was used.

2. Potential sampling bias resulting from the study design

The authors simply state that “A non-response bias analysis was not conducted because alternative contact information (e.g., telephone number) required for such an analysis was not obtainable. Due to this limitation, we do extrapolate our results to all landowners in the Southern Plains.” (lines 127-129). This sounds a bit like "we couldn't do it, therefore it was not necessary", but the possible response biases (e.g. age) and their likely an influence on the findings needs to be properly acknowledged and dealt with. The statistical analyses used assume that the data represent a random sample from the population, and if this assumption is violated then the inferences drawn from the sample cannot be applied to the population.

I identified two (possible) sources of sampling bias that are likely to have influenced the results of the analyses and these need to be better discussed and taken into account when drawing conclusions from the research.

The first is that the overall sample was chosen as a random subset of 100 landowners in a county, but to this the authors added 318 purposively chosen members of PBAs (making up 16.6 % of the total questionnaires mailed and > 30 % of the responses received). The authors do not tell the reader what percentage the 100 randomly selected landowners per county represents of the total eligible population (landowners with a minimum 100 acre property size). Similarly, it is not clear what percentage of PBA members is represented by the 318 PBA members that were contacted. Including the PBA members as part of the otherwise random sample violates the assumptions of the regression analyses that used to analyse the influence of various demographic factors on woody cover preference and use of fire. The magnitude of the possible bias this introduces is impossible to estimate since no information is given on the overall ratio of PBA members to the general population, and how this compares to the ratio of respondents in the data (> 30 % of the survey respondents were PBA members who were purposively selected as survey participants).

Was there any exploration of what led PBA members to become members of PBAs? One of the study’s findings if that PBA members were more likely to have used fire in the past, but what can one infer about cause and effect? Would landowners given to using fire not be more likely to join a PBA in the first place? Or does joining a PBA result in landowners being more comfortable using fire?

The second possible source of bias is the use of a mail survey. I don’t know what the current situation is in the rural areas of Texas and Oklahoma, but in many parts of the world few people read and respond to letters via “snail” mail. There are likely to be valid reasons for choosing this method, but they need to be justified and explained, and possible biases that may have arisen need to be discussed, ideally with reference to the international literature. I would expect, for example, that older landowners would be more likely to respond to a mail survey. The sample is very strongly skewed towards older respondents but it is unclear whether this reflects the demographics of landowners in the area or whether this is a respondent bias part of the survey. If the latter then the survey results may under-represent the perceptions, preferences and practices of younger farmers who are significantly different in these respects. PBA members made up a third of respondents, what is their age profile compared to that of the non-PBA members? At the very least, the authors should make an effort to compare this to the actual demographics of the study area, which should be available from population census or similar data, to see if a significant age bias exists.

Both these sources of (possible) bias have a bearing on the results reported. Age was a significant predictor of woody cover preference, and both age and PBA membership were found to significantly predict use of prescribed burning. Without consideration (and where possible, elimination) of sampling bias, these results cannot be considered to be robust.

3. Inclusion of woody cover trend data from Hartfield & van Leeuwen (2018)

The obvious gap when trying to determine whether farmer’s perception of current woody cover and past change are accurate is some objective measure of actual cover. The authors address this by using data from a remote sensing study (Hartfield & van Leeuwen 2018) to compare perceptions of change to published rates of change on a per-county basis. While this is a good idea, there are several problems with how this was done and reported.

The first issue concerns the scale mismatch between the perceptions at the individual landowner level and the change data at the level of the county. The authors explain that this was unavoidable, but the implications are not adequately explored. No data of actual woody cover in each county, and some indication of how variable it is in space, are presented. If woody cover in a county varies a lot spatially, then linking it to individual landowner responses will be a bit of a stretch.

The survey was done in 2015, but the study that is cited as the source of the actual change data was published a couple of years later in 2018. I assume that the remote sensing study therefore did not inform the choice of counties for conducting the survey, but that the data were later accessed to compare perceptions of change to actual change. This is a pity, since the motivation for the survey was that woody encroachment is a significant issue, whereas the remote sensing study found that many counties did not actually experience an increase in woody cover. Possible reasons for the discrepancy between landowners’ perceptions are discussed, but in a rather speculative fashion.

I was unable to ascertain where the authors obtained the data shown in Table 7. No source is given in the table caption, but the source is indicated in the text to be Hartfield & van Leeuwen (2018). I looked at the Hartfield & van Leeuwen (2018) paper and the data do not appear in it. Instead, Hartfield and van Leeuwen state in their discussion that “In a future paper, we will use estimates of woody cover from two dates, to compute the change in woody cover for each pixel in the two ecoregions.” So where do the data cited in the submitted manuscript come from?

Since the authors did manage to access woody change cover for the counties in which their survey was done (and will hopefully be able to provide a correct citation), they should also have been able to source and report woody cover in 2004 and 2014, the latter being the year before the survey. This would allow not only comparison of woody cover estimates and actual woody cover, but also whether actual woody cover in a county influenced perceptions and use of prescribed burning.

Lastly, the perceptions of trends and the comparison with published woody cover trends should have been reported in the results section. Instead, they come as something of a surprise in the form of a table and accompanying text at the beginning of the discussion section. The discussion should not introduce new findings but needs to interpret the findings reported in the results section.

Other comments

The introduction reads as if the paper is aimed at a USA-based rangeland science and management audience. For publication in an international journal, the introduction is rather narrowly focused on a local problem. As someone not based in the USA (but interested in woody encroachment and rangeland management), I would have liked to see more international literature to contextualize the situation in Great Plains. The discussion should similarly draw conclusions that go beyond the regional.

Line 32: Were landowners just asked about the “amount” or also about perceived trends?

Lines 36 and 82: “lability” should be “liability”

Line 39: Does membership of PBA cause landowners to be more likely to use fire, or do landowners who are comfortable using fire join PBAs more readily?

Lines 44-46: This final sentence of the abstract is rather vague.

Line 50: The values of woody cover increase given here contrast with the declines in woody cover reported in Table 7.

Lines 55-56: Use the singular (“climate change” and “land use”)?

Lines 59-61: Why were PBA members explicitly included in the survey while membership of FMAs and the other organisations cited was seemingly not considered or explored at all? Also, since FMAs and NRCS are never mentioned again, there is no need to introduce the abbreviations here.

Line 71: The argument about under- or overestimating woody cover is valid but in the absence of actual woody cover data, the study does not actually determine whether farmers under- or over-estimated woody cove on their land.

Lines 84-85: The first question is about landowner perceptions about woody plant expansion and how it correlates with actual trends but much of the focus of the study (and the reported analyses) is farmer’s estimation of current woody cover (rather than trend). “Trend” and “current cover” are not always clearly distinguished in the writing.

Lines 96 and 99: Change “included” to “were”. “Included” implies that the subsequent list is a subset, when in fact all counties sampled are listed.

Lines 116-117: What were the “five areas of enquiry including attitudes concerning woody plants and fire” and are they all reported on in this manuscript? If not, why not?

Line 119 and elsewhere: Journal requirements include the use of SI units. Acres are not SI units. The values given need to be converted to hectares, or at least report equivalent values in SI units for the benefit of readers accustomed to the metric system.

Line 119: What proportion of landowners with properties > 100 acres does the sample of 1600 represent?

Line 120: What is the basis for the 100-acre minimum? Can the choice of this minimum value, and the assumption that owners of smaller properties would be less likely to use fire, somehow be substantiated (e.g. with a reference from the literature)?

Line 123: “Pulled from” seems rather colloquial. “Extracted from”?

Lines 124-125 – what percentage of PBA members do the 126 / 192 members from TX /OK represent?

Line 131: “Participants” needs an apostrophe.

Line 132: Close bracket after “trees”.

Line 143: What this a “yes/no” type of question?

Line 178: Should “assumption” be replaced with “prediction” or “hypothesis”?

Table 3, last row: OK is abbreviated here but this is the only instance – keep use of appreviations consistent or avoid them.

Lines 198-199 – how do the sample demographics compare to the general population? Also, all results should be consistently written in the past tense.

Line 204: But this figure is influenced by the fact that PBA members were actively targeted. What is the rate of PBA membership actually like among landowners?

Table 4. It would be useful to also include what percentage of respondents fall within the "ranching", "heritage", "hunting" and "farmer" land use categories.

Paragraph following Line 207: It would be useful to include the published data on woody cover and trends in each county here and analyse the differences by county.

Line 240: "Reported" is ambiguous – does this refer to woody cover reported by respondents or in the literature? Would “perceived” not be a more appropriate word here than “reported”?

Tables 5 and 6: p-values are not 0.000 – they are greater than zero and should be presented as < 0.001.

Line 248: “low-cost, effective”: Is that how it was put in the questionnaire? If not, I would leave these value-laden terms out of the results section.

Line 250: delete “By contrast”.

Lines 255-256: Rather state this more intuitively, e.g. "older land owners were significantly less likely to have used fire...". The sentence needs a full stop.

Lines 261 onwards: This needs to be reported in the results section, not the discussion.

Table 7 needs to cite the source of the woody cover change data. Also include actual woody cover in 2004 and 2014 for each county, as these values are of interest. For example, a small decline in woody cover might be outweighed in people’s perceptions by overall high woody cover values.

Line 279: Any data / evidence / literature to support this?

Line 283: farmers may be frustrated by the woody cover, but the reader is frustrated by the absence of data on actual woody cover.

Lines 285-287: What does that mean? Not very enlightening when put like this unless one goes and reads the paper (beyond its title). Not only is it unclear what argument the cited paper makes, but there is also no explanation of how/why this would influence perceptions of actual woody cover.

Lines 295 and 297: It would be useful here to know what the percentage of landowners of the rancher, farmer, heritage, hunter types was in in each state.

Line 298: Replace the question mark with a full stop.

Line 322: From what I can ascertain, actual willingness was not measured – “used fire more frequently” would seem more accurate here than “more willing”.

Line 235: Did the questionnaire not probe landowners’ reasons for using / not using fire?

Line 235: What is meant by “legitimate” in this context?

Line 332: How about training and support in the use of fire?

Figure 1: For the benefit of an international audience, please label the states of Texas and Oklahoma.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Sep 8;15(9):e0238688. doi: 10.1371/journal.pone.0238688.r002

Author response to Decision Letter 0

5 Feb 2020

Reviewer #1:

R1.1: Discuss how potential bias was handled

Response R1.1a: Using PBA membership lists, we intentionally included all PBA members in each study county because if we relied on a randomly drawn PBA membership sample drawn from the general tax records, the chance of selecting an adequately large sample of PBA members for comparison with non-member respondents would have been extremely unlikely. One of the objectives of our study was to understand the effect of PBA membership on perspectives about woody plant encroachment and the use of fire. It is unclear how PBA membership would have enhanced awareness of the presence of woody plants on their property; woody plants are equally visible to both members and non-members. It is indeed likely that PBA members have a better understanding of the use of prescribed fire; confirming this with our data is the basis for arguing that greater landowner access to PBA membership could enhance the broader use of this critically needed rangeland management tool. Previous research has also shown that many PBA members were not always supportive of the use of prescribed fire but rather that joining such a landowner organization enabled them to change their mind about prescribed fire being a risky land management tool and, consequently, to decide to apply fire on their own land (Kreuter et al. 2008). Finally, when we report and discuss our findings, we are careful to discuss only in terms of our respondents in order not to extrapolate any possible response bias to a larger population.

b. Were any other sources of bias tested for? (e.g. violations of assumptions or outliers)

Response R1.1b: We did not test for other potential sources of response bias. Since our data were derived primarily from specified response scales, outliers are not present. Additionally, as indicated above, to avoid extrapolating potential bias, we refer to respondents only, not landowners in general, in our presentation and discussion of results and not extrapolated to the landowner population from which the sample was drawn.

c. How were unanswered questions handled? (i.e. was there any influence created by non-responses to specific items within questionnaire areas of inquiry)

Response R1.1c:

We used null values for unanswered questions, resulting in some variation of n values for each variable. We based this on the assumption that data were missing completely at random (MCAR) as we could not determine respondent’s reasons for not answering any particular question. Based on this assumption, we did not impute MCAR data.

R1.2: Manuscript organization

a. Information in the Results section should only report results; discussion should be located in the Discussion section (line 240-243)

Response R1.2a: All discussion points originally included in the results section were moved to the discussion.

b. Data should be reported in the Results section (Table 7)

Response R1.2b: Table 7 (now Table 5) and the associated text have been moved for the results section.

R1.3: Results relationship to Discussion and Conclusions

a. It is important to use the Discussion and Conclusion section to show how the study complements the existing literature and answers the research questions. More of this study’s research findings should be listed, as well as referencing tables and figures from the Results sections.

Response R1.23a: We have expanded the discussion section to address additional significant findings (ex: age variable influence on woody cover preference and use of fire) and have connected our findings back to the literature more thoroughly, including more international literature.

R1.4: Data/analyses transparency

a. What software package was used for data analyses?

Response R1.4a: We used STATA 13 and have included that information in the Methodology section and the works cited

Response R1.4b: We have added an abbreviated questionnaire to the supplemental information so that readers can read the questions asked of participants. We only developed latent index variables for ascertaining respondent’s landownership motivations as outlined in Table 2. The question concerning woody cover preference is detailed in Table 1 under the dependent variable label and again in Figure 2 where readers can see the question prompt as it appears in the questionnaire instrument.

c. Table 2: What type of rotation was used? List rotated factor loading results for all items. Include % of Variance as a separate row.

Response R1.4c: Requested changes have been included in Table 2.

d. Table 5 and 6: Include standard errors for β coeff. Consider reporting odds ratio and confidence interval instead of % Δ odds, as this will help with interpretation of the results.

Response R1.4d: SE of β coeff. Have been included as suggested. We believe that it is simply a matter of interpretation wording preference since % change in odds is simply subtracting the odds ratio from 1 and multiplying the difference by 100. Either method (odds ratio or % change in odds) will provide the same information. Our preferred standard reporting method is % change in odds.

R1.5: Include supporting references

Line 284; line 296-7 (in that area – statewide?); line 315; line 241-3 (how does this relate to Table 4?); line 315-6

Response R1.5: Line 284 has been deleted from the manuscript per other reviewer comments. Line 296-7 reference has been added. Line 315 references have been added. Line 241-243 has been removed from the results section and moved to the discussion section and references have been added. Line 315-316 references concerning increasing wildfire risk have been added.

R1.6: Clarity of statements

a. line 250, 254 – be careful of interpreting % Δ odds as probability; reporting the odds ratio will allow for more accurate interpretation of data

Response R1.6a: Please see response R1.4d above

b. line 282-3 – what about the possible influence of PBA member bias?

Response R1.6b: Please see R1.1a above.

c. line 286 – expand on this thought

Response R1.6c: We have deleted the sentence and the preceding one as these are perhaps too speculative.

d. line 303-5 – how does this species relate to the study, or include reference for why it should be specifically mentioned

Response R1.6d: We have added references concerning landowner preference for retaining oak trees on their land (Olenick et al 2005) (Juzwick et al 2011)

e. line 314 – discuss how/why would it escalate wildfire risk

Response R1.6e: We have added language clarifying why cedar/juniper expansion increases risk of wildfire and added 2 references as support. (Brockway et al 2009 and Twidwell et al 2013)

f. line 344 – what preceding questions?

Response R1.6f: That sentence has been reworded for clarity

R1.7: Completeness of reporting information

Discuss all statistically significant findings – ex. age relationship and woody cover preference (Table 5) – all other significant findings were mentioned, including age for Table 6

Response R1.7: We have added language to the results section about the significant age variable in Table 5 (now Table 6)

R1.8: Proofreading

Several typos and grammatical errors exist and need corrected

Response R1.8: We have carefully proof read the revised manuscript to ensure correct grammar and eliminate typos.

Reviewer #2:

R2.1: In lines143-148 you note that you do not conduct an assessment of cover change on respondent's properties--a weakness--but you adequately attempt to qualify your results / observations.

Response R2.1: Since there was no recommendation for change, we did not change the text.

R2.2: L120 has an extra 'to', and presumably L129 is missing a 'not'

Response R2.2: Corrections included.

Reviewer #3:

R3.1. It seems odd to conduct this research on issues surrounding woody cover increase in counties that actually have had decreasing woody cover (Table 7). Was there any baseline data on woody cover and woody cover change in the study area that informed the study and choice of study sites?

Response R3.1. The spatial analysis of woody plant cover changes in the Southern Plains that concluded some counties experienced woody plant reduction was being conducted at the same time as the landowner survey. These assessments were two parts of a larger research project. Generally, woody plant expansion is thought to have occurred throughout the Southern Plains. We included explanations for the disparity: “The land cover change estimates may have been influenced by the effects of a severe drought between 2010 and 2015 in that dead standing trees, which were more abundant in the Edwards Platea, may have been undetected from space whereas landowners could easily see them from ground level. Additionally, landowners are also able to see small saplings that indicate woody plant increase but these may also not be detectable from space. Conversely, some respondents may have erroneously indicated woody plant expansion because of their growing frustration with the higher than desired woody plant concentration on their land. This may be especially the case on the counties sampled in Texas which have more woody plant cover than those in Oklahoma. Additionally, Roberts et al. (2018) discovered that nearly all policies aimed at containing juniper tend to promote rather than contain the spread of the species due to “doublethink and policy-ecology mismatches.”(17)”

R3.2: Incomplete information on the questionnaire

a. It would have been useful to have access to the whole (blank) questionnaire. This would allow readers to assess how the questions were phrased, and whether some information that would have been useful to have (e.g. reasons given for using/not using prescribed burning) was collected but not reported, or whether it was not collected. It would also allow for repeatability of the work.

Response R3.2a. We are including an abbreviated copy of the questionnaire in the supplemental information (S1).

b. Are there substantial aspects of the questionnaire that are not reported in this manuscript? If so, what are the reasons for this? I feel that I am getting an incomplete picture of the questionnaire and its overall aims. Are they the same as reported in this paper or is the work presented here just a subset?

Response R3.2b. Please see the previous response. Not all of the information from a multi- questionnaire aimed at addressing multiple research question can be presented in one journal manuscript. We provide the reference for one other publication so far above and cite it in the revised manuscript. as it includes aspects that have been published elsewhere. See: Kreuter, U.P., D.A. Stroman, C. Wonkka, J. Weir, A.A. Abney, J.K. Hoffman. 2019. Landowner perceptions of legal liability for using prescribed fire in the Southern Plains, USA. Rangeland Ecology & Management. 72: 959-967. This very recent publication has been added in the discussion of liability concerns.

c. It would have been useful to have asked farmers to estimate the woody cover on their farms at some earlier date using the sample schematics used to elicit perceptions of present and desirable woody cover.

Response R3.2b. We did ask survey participants to respond to the following three statements, although they were not asked to refer to the schematic in selecting their response option, which could have been beneficial: Cedar/juniper has increased on my property in the last ten years; Mesquite has increased on my property in the last ten years; Other woody plants have increased on my property in the last ten years. We report this information in the revised manuscript.

d. The authors state that the survey data cannot be freely shared because of ethics/confidentiality concerns, but this restriction should not apply to the actual questionnaire that was used.

Response R3.2d. Please see Response 3.2a. and 3.2b.

R3.3: Potential sampling bias resulting from the study design

a. The authors simply state that “A non-response bias analysis was not conducted because alternative contact information (e.g., telephone number) required for such an analysis was not obtainable. Due to this limitation, we do extrapolate our results to all landowners in the Southern Plains.” (lines 127-129). This sounds a bit like "we couldn't do it, therefore it was not necessary", but the possible response biases (e.g. age) and their likely an influence on the findings needs to be properly acknowledged and dealt with. The statistical analyses used assume that the data represent a random sample from the population, and if this assumption is violated then the inferences drawn from the sample cannot be applied to the population.

Response R3.3a. See response in R3.3c. We also changed the wording in line 127-129 to read “Due to this limitation, we do not extrapolate our results to all landowners in the Southern Plains.”

b. I identified two (possible) sources of sampling bias that are likely to have influenced the results of the analyses and these need to be better discussed and taken into account when drawing conclusions from the research.

Response R3.3b. See response in R3.3c

c. The first is that the overall sample was chosen as a random subset of 100 landowners in a county, but to this the authors added 318 purposively chosen members of PBAs (making up 16.6 % of the total questionnaires mailed and > 30 % of the responses received). The authors do not tell the reader what percentage the 100 randomly selected landowners per county represents of the total eligible population (landowners with a minimum 100 acre property size). Similarly, it is not clear what percentage of PBA members is represented by the 318 PBA members that were contacted. Including the PBA members as part of the otherwise random sample violates the assumptions of the regression analyses that used to analyze the influence of various demographic factors on woody cover preference and use of fire. The magnitude of the possible bias this introduces is impossible to estimate since no information is given on the overall ratio of PBA members to the general population, and how this compares to the ratio of respondents in the data (> 30 % of the survey respondents were PBA members who were purposively selected as survey participants).

Response R3.3c.

First, it is impossible to report the total eligible population of landowners in a county who own 100 acres of land or more because many landowners own multiple plats that equal or are greater than 100 acres. We compiled a list of all landowners, according to county tax records, who met the minimum acreage threshold, eliminated any duplicate names and pulled our sample from there. Second, had we relied only on the county tax records to randomly draw landowners who are PBA members, we would have obtained a very small sample of such landowners. Since we were interested in exploring differences in prescribed fire perspectives between PBA member and non-member landowners, we needed to ensure an adequate number of PBA member landowners in our study. Therefore, whereas the non-member landowners represent a random sample of landowners who own at least one plat of land = > 100 acres, the survey participants who were PBA members represent the entire population of PBA members who reside in the counties we surveyed.

d. Was there any exploration of what led PBA members to become members of PBAs? One of the study’s findings if that PBA members were more likely to have used fire in the past, but what can one infer about cause and effect? Would landowners given to using fire not be more likely to join a PBA in the first place? Or does joining a PBA result in landowners being more comfortable using fire?

Response R3.3d. In this study, we did not inquire about motivations for becoming a PBA member. Previous research indicates that membership motivations vary including safe fire training, access to fire management equipment on burn days, assistance with application of prescribed fire, and access to prescribed fire liability insurance. However, we did ask some questions about how the PBA influences their social networks. While we do not know from this study if joining a PBA resulted in more landowners being more comfortable using fire, previous research (Kreuter et al. 2008, Toledo et al. 2014) as well as this study show there is strong correlation between PBA membership and application of fire on the members own property and assistance with the application of prescribed fire on other people’s properties.

e. The second possible source of bias is the use of a mail survey. I don’t know what the current situation is in the rural areas of Texas and Oklahoma, but in many parts of the world few people read and respond to letters via “snail” mail. There are likely to be valid reasons for choosing this method, but they need to be justified and explained, and possible biases that may have arisen need to be discussed, ideally with reference to the international literature. I would expect, for example, that older landowners would be more likely to respond to a mail survey. The sample is very strongly skewed towards older respondents but it is unclear whether this reflects the demographics of landowners in the area or whether this is a respondent bias part of the survey. If the latter then the survey results may under-represent the perceptions, preferences and practices of younger farmers who are significantly different in these respects. PBA members made up a third of respondents, what is their age profile compared to that of the non-PBA members? At the very least, the authors should make an effort to compare this to the actual demographics of the study area, which should be available from population census or similar data, to see if a significant age bias exists.

Response R3.3e. We have added information about age to the discussion, including several references. The reported average age (66) is in line with several similar studies both within our study area and throughout the U.S. (see references in discussion section). Our age range of respondents was between 30-93, which indicates that the use of a mail survey is still an accessible method for reaching private landowners of all ages. County land ownership databases in our study area do not include information such as phone or email contacts that may allow for other survey methods. In addition, older landowners may be less likely to answer long form surveys by phone or email. Moreover, we have conducted mail surveys for over 20 years with consistently satisfactory response rates ranging from 30-50% whereas the email/web-based surveys that we have conducted have resulted in extremely low response rates of 5% or less.

f. Both these sources of (possible) bias have a bearing on the results reported. Age was a significant predictor of woody cover preference, and both age and PBA membership were found to significantly predict use of prescribed burning. Without consideration (and where possible, elimination) of sampling bias, these results cannot be considered to be robust.

Response R3.3f. Please see R1.1a above.

R3.4: Inclusion of woody cover trend data from Hartfield & van Leeuwen (2018)

a. The obvious gap when trying to determine whether farmer’s perception of current woody cover and past change are accurate is some objective measure of actual cover. The authors address this by using data from a remote sensing study (Hartfield & van Leeuwen 2018) to compare perceptions of change to published rates of change on a per-county basis. While this is a good idea, there are several problems with how this was done and reported.

Response R3.4a. See response R3.4c

b. The first issue concerns the scale mismatch between the perceptions at the individual landowner level and the change data at the level of the county. The authors explain that this was unavoidable, but the implications are not adequately explored. No data of actual woody cover in each county, and some indication of how variable it is in space, are presented. If woody cover in a county varies a lot spatially, then linking it to individual landowner responses will be a bit of a stretch.

Response R3.4b. See response R3.4c

c. The survey was done in 2015, but the study that is cited as the source of the actual change data was published a couple of years later in 2018. I assume that the remote sensing study therefore did not inform the choice of counties for conducting the survey, but that the data were later accessed to compare perceptions of change to actual change. This is a pity, since the motivation for the survey was that woody encroachment is a significant issue, whereas the remote sensing study found that many counties did not actually experience an increase in woody cover. Possible reasons for the discrepancy between landowners’ perceptions are discussed, but in a rather speculative fashion.

Response R3.4c. We did not have access to the woody cover data cited in the paper during project design. However we have added data from the same source as the % cover change about actual woody plant coverage by county into Table 5.

d. I was unable to ascertain where the authors obtained the data shown in Table 7. No source is given in the table caption, but the source is indicated in the text to be Hartfield & van Leeuwen (2018). I looked at the Hartfield & van Leeuwen (2018) paper and the data do not appear in it. Instead, Hartfield and van Leeuwen state in their discussion that “In a future paper, we will use estimates of woody cover from two dates, to compute the change in woody cover for each pixel in the two ecoregions.” So where do the data cited in the submitted manuscript come from?

Response R3.4e. The data were obtained directly from Hartfield and van Leeuwen, who were Co-PIs on a larger project of which our survey was one of our components.

e. Since the authors did manage to access woody change cover for the counties in which their survey was done (and will hopefully be able to provide a correct citation), they should also have been able to source and report woody cover in 2004 and 2014, the latter being the year before the survey. This would allow not only comparison of woody cover estimates and actual woody cover, but also whether actual woody cover in a county influenced perceptions and use of prescribed burning.

Response R3.4e. We have added in the woody cover estimates for 2002 and 2014 from the same source as the % change in woody cover. However, we elected not to add county cover estimates into our models concerning perception and use of fire because some respondents may own property in other counties and accordingly their perceptions are not informed only from woody plant cover in the counties in which their sampled property is located.

f. Lastly, the perceptions of trends and the comparison with published woody cover trends should have been reported in the results section. Instead, they come as something of a surprise in the form of a table and accompanying text at the beginning of the discussion section. The discussion should not introduce new findings but needs to interpret the findings reported in the results section.

Response R3.4a. These results have been moved to the results subsection, “Woody plant cover perceptions and preferences.”

R3.5: For publication in an international journal, the introduction is rather narrowly focused on a local problem. I would have liked to see more international literature to contextualize the situation in Great Plains. The discussion should similarly draw conclusions that go beyond the regional.

Response R3.5. We have added more international literature to our discussion in order to draw comparisons with similar issues globally, e.g., Sutherland et al. 2011; Eburn and Cary 2018.

R3.6: Line 32: Were landowners just asked about the “amount” or also about perceived trends?

Response R3.6. They were also asked about their perceptions about increases in woody plant cover on their land. See R3.24 for more details.

R3.7: Lines 36 and 82: “lability” should be “liability”

Response R3.7. Done

R3.8: Line 39: Does membership of PBA cause landowners to be more likely to use fire, or do landowners who are comfortable using fire join PBAs more readily?

Response R3.8. That is a good question. Because of the requirement by many PBAs for members to assist with burns on other members properties before they receive assistance with a burn on their own property, the first alterative is more likely, i.e., that PBA cause landowners to be more likely to use fire. This is emphasized in the discussion where the following sentence has been added: “it has been found that that PBA’s are very effective in engaging landowners in the use of prescribed fire, because they provide their members with fire safety training and equipment and they promote safe fire application experience among members who assist each other on burn days.” We also added another reference to support this point.

R3.9: Lines 44-46: This final sentence of the abstract is rather vague.

Response R3.9. The sentence has been changed to: “Future outreach efforts to promote prescribed fire to maintain grasslands should more actively support prescribed burn associations, which an effective vehicle for increasing prescribed fire use by private landowners.”

R3.10: Line 50: The values of woody cover increase given here contrast with the declines in woody cover reported in Table 7.

Response R3.10. To address this issue, we come back to the trends reported by Berg et al. 2015 (Ref #1) in discussing the information derived from Hartfield & van Leeuwen (2018) and presented in Table 5.

R3.11: Lines 55-56: Use the singular (“climate change” and “land use”)?

Response R3.11. Done

R3.12: Lines 59-61: Why were PBA members explicitly included in the survey while membership of FMAs and the other organizations cited was seemingly not considered or explored at all? Also, since FMAs and NRCS are never mentioned again, there is no need to introduce the abbreviations here.

Response R3.12. To avoid confusion, we eliminated the term Fire Management Associations (FMAs), because it is equivalent to Prescribed Burn Associations (PBAs), which is the more generally used term and the term we use throughout the manuscript. We also deleted (NRCS).

R3.13: Line 71: The argument about under- or overestimating woody cover is valid but in the absence of actual woody cover data, the study does not actually determine whether farmers under- or over-estimated woody cove on their land.

Response R3.13. We changed the sentence to: “Accurately estimating woody cover density is difficult from ground level because visual foreshortening of dispersed trees seen from that perspective may suggest higher woody plant density than overhead measurements of canopy cover would indicate.”

R3.14: Lines 84-85: The first question is about landowner perceptions about woody plant expansion and how it correlates with actual trends but much of the focus of the study (and the reported analyses) is farmer’s estimation of current woody cover (rather than trend). “Trend” and “current cover” are not always clearly distinguished in the writing.

Response R3.14. We changed “actual trends” to “estimated changes” which more accurately represents the data we presented in Table 7. We did not use the term “current cover.

R3.15: Lines 96 and 99: Change “included” to “were”. “Included” implies that the subsequent list is a subset, when in fact all counties sampled are listed.

Response R3.15. Done

R3.16: Lines 116-117: What were the “five areas of enquiry including attitudes concerning woody plants and fire” and are they all reported on in this manuscript? If not, why not?

Response R3.16. Given that the not all 5 areas of inquiry are presented in the manuscript, we deleted this sentence as it creates ambiguity.

R3.17: Line 119 and elsewhere: Journal requirements include the use of SI units. Acres are not SI units. The values given need to be converted to hectares, or at least report equivalent values in SI units for the benefit of readers accustomed to the metric system.

Response R3.17. All property sizes were changed to ha with the acres shown in parentheses.

R3.18: Line 119: What proportion of landowners with properties > 100 acres does the sample of 1600 represent?

Response R3.18. 100%. Landowners with less than 100 acres were disqualified from inclusion.

R3.19: Line 120: What is the basis for the 100-acre minimum? Can the choice of this minimum value, and the assumption that owners of smaller properties would be less likely to use fire, somehow be substantiated (e.g. with a reference from the literature)?

Response R3.19. As we stated in the manuscript: “We applied the minimum acreage requirement with the assumption that landowners with smaller acreage properties are less likely to apply prescribed fire …”. This was based on information gathered during pre-survey meetings with personnel from the Natural Resources Conservation Services, Texas A&M AgriLife Extension, an Oklahoma Extension Service.

R3.20: Line 123: “Pulled from” seems rather colloquial. “Extracted from”?

Response R3.20. Done.

R3.21: Lines 124-125 – what percentage of PBA members do the 126 / 192 members from TX /OK represent?

Response R3.21. All identified PBA members in each study county were included in the study. If we had relied only on the sample extracted from the county tax records, we would have had a very small number of PBA-member respondents, which would have prevented us from making member/non-member comparisons.

R3.22: Line 131: “Participants” needs an apostrophe.

Response R3.22. Done

R3.23: Line 132: Close bracket after “trees”.

Response R3.23. Done

R3.24: Line 143: Was this a “yes/no” type of question?

Response R3.24. We used the 7-point response scale to answer the question: “To what extent do you DISAGREE OR AGREE with each of the following statements about woody plants?” … with respect to increases in juniper, mesquite and other woody plants during the last 10 years. This is now evident in the survey questionnaire we have added in the Supplemental information.

R3.25: Line 178: Should “assumption” be replaced with “prediction” or “hypothesis”?

Response R3.25. We replaced “assumption” with “prediction”.

R3.26: Table 3, last row: OK is abbreviated here but this is the only instance – keep use of abbreviations consistent or avoid them.

Response R3.26. Oklahoma has been spelled out throughout the manuscript.

R3.27: Lines 198-199 – how do the sample demographics compare to the general population? Also, all results should be consistently written in the past tense.

Response R3.27. We are uncertain how we should report general population statistics. We could report general demographic statistics for the 16 study counties or for each of the two states but do not see what this would add to the manuscript. Suffice it to say, that the demographic results provided approximate those from a plethora of other surveys conducted by the researchers during the last 20 years. The past tense has been used throughout in the landowner characteristics results.

R3.28: Line 204: But this figure is influenced by the fact that PBA members were actively targeted. What is the rate of PBA membership actually like among landowners?

Response R3.28. We could not determine which figure in line 204 the reviewer is referring to. No figure was referenced in that line or on that page or the subsequent page.

R3.29: Table 4. It would be useful to also include what percentage of respondents fall within the "ranching", "heritage", "hunting" and "farmer" land use categories.

Response R3.29. They are not landowner categories they are motivations that landowners have for owning their land. Each respondent may fit into more than one motivation category. We have changed the manuscript to reflect that they are landownership motivations and not categories of landowners (see table 6 and 7). The mean response score for each motivation is also presented in Table 2 to give readers an idea of which motivations were more common. For example, lifestyle/recreation motivations were overall high compared to the farming motivations.

R3.30: Paragraph following Line 207: It would be useful to include the published data on woody cover and trends in each county here and analyze the differences by county.

Response R3.30. Table 7 (now Table 5) and the associated text have been moved to this subsection to address this comment R3.39.

R3.31: Line 235: Did the questionnaire not probe landowners’ reasons for using / not using fire?

Response R3.31. We did not see how this comment relates to the statement on line 235.

R3.32: Line 235: What is meant by “legitimate” in this context?

Response R3.32. The line reference provided should have been 325. We changed the “legitimate” to “safe and effective.”

R3.33: Line 240: "Reported" is ambiguous – does this refer to woody cover reported by respondents or in the literature? Would “perceived” not be a more appropriate word here than “reported”?

Response R3.33. Reported was changed to perceived.

R3.34: Tables 5 and 6: p-values are not 0.000 – they are greater than zero and should be presented as < 0.001.

Response R3.34. Done

R3.35: Line 248: “low-cost, effective”: Is that how it was put in the questionnaire? If not, I would leave these value-laden terms out of the results section.

Response R3.35. Survey participants were asked the extent to which they agreed with the notion that prescribed fire is “more effective” and “less costly” “than other woody methods for controlling woody plants.” This language was based on the finding by Van Liew et al. 2012, which is cited and refenced in the manuscript. We have added “relatively” to the sentence to connect it to that paper, which is now also cited in this sentence to support the statement.

R3.36: Line 250: delete “By contrast”.

Response R3.36. Done

R3.37: Lines 255-256: Rather state this more intuitively, e.g. "older land owners were significantly less likely to have used fire...". The sentence needs a full stop.

Response R3.37. Done.

R3.38: Lines 261 onwards: This needs to be reported in the results section, not the discussion.

Response R3.38. Done

R3.39: Table 7 needs to cite the source of the woody cover change data. Also include actual woody cover in 2004 and 2014 for each county, as these values are of interest. For example, a small decline in woody cover might be outweighed in people’s perceptions by overall high woody cover values.

Response R3.39. The source of the woody cover change date was already cited as reference #14 in footnote b. The 2004 and 2014 woody plant cover data have also been added to Table 7 (now Table 5).

R3.40: Line 279: Any data / evidence / literature to support this?

Response R3.40. We have added three references about the effect of drought on woody plant mortality. They are Twidwell et al. 2014, Moore et al. 2016 and Croucher et al. 2019.

R3.41: Line 283: farmers may be frustrated by the woody cover, but the reader is frustrated by the absence of data on actual woody cover.

Response R3.41. We have deleted the statement. We did provide data about county level woody plant cover (Table 5) as well as landowner perceptions of woody cover on their own land, based on Figure 2. Measuring actual woody cover on each respondent’s land would be extremely difficult and beyond the scope of this study. For this reason, we obtained information about landowner perception of woody plant cover on their land.

R3.42: Lines 285-287: What does that mean? Not very enlightening when put like this unless one goes and reads the paper (beyond its title). Not only is it unclear what argument the cited paper makes, but there is also no explanation of how/why this would influence perceptions of actual woody cover.

Response R3.42. We have deleted the statement as it relates to the previous sentence.

R3.43: Lines 295 and 297: It would be useful here to know what the percentage of landowners of the rancher, farmer, heritage, hunter types was in in each state.

Response R3.43. These percentages have been included in Table 4.

R3.44: Line 298: Replace the question mark with a full stop.

Response R3.44. Done.

R3.45: Line 322: From what I can ascertain, actual willingness was not measured – “used fire more frequently” would seem more accurate here than “more willing”.

Response R3.45. We changed “more willing” to “appeared to be more accepting” as we did not measure frequency of prescribed fire use either.

R3.46: Line 332: How about training and support in the use of fire?

Response R3.46. The sentence has been modified to included this suggestion.

R3.47: Figure 1: For the benefit of an international audience, please label the states of Texas and Oklahoma.

Response R3.47. We have modified Figure 1, including labeling Texas and Oklahoma.

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(50.6KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0238688.r003

Decision Letter 1

Francisco X Aguilar

26 Mar 2020

PONE-D-19-20616R1

Landowner perceptions of woody plants and prescribed fire in the southern Great Plains, USA

PLOS ONE

Dear Dr. Stroman,

==============================

ACADEMIC EDITOR: The issue of non-random participation in the PBA was raised during the first review and it has not been addressed in this revised version. I agree with this concern clearly emphasized by Reviewer 3.

One statistical tool to compensate for the non-random nature of participation is the use of propensity score matching. Within STATA 13, which the authors are using, this is relatively simple to do these days. Please see: https://www.stata.com/manuals13/teteffectspsmatch.pdf

This step will strengthen the study tremendously. The author should, nonetheless, see this as a suggestion and may take a different analytical approach when addressing the issue of non-random participation.

The authors seem to dismiss the comment raised by the reviewer regarding non-response bias by indicating that because non follow-up phone calls were made this cannot be tested. The authors could conduct a simple comparison between early respondents and later respondents (e.g. first and second waves) and infer non-response bias from differences between these two groups.

See: Armstrong, J.S., Overton, T.S., 1977. Estimating non-response bias in mail surveys. Journal of Marketing Research 14 (3), 396–402.

The authors shall address the minor issues noted by reviewer 1.

==============================

We would appreciate receiving your revised manuscript by May 10 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Francisco X Aguilar

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Partly

Reviewer #3: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Reviewer #1: R1.1: Discuss how potential bias was handled

a. Survey recipients included 100 random landowners from each county in the study area,

plus PBA members. It seems that including a specific subset of the sample population

would potentially influence findings and make it difficult to draw conclusions, especially

since PBA members would already understand the value of prescribed fire and therefore

are potentially more aware of the presence of woody plants on their property. How was

this potential source of bias accounted for during the analyses?

Response R1.1a: Using PBA membership lists, we intentionally included all PBA

members in each study county because if we relied on a randomly drawn PBA

membership sample drawn from the general tax records, the chance of selecting an

adequately large sample of PBA members for comparison with non-member respondents

would have been extremely unlikely. One of the objectives of our study was to

understand the effect of PBA membership on perspectives about woody plant

encroachment and the use of fire. It is unclear how PBA membership would have

enhanced awareness of the presence of woody plants on their property; woody plants are

equally visible to both members and non-members. It is indeed likely that PBA members

have a better understanding of the use of prescribed fire; confirming this with our data is

the basis for arguing that greater landowner access to PBA membership could enhance

the broader use of this critically needed rangeland management tool. Previous research

has also shown that many PBA members were not always supportive of the use of

prescribed fire but rather that joining such a landowner organization enabled them to

change their mind about prescribed fire being a risky land management tool and,

consequently, to decide to apply fire on their own land (Kreuter et al. 2008). Finally,

when we report and discuss our findings, we are careful to discuss only in terms of our

respondents in order not to extrapolate any possible response bias to a larger population.

- It still appears there is potential sampling bias. When a specific subset of a larger population (PBA members) is sampled and responses combined with a random sampling of the larger population (randomly selected landowners), there is the possibility that the responses from the subset will influence the data set when all responses are analyzed together. It would be more appropriate to analyze each population separately and compare responses from each demographic. If one dataset is used for analyses then a clear description of how this would not influence the results needs to be stated.

b. Were any other sources of bias tested for? (e.g. violations of assumptions or outliers)

Response R1.1b: We did not test for other potential sources of response bias. Since our

data were derived primarily from specified response scales, outliers are not present.

Additionally, as indicated above, to avoid extrapolating potential bias, we refer to

respondents only, not landowners in general, in our presentation and discussion of results

and not extrapolated to the landowner population from which the sample was drawn.

- Providing information on potential bias and how it was handled will help to ensure that results from the data analyses are valid and accurate. Not only does this include outliers (within the data set, not just outside of the specified response scale), but violations of assumptions of statistical tests (e.g. additivity and linearity; normality; homoscedasticity/homogeneity of variance; independence). This information might also help address potential sampling bias.

c. How were unanswered questions handled? (i.e. was there any influence created by nonresponses

to specific items within questionnaire areas of inquiry)

Response R1.1c:

We used null values for unanswered questions, resulting in some variation of n values for

each variable. We based this on the assumption that data were missing completely at

random (MCAR) as we could not determine respondent’s reasons for not answering any

particular question. Based on this assumption, we did not impute MCAR data.

- Again, this is information that should be included in the manuscript.

R1.8: Proofreading

Several typos and grammatical errors exist and need corrected

Response R1.8: We have carefully proof read the revised manuscript to ensure correct

grammar and eliminate typos.

- Grammatical errors and typos persist; continue to revise. Replace colloquial wording with specific language.

Discussion and Conclusion

a. Discussion of perceived v. actual land cover (lines 299-317) – what about potential influence of PBA members due to their higher response rate (i.e. is it possible for PBA members to be more aware of woody encroachment)?

b. Clarify or expand on “doublethink and policy-ecology mismatches” (lines 313-314)

c. Be sure to relate discussion of prescribed fire and woody plant cover back to this research. Lines 354-372 focus on other research; there is an opportunity to relate discussion about PBAs (lines 365-366) to this research (Table 7)

d. Lines 375-376 – include reference

e. Conclusion paragraph (lines 381-392) – be sure to clearly summarize relationships to the three research questions and how this relates to opportunities for future research opportunities

Reviewer #3: The authors have not adequately addressed the concern raised in my review (which was also raised by Reviewer 1) that the addition of purposively sampled members of PBAs to a random sample of landowners results in a sample that, when analysed as a single data set, is not random. To be a truly random sample, every subject in the target population must have an equal chance of being selected in the sample. In the case of the combined random + PBA member data set, this is clearly not the case. I understand the reason for purposively sampling PBA members, as there are relatively few of them, but to include them in the overall analysis violates this key assumption. This is not acknowledged in the revised manuscript and no attempt was made to correct for this in the analyses.

The other possible sources of bias were evidently difficult to eliminate (mail survey – possible age bias), but the authors make a reasonable case in their rebuttal. However, this information which is of interest to the reader is not included in the revised text. All they have done is to state “Due to this limitation, we do NOT extrapolate our results to all landowners in the Southern Plains.” This may well be honest but if the aim was not to extrapolate to the bigger population, why conduct a massive survey?

I understand that the data collection can no longer be changed, and that the data generated from the survey are still valuable and deserve to be published. Nevertheless, in analyzing the data, and reporting and interpreting the results, care needs to be taken to acknowledge and deal with the issues.

For example, I commented:

R3.28: Line 204: But this figure is influenced by the fact that PBA members were actively targeted. What is the rate of PBA membership actually like among landowners?

The “figure” referred to was the value of 32 % in the sentence “About a third of respondents (32%) reported being a member of a PBA.” The percentage of PBA members relative to the general population is inflated by the inclusion of a non-random proportion of PBA members. So what then is the use of knowing that 32 % of the sample are self-reported PBA members? It does not reflect the true percentage of PBA members among the general population. Can this value be used to interpret findings that PBA membership significantly influences attitudes to fire management?

I think the authors need to make a better effort (perhaps with the support of a statistician or data analyst) to address this concern, so that the way the data are reported and interpreted is robust and can allow the reader to draw conclusions about land owners in the greater region, with the appropriate caveats if required. I am not a statistician and while I can see the problem, I cannot offer an alternative way of analyzing the data that gets around the assumption of a random sample – but a statistician or data analyst would no doubt be able to help.

Apart from this I am satisfied with the revisions and author responses.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: No

PLoS One. 2020 Sep 8;15(9):e0238688. doi: 10.1371/journal.pone.0238688.r004

Author response to Decision Letter 1

20 Jul 2020

PONE-D-19-20616R1 – Response to Reviewer Comments

Comment AE 1.1: The issue of non-random participation in the PBA was raised during the first review and it has not been addressed in this revised version. I agree with this concern clearly emphasized by Reviewer 3. One statistical tool to compensate for the non-random nature of participation is the use of propensity score matching. Within STATA 13, which the authors are using, this is relatively simple to do these days. Please see:https://www.stata.com/manuals13/teteffectspsmatch.pdf. This step will strengthen the study tremendously. The author should, nonetheless, see this as a suggestion and may take a different analytical approach when addressing the issue of non-random participation.

Response AE 1.1: To address the ongoing concern about non-random participation of PBA members, we have rewritten the description of our sampling approach as follows: “The survey included 1,918 landowners in Texas and Oklahoma who owned a minimum of 40.5 ha (100 acres). We applied this minimum size requirement with the assumption that landowners with smaller sized properties are less likely to apply prescribed fire, regardless of their perspectives of woody plant density. This survey sample was derived using a stratified sample selection approach. The first stratum consisted of general landowners included in the tax databases of each of the 16 selected counties. From each of these databases, 100 landowners who owned at least 40.5 ha were randomly selected for a first stratum sub-sample of 1,600 landowners. The second stratum consisted all members of PBA in the 16 selected counties. The population of PBA members in each of the two states was small enough that our PBA sample consisted the whole subgroup. This resulted second stratum sub-sample of 318 PBA members (126 in Texas and 192 in Oklahoma). To ensure that the PBA members were not a biased subset of the combined sample of 1,918 landowners, we compared responses from the two subsamples regarding perceptions of woody plants and prescribed fire and about the proportion of income derived from the property.”

The results of this analysis are included at the beginning of the results section. We found no significant difference between PBA members and non-members regarding the proportion of annual income derived from the property or about existing and preferred woody plant density (see Table 4 in the manuscript). However, PBA members were more likely to respond that woody plants are a problem on their land (non -members also agreed with that statement but not as strongly), and they were also much more likely to respond that they were in favor of applying prescribed fire to their own land. In the discussion we included the following statement to address these results: “While we did not find any difference between PBA members and non-members in their perceptions and preferences of woody plant densities on their land, we did find that PBA members were more likely to view woody plant expansion as a problem on their property and to apply prescribed fire to address this concern. However, the directionality of this difference is unclear; i.e., whether PBA membership elevated the perception of woody plants as a problem or if landowners who viewed woody plants as a problem were more likely to become a PBA member in order to apply fire using PBA membership benefits including access to labor, fire management equipment and fire training.”

Comment AE 1.2: The authors seem to dismiss the comment raised by the reviewer regarding non-response bias by indicating that because non follow-up phone calls were made this cannot be tested. The authors could conduct a simple comparison between early respondents and later respondents (e.g. first and second waves) and infer non-response bias from differences between these two groups. See: Armstrong, J.S., Overton, T.S., 1977. Estimating non-response bias in mail surveys. Journal of Marketing Research 14 (3), 396–402.

Response AE 1.2: As previously stated, we were not able to follow up with non-respondents so conducting traditional non-response tests was not possible for this study. Per the reviewers’ recommendation based on Armstrong and Overton (1977), which has been cited over 15,000, we have now attempted to test for potential non-response bias by comparing responses from early and late respondents, even though we could find no evidence that this method has been quantitatively verified to support it as an accurate estimate of non-response bias. Respondents were included in one of two groups based on receipt of their questionnaire during or after the first 3 weeks after the mailing date of the questionnaire (i.e., first and second wave responders). We used Mann-Whitney rank tests to compared the variable values of the two groups for five key variables: age of respondent; reliance on income from their property; perception of and preference for woody plant cover on their land, and whether or not they thought woody plants were a problem on their land. We found no statistically significant differences for any of the five variables between the two groups (p = 0.0707 - 0.7721). However, we are not convinced that testing early versus later respondents is a legitimate method for dealing with non-response bias as all the data are from respondents. Therefore, we have continued to be cautious about extrapolating our results to the broader population of landowners in the 16 counties from which the sample of general landowners was drawn. We have included this concern in the limitations paragraph in the discussion section.

Comment AE 1.3: The authors shall address the minor issues noted by reviewer 1.

Response AE 1.2: These responses are provided below.

Comment R 1.1: It still appears there is potential sampling bias. When a specific subset of a larger population (PBA members) is sampled and responses combined with a random sampling of the larger population (randomly selected landowners), there is the possibility that the responses from the subset will influence the data set when all responses are analyzed together. It would be more appropriate to analyze each population separately and compare responses from each demographic. If one dataset is used for analyses then a clear description of how this would not influence the results needs to be stated.

Response R 1.1: See response AE 1.1. above.

Previous Comment R1.1: Discuss how potential bias was handled

Previous Response R1.1a: Using PBA membership lists, we intentionally included all PBA members in each study county because if we relied on a randomly drawn PBA membership sample drawn from the general tax records, the chance of selecting an adequately large sample of PBA members for comparison with non-member respondents would have been extremely unlikely. One of the objectives of our study was to understand the effect of PBA membership on perspectives about woody plant encroachment and the use of fire. It is unclear how PBA membership would have enhanced awareness of the presence of woody plants on their property; woody plants are equally visible to both members and non-members. It is indeed likely that PBA members have a better understanding of the use of prescribed fire; confirming this with our data is the basis for arguing that greater landowner access to PBA membership could enhance the broader use of this critically needed rangeland management tool. Previous research has also shown that many PBA members were not always supportive of the use of prescribed fire but rather that joining such a landowner organization enabled them to change their mind about prescribed fire being a risky land management tool and, consequently, to decide to apply fire on their own land (Kreuter et al. 2008). Finally, when we report and discuss our findings, we are careful to discuss only in terms of our respondents in order not to extrapolate any possible response bias to a larger population.

Comment R 1.2: Providing information on potential bias and how it was handled will help to ensure that results from the data analyses are valid and accurate. Not only does this include outliers (within the data set, not just outside of the specified response scale), but violations of assumptions of statistical tests (e.g. additivity and linearity; normality; homoscedasticity/ homogeneity of variance; independence). This information might also help address potential sampling bias.

Response R 1.2: See response AE 1.1 and AE 1.2 above. We do not understand what the reviewer means by outliers within the data set – we had previously stated that the response data were primarily based on specified response options (generally 5).

Previous Comment R1.1b: Were any other sources of bias tested for? (e.g. violations of assumptions or outliers)

Previous Response R1.1b: We did not test for other potential sources of response bias. Since our

data were derived primarily from specified response scales, outliers are not present.

Additionally, as indicated above, to avoid extrapolating potential bias, we refer to

respondents only, not landowners in general, in our presentation and discussion of results

and not extrapolated to the broader landowner population from which the sample was drawn.

Comment R 1.3: Again, this is information that should be included in the manuscript.

Response R 1.3: We have included this information in the methods section of the manuscript as suggested.

Previous comment R 1.1c. How were unanswered questions handled? (i.e. was there any influence created by nonresponses to specific items within questionnaire areas of inquiry)

Previous Response R1.1c: We used null values for unanswered questions, resulting in some variation of n values for each variable. We based this on the assumption that data were missing completely at random (MCAR) as we could not determine respondent’s reasons for not answering any particular question. Based on this assumption, we did not impute MCAR data.

Comment R 1.4: Grammatical errors and typos persist; continue to revise. Replace colloquial wording with specific language.

Response R 1.4: We have again proofread the entire manuscript and asked the additional co-author, who is a statistician, to also read the document in order to identify the unspecified persistent “grammatical errors” and “colloquialisms” and made changes where appropriate. Without explicit specification, we do not know what other errors and colloquialisms the reviewer might be referring to.

Comment R 1.5: Discussion of perceived v. actual land cover (lines 299-317) – what about potential influence of PBA members due to their higher response rate (i.e. is it possible for PBA members to be more aware of woody encroachment)?

Response R1.5: See response AE 1.1. above.

Comment R 1.6: Clarify or expand on “doublethink and policy-ecology mismatches” (lines 313-314)

Response R1.6: This is a quotation from a cited reference which stated: “Mismatches between policy and ecology can lead natural resource agencies toward a “doublethink” mentality (where contradictory thoughts exist without acknowledged cognitive dissonance) that produces policies that simultaneously promote and control invasive species.” We have replaced the above quote with an expanded explanation based on this statement.

The new wording reads “Additionally, Roberts et al. (2018) discovered that nearly all policies aimed at containing juniper tend to promote rather than contain the spread of the species due to mismatches between policy and ecology and contradictory programs within natural resources agencies that simultaneously promote and control invasive species”

Comment R 1.7: Be sure to relate discussion of prescribed fire and woody plant cover back to this research. Lines 354-372 focus on other research; there is an opportunity to relate discussion about PBAs (lines 365-366) to this research (Table 7)

Response R1.7: The line numbers to which the reviewer refers are not included in the revised version of the manuscript that goes only to line 368.

Comment R 1.8: Lines 375-376 – include reference

Response R1.8: Unclear to what the reviewer is referring – in our manuscript the maximum line number is 368.

Comment R1.9: Conclusion paragraph (lines 381-392) – be sure to clearly summarize relationships to the three research questions and how this relates to opportunities for future research opportunities.

Response R1.9: The concluding paragraph has been revised to summarize the relationship to the three research questions and to indicate opportunities for future research.

Reviewer Comment 3.1: The authors have not adequately addressed the concern raised in my review (which was also raised by Reviewer 1) that the addition of purposively sampled members of PBAs to a random sample of landowners results in a sample that, when analyzed as a single data set, is not random. To be a truly random sample, every subject in the target population must have an equal chance of being selected in the sample. In the case of the combined random + PBA member data set, this is clearly not the case. I understand the reason for purposively sampling PBA members, as there are relatively few of them, but to include them in the overall analysis violates this key assumption. This is not acknowledged in the revised manuscript and no attempt was made to correct for this in the analyses.

Response R3.1: See response AE 1.1. above.

Reviewer Comment 3.2: The other possible sources of bias were evidently difficult to eliminate (mail survey – possible age bias), but the authors make a reasonable case in their rebuttal. However, this information which is of interest to the reader is not included in the revised text. All they have done is to state “Due to this limitation, we do NOT extrapolate our results to all landowners in the Southern Plains.” This may well be honest but if the aim was not to extrapolate to the bigger population, why conduct a massive survey?

Response R3.2: See response AE 1.2 above.

Reviewer Comment 3.3: I understand that the data collection can no longer be changed, and that the data generated from the survey are still valuable and deserve to be published. Nevertheless, in analyzing the data, and reporting and interpreting the results, care needs to be taken to acknowledge and deal with the issues. For example, I commented: R3.28: Line 204: But this figure is influenced by the fact that PBA members were actively targeted. What is the rate of PBA membership actually like among landowners? The “figure” referred to was the value of 32 % in the sentence “About a third of respondents (32%) reported being a member of a PBA.” The percentage of PBA members relative to the general population is inflated by the inclusion of a non-random proportion of PBA members. So what then is the use of knowing that 32 % of the sample are self-reported PBA members? It does not reflect the true percentage of PBA members among the general population. Can this value be used to interpret findings that PBA membership significantly influences attitudes to fire management? I think the authors need to make a better effort (perhaps with the support of a statistician or data analyst) to address this concern, so that the way the data are reported and interpreted is robust and can allow the reader to draw conclusions about land owners in the greater region, with the appropriate caveats if required. I am not a statistician and while I can see the problem, I cannot offer an alternative way of analyzing the data that gets around the assumption of a random sample – but a statistician or data analyst would no doubt be able to help.

Response R3.3: We have consulted with a statistician, who is now a co-author of the manuscript, about the potential bias concerns. The statistician helped us craft revisions in the manuscript and the preceding responses to address this issue to the extent possible. We have also addressed these issues in a limitations and future research paragraph at the end of the discussion section.

Attachment

Submitted filename: PONE-D-19-20616R1 Reviewer Response.docx

Click here for additional data file.^{(30.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0238688.r005

Decision Letter 2

Francisco X Aguilar

5 Aug 2020

PONE-D-19-20616R2

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

PLOS ONE

Dear Dr. Stroman,

Please submit your revised manuscript by Sep 19 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Francisco X Aguilar

Academic Editor

PLOS ONE

Additional Editor Comments:

The authors have engaged in a major revision of the study. Although some differences remain between the reviewers' identified criticisms and the authors' approach to dealing with them, I deemed these to be acceptable.

As a minor revision, I will suggest the authors have a separate 'Conclusions', from a 'Discussion and Conclusion', section. This would largely be the last paragraph in the current section.

As the authors engage in this minor revision, I will also ask them to try to offer clear answers in the Conclusions section to their three specific questions: 1) Do landowner perceptions about woody plant expansion correspond with estimated changes in their area? 2) Do landowners within the Southern Plains desire less woody cover on their land? 3) Do expressed woody cover preferences influence the adoption of prescribed fire on private land?

This might seem like a small change but it will help with readability and the potential impact of this paper.

We appreciate you considering PLOS One as an outlet for your research.

[Note: HTML markup is below. Please do not edit.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Sep 8;15(9):e0238688. doi: 10.1371/journal.pone.0238688.r006

Author response to Decision Letter 2

20 Aug 2020

Editor comment:

As a minor revision, I will suggest the authors have a separate 'Conclusions', from a 'Discussion and Conclusion', section. This would largely be the last paragraph in the current section.

This might seem like a small change but it will help with readability and the potential impact of this paper.

Response: We have made the requested changes in the manuscript, focusing the conclusion section on the 3 research questions.

Attachment

Submitted filename: Reviewer response 8-20-2020.docx

Click here for additional data file.^{(12.9KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0238688.r007

Decision Letter 3

Francisco X Aguilar

24 Aug 2020

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

PONE-D-19-20616R3

Dear Dr. Stroman,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Francisco X Aguilar

Academic Editor

PLOS ONE

PLoS One. doi: 10.1371/journal.pone.0238688.r008

Acceptance letter

Francisco X Aguilar

27 Aug 2020

PONE-D-19-20616R3

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

Dear Dr. Stroman:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Francisco X Aguilar

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. Mail survey questionnaire.

(PDF)

Click here for additional data file.^{(584KB, pdf)}

S2 File. Anonymized data file.

(XLS)

Click here for additional data file.^{(289.5KB, xls)}

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file.^{(50.6KB, docx)}

Attachment

Submitted filename: PONE-D-19-20616R1 Reviewer Response.docx

Click here for additional data file.^{(30.2KB, docx)}

Attachment

Submitted filename: Reviewer response 8-20-2020.docx

Click here for additional data file.^{(12.9KB, docx)}

Data Availability Statement

All relevant data are within the manuscript and its Supporting Information files.

[pone.0238688.ref001] 1.Berg MD, Sorice MG, Wilcox BP, Angerer JP, Rhodes EC, et al. (2015) Demographic Changes Drive Woody Plant Cover Trends—An Example from the Great Plains. Rangeland Ecology & Management 68: 315–321. [Google Scholar]

[pone.0238688.ref002] 2.Harr RN, Morton LW, Rusk SR, Engle DM, Miller JR, et al. (2014) Landowners perceptions of risk in grassland management: woody plant encroachment and prescribed fire. Ecology and Society 19. [Google Scholar]

[pone.0238688.ref003] 3.Huxman TE, Wilcox BP, Breshears DD, Scott RL, Snyder KA, et al. (2005) Ecohydrological implications of woody plant encroachment. Ecology 86: 308–319. [Google Scholar]

[pone.0238688.ref004] 4.Twidwell D, Rogers WE, Fuhlendorf SD, Wonkka CL, Engle DM, et al. (2013) The rising Great Plains fire campaign: citizens' response to woody plant encroachment. Frontiers in Ecology and the Environment 11: e64–e71. [Google Scholar]

[pone.0238688.ref005] 5.Archer S, Schimel DS, Holland EA (1995) Mechanisms of shrubland expansion: land use, climate or CO2? Climatic Change 29: 91–99. [Google Scholar]

[pone.0238688.ref006] 6.Kreuter UP, Woodard JB, Taylor CA, Teague WR (2008) Perceptions of Texas landowners regarding fire and its use. Rangeland Ecology & Management 61: 456–464. [Google Scholar]

[pone.0238688.ref007] 7.Lee AC, Conner JR, Mjelde JW, Richardson JW, Stuth JW (2001) Regional Cost Share Necessary for Rancher Participation in Brush Control. Journal of Agricultural and Resource Economics 26: 478–490. [Google Scholar]

[pone.0238688.ref008] 8.Kahneman D (2011) Thinking, fast and slow 1st ed Kahneman Daniel. New York: Farrar, Straus and Giroux. [Google Scholar]

[pone.0238688.ref009] 9.Van Liew D, Conner JR, Kreuter UP, Teague WR (2012) An economic assessment of prescribed extreme fire and alternative methods for managing invasive brush species in Texas: A modeling approach. The Open Agricultural Journal 6: 17–26. [Google Scholar]

[pone.0238688.ref010] 10.Weir JR, Kreuter UP, Wonkka CL, Stroman DA, Russell M, et al. (2019) Liability and prescribed fire: Perception and reality. Rangeland Ecology & Management 72: 533–538. [Google Scholar]

[pone.0238688.ref011] 11.Morton LW, Regen E, Engle DM, Miller JR, Harr RN (2010) Perceptions of Landowners Concerning Conservation, Grazing, Fire and Eastern Redcedar Management in Tallgrass Prairie. Rangeland Ecology & Management 63: 645–654. [Google Scholar]

[pone.0238688.ref012] 12.Toledo D, Kreuter UP, Sorice MG, Taylor CA (2012) To burn or not to burn: Ecological restoration, liability concerns and the role of prescribed burning associations. Rangelands: 18–23. [Google Scholar]

[pone.0238688.ref013] 13.Dillman DA (2000) Mail and internet surveys: The tailored design method. New York, NY: John Wiley and Sons. [Google Scholar]

[pone.0238688.ref014] 14.Armstrong JS, Overton TS (1977) Estimating Nonresponse Bias in Mail Surveys. Journal of Marketing Research 14: 396–402. [Google Scholar]

[pone.0238688.ref015] 15.Hartfield K, van Leeuwen W (2018) Woody Cover Estimates in Oklahoma and Texas Using a Multi-Sensor Calibration and Validation Approach. Remote Sensing 10: 632. [Google Scholar]

[pone.0238688.ref016] 16.Schmitt N (1996) Uses and abuses of coefficient alpha. 8: 350–353. [Google Scholar]

[pone.0238688.ref017] 17.Cortina JM (1993) What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology 78: 98–104. [Google Scholar]

[pone.0238688.ref018] 18.StataCorp (2015) Stata Statistical Software: Release 14. 14.0 ed. College Station.

[pone.0238688.ref019] 19.Twidwell D, Wonkka CL, Taylor CA Jr, Zou CB, Twidwell JJ, et al. (2014) Drought-induced woody plant mortality in an encroached semi-arid savanna depends on topoedaphic factors and land management. Applied Vegetation Science 17: 42–52. [Google Scholar]

[pone.0238688.ref020] 20.Crouchet SE, Jensen J, Schwartz BF, Schwinning S (2019) Tree Mortality After a Hot Drought: Distinguishing Density-Dependent and -Independent Drivers and Why It Matters. Frontiers in Forests and Global Change 2. [Google Scholar]

[pone.0238688.ref021] 21.Moore GW, Edgar CB, Vogel JG, Washington-Allen RA, March Rosaleen G, et al. (2016) Tree mortality from an exceptional drought spanning mesic to semiarid ecoregions. Ecological Applications 26: 602–611. 10.1890/15-0330 [DOI] [PubMed] [Google Scholar]

[pone.0238688.ref022] 22.Roberts CP, Uden DR, Allen CR, Twidwell D (2018) Doublethink and scale mismatch polarize policies for an invasive tree. PLoS ONE 13. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0238688.ref023] 23.Hurst KF, Ramsdell CP, Sorice MG (2017) A life course approach to understanding social drivers of rangeland conversion. Ecology and Society 22. [Google Scholar]

[pone.0238688.ref024] 24.Sorice MG, Kreuter UP, Wilcox BP, Fox Iii WE (2014) Changing landowners, changing ecosystem? Land-ownership motivations as drivers of land management practices. Journal of Environmental Management 133: 144–152. 10.1016/j.jenvman.2013.11.029 [DOI] [PubMed] [Google Scholar]

[pone.0238688.ref025] 25.Bertone R (2019) Oklahoma's top 10 agricultural products. Farm flavor. [Google Scholar]

[pone.0238688.ref026] 26.Becerra TA, Engle DM, Elmore RD, Fuhlendorf SD (2013) Contrasting Preference for Grassland Landscapes Among Population Groups in the Central and Southern Great Plains. Rangeland Ecology & Management 66: 529–538. [Google Scholar]

[pone.0238688.ref027] 27.Juzwik J, Appel DN, MacDonald WL, Burks S (2011) Challenges and Successes in Managing Oak Wilt in the United States. Plant disease 2011 v.95 no.8: pp. 888–900. 10.1094/PDIS-12-10-0944 [DOI] [PubMed] [Google Scholar]

[pone.0238688.ref028] 28.Olenick KL, Kreuter UP, Conner JR (2005) Texas landowner perceptions regarding ecosystem services and cost-sharing land management programs. Ecological Economics 53: 247–260. [Google Scholar]

[pone.0238688.ref029] 29.Brockway DG, Outcalt KW, Estes BL, Rummer RB (2009) Vegetation response to midstorey mulching and prescribed burning for wildfire hazard reduction and longleaf pine (Pinus palustris Mill.) ecosystem restoration. Forestry: An International Journal of Forest Research 82: 299–314. [Google Scholar]

[pone.0238688.ref030] 30.Wonkka CL, Rogers WE, Kreuter UP (2015) Legal barriers to effective ecosystem management: exploring linkages between liability, regulations, and prescribed fire. Ecological Applications 25: 2382–2393. 10.1890/14-1791.1 [DOI] [PubMed] [Google Scholar]

[pone.0238688.ref031] 31.Eburn M, Cary GJ (2017) You own the fuel, but who owns the fire? International Journal of Wildland Fire 26: 999–1008. [Google Scholar]

[pone.0238688.ref032] 32.Loomis JB, Bair LS, González-Cabán A (2001) Prescribed Fire and Public Support: Knowledge Gained, Attitudes Changed in Florida. Journal of Forestry 99: 18–22. [Google Scholar]

[pone.0238688.ref033] 33.Toledo D, Kreuter UP, Sorice MG, Taylor CA Jr (2014) The role of prescribed burn associations in the application of prescribed fires in rangeland ecosystems. Journal of Environmental Management 132: 323–328. 10.1016/j.jenvman.2013.11.014 [DOI] [PubMed] [Google Scholar]

[pone.0238688.ref034] 34.Weir JR, Twidwell D, Wonkka CL (2016) From Grassroots to National Alliance: The Emerging Trajectory for Landowner Prescribed Burn Associations. Rangelands. [Google Scholar]

[pone.0238688.ref035] 35.Kreuter UP, Stroman DA, Wonkka CL, Weir J, Abney AA, et al. (2019) Landowner Perceptions of Legal Liability for Using Prescribed Fire in the Southern Plains, United States. Rangeland Ecology & Management 72: 959–967. [Google Scholar]

[pone.0238688.ref036] 36.Sutherland L-A, Burton RJF (2011) Good Farmers, Good Neighbours? The Role of Cultural Capital in Social Capital Development in a Scottish Farming Community. Sociologia Ruralis 51: 238–255. [Google Scholar]

[pone.0238688.ref037] 37.Brunson MW, Huntsinger L (2008) Ranching as a Conservation Strategy: Can Old Ranchers Save the New West? Rangeland Ecology & Management 61: 137–147. [Google Scholar]

[pone.0238688.ref038] 38.Sorice MG, Kreuter UP, Wilcox BP, Fox WE III (2012) Classifying land-ownership motivations in central Texas, USA: A first step in understanding drivers of large-scale land cover change. Journal of Arid Environments 80: 56–64. [Google Scholar]

[pone.0238688.ref039] 39.Jackson-Smith D, Kreuter UP, Krannich RS (2005) Understanding the multidimensionality of property rights orientations: Evidence from Utah and Texas ranchers. Society and Natural Resources 18: 587–610. [Google Scholar]

PERMALINK

Landowner perceptions of woody plants and prescribed fire in the Southern Plains, USA

Dianne A Stroman

Urs P Kreuter

Carissa L Wonkka

Roles

Abstract

Introduction

Methodology

Study area

Fig 1. Map of study area.

Mail survey

Measuring woody cover perceptions and preferences

Fig 2. Survey questions used to gauge survey participants’ perception of and preference for woody plant cover on their property.

Data and regression analyses

Table 1. Dependent and independent variables used in cover preference regression model.

Table 2. Orthogonal varimax rotated factor loading results of PCA analysis of landownership motivations with Cronbach’s α measuring internal scale reliability.

Results

Response rates and non-response bias assessment

Table 3. Survey questionnaire response rates.

Table 4. Mean response scores from PBA members and non-member survey questionnaire responses testing non-random participation bias.

Landowner characteristics

Table 5. Survey respondents’ demographic and property characteristics.

Woody plant cover perceptions and preferences

Fig 3. Comparison of Texas and Oklahoma respondents’ perceptions of current woody plant cover on their land (black bars) with their preferred amount of woody plant cover (striped bars).

Fig 4. Difference between respondent’s preference for woody plant cover and the current amount of cover on their land.

Table 6. Landowner estimates of woody cover increases and documented woody cover increases by county between 2004–2014.

Regression analyses

Table 7. Ordinal logistic regression model for factors influencing woody plant cover preferences of respondents in Texas and Oklahoma.

Table 8. Logistic regression results on factors influencing use of prescribed fire on private lands.

Discussion

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Francisco X Aguilar

Roles

Author response to Decision Letter 0

Decision Letter 1

Francisco X Aguilar

Roles

Author response to Decision Letter 1

Decision Letter 2

Francisco X Aguilar

Roles

Author response to Decision Letter 2

Decision Letter 3

Francisco X Aguilar

Roles

Acceptance letter

Francisco X Aguilar

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases