Reference Cited

Trip Characteristics

Preference for certain waterfowl hunting trip characteristics is a foundational component of a respondent’s experience where respondents self-select trip characteristics that they believe will lead to a satisfying experience that fulfills their motivations (Schroeder and others, 2019). Respondents have substantial control over trip characteristics within the bounds of economic and logistical barriers and make those decisions to maximize well-being.

When asked about the circumstances under which they choose to participate in waterfowl hunting, respondents in both subflyways chose “both”; they regularly planned waterfowl hunting for themselves and went hunting when invited by someone else (table 1; fig 1A–E). There were differences between types of trip choices respondents made in the two subflyways, but the effect sizes were still classified as “low.” In the Mississippi subflyway, 57 percent of respondents reported seeking ducks and geese (fig. 1B). Respondents in the Atlantic subflyway were equally likely to seek both ducks and geese and 40 percent of respondents reported only seeking ducks. Respondents in both subflyways differed substantially from the national survey and were less likely to seek ducks and geese in a normal year. In the Atlantic subflyway, 19 percent of respondents reported that they did not seek either ducks or geese.

When asked about the flyways they tended to hunt in 2016, respondents in the Mississippi and Atlantic subflyways preferentially hunted in the flyway where they live (fig. 1C). There were differences in the land types that respondents reported they primarily hunted on, but these differences had low effect sizes. Most respondents in the Mississippi subflyway (44 percent) and Atlantic subflyway (56 percent) reported that they primarily hunted on public land. National survey respondents also primarily hunted on public land but were slightly more likely to hunt on land owned by a friend or family member than respondents in the Southeast region (fig. 1D). Respondents in the Atlantic subflyway (76 percent) and Mississippi subflyway (73 percent) reported they primarily took day trips, which was highly in line with the national survey response (fig. 1E). Respondents who took multiday trips hunting trips were a substantial minority in the Southeast region and the national survey.

Harvest Characteristics

Harvest characteristics for ducks and geese were statistically different between the Mississippi and Atlantic subflyway and had a low effect size. Respondents in the Atlantic subflyway reported harvesting substantially fewer ducks and geese than respondents in the Mississippi subflyway; respondents from the national survey reported harvesting even fewer ducks and geese. Respondents in both subflyways reported harvesting more ducks than geese and that they averaged 10 or fewer geese per year (table 2; fig. 2A, B). Thus, respondents in the Mississippi subflyway are more inclined to hunt geese than respondents in the Atlantic subflyway, but respondents in both subflyways preferentially seek and harvest ducks more than geese.

In the Mississippi subflyway, 51 percent of respondents reported that they harvested an average of 21–50 or more ducks annually in 2011–15; only 30 percent of respondents in the Atlantic subflyway reported the same (fig. 2A). In comparison, only 17 percent of respondents in the Mississippi subflyway and 7 percent of respondents in the Atlantic subflyway reported harvesting an average of 21–50 or more geese. In the Atlantic subflyway, 29 percent of respondents reported that they harvested 5 or fewer ducks on average across 5 years and 66 percent harvested 5 or fewer geese (fig. 2B). The respondents in the Mississippi subflyway reported harvesting more ducks than geese than the national survey respondents.

Avidity Characteristics

Avidity is a key metric when measuring a participant’s involvement in an activity and describes their level of participation (Salz and others, 2001). A participant’s level of involvement is an important factor in describing how a person feels about the activity, how likely they are to continue to participate, and how important it is to how they view themselves. Respondents in the Mississippi subflyway reported being more avid than respondents in the Atlantic subflyway and most hunted at least 11 days per year in the previous 5 years (table 3; fig. 3A–C). Similarly, the Mississippi subflyway had less turnover between years, and 70 percent of respondents reported hunting every year in the previous 5 years.

Most respondents in the Atlantic subflyway reported hunting all 5 years (56 percent) but had more turnover than the Mississippi subflyway and the national survey. In the Mississippi subflyway, 30 percent of respondents averaged hunting 21 or more days, but only 18 percent of respondents in the Atlantic subflyway hunted 21 or more days and most hunted 10 days or fewer (table 3; fig. 3B). The effect sizes were low for these differences, but important management implications may still exist given the role that avidity plays in describing waterfowl hunting involvement. The finding that most respondents tended to hunt each year across the previous 5 years supports findings that waterfowl hunting tends to have lower turnover compared with other recreational hunting activities (Hinrichs and others, 2020).

In the Mississippi and Atlantic subflyways and the national survey, respondents overwhelmingly reported hunting 0–10 days in 2016 (fig. 3C). The Mississippi subflyway reported hunting more days than respondents in the Atlantic subflyway, but most respondents in both subflyways reported hunting 10 days or fewer (table 3; fig. 3C).

Hunter Satisfaction and Motivations

Satisfaction with any recreational activity can be assessed using a variety of mechanisms, but satisfaction is always based on a person’s ability to achieve a range of expectations for participating (Brunke and Hunt, 2008). Some expectations will exert an increased effect on satisfaction overall based on its importance to that individual. Often, expectations are split into harvest- and nonharvest-related forms because participants tend to control directly for their nonharvest-related motivations in choosing when and where to hunt (Schroeder and others, 2018). Harvest-related satisfaction in waterfowl hunting is often assessed in the number of birds harvested or the number of times respondents are able to meet the daily bag limit in their flyway.

When respondents were asked about the minimum number of ducks they needed to harvest per trip to feel satisfied, respondents in the Mississippi subflyway tended to base their satisfaction on harvest more than respondents in the Atlantic subflyway; however, the effect size for the Mississippi subflyway indicates that the differences should be carefully considered when making management decisions. In the Atlantic subflyway, 60 percent of respondents were satisfied with 1–3 birds harvested per trip, but only 44 percent of respondents in the Mississippi subflyway felt the same, and most respondents required 3–5 ducks harvested (table 4; fig. 4A–C).

When asked about the number of times respondents needed to shoot their daily bag limit on their hunts to be satisfied, 17 percent of respondents in the Mississippi subflyway reported needing to occasionally shoot their limit and 38 percent in the Atlantic subflyway reported needing to occasionally shoot their limit (table 4; fig. 4B). In comparison, only 10 percent of respondents in the Atlantic subflyway reported needing to shoot their daily bag limit most of the time and 27 percent reported needing to occasionally shoot their limit. Few respondents in both subflyways reported needing to shoot their daily bag limit every time they hunted to be satisfied. Respondents were also asked how many times they were able to meet these expectations and shoot their daily bag limit in 2016. The responses for both subflyways corresponded well with their expectations for satisfaction (table 4; fig. 4C). This trend indicates that respondents in both subflyways are likely satisfied with their ability to regularly harvest their daily bag limit. Across these harvest-oriented motivations, respondents in the Mississippi subflyway tended to prioritize harvest in determining satisfaction more than respondents in the Atlantic subflyway and the national survey. This orientation was represented in the number of birds harvested and in meeting the bag limit allowed.

When asked about satisfaction with hunting regulations, like daily limits or the number of days in the season, most respondents were neutral (fig. 5A–G). The number of ducks typically present in the areas respondents hunted or the number of ducks respondents typically saw had the lowest satisfaction among both subflyways; respondents in the Mississippi subflyway were more satisfied than respondents in the Atlantic subflyway (fig. 5A–G). Respondents in both subflyways were most satisfied with nonharvest-related attributes, like the quality of the habitat where they hunted or their overall hunting experience. Mississippi subflyway respondents were consistently more satisfied with management and regulatory aspects of waterfowl hunting (daily limits, days in the season, ducks present, and quality of habitats), and this increased satisfaction accounted for the significant differences identified between the two subflyways (table 5).

Species Importance

The most relevant species to each flyway were selected by the NAWMP Human Dimensions Working Group and Public Engagement Team and tailored based on the opinions of experts in those flyways (Patton, 2018). Respondents from both subflyways were asked to describe the importance of selected species to their preferences of waterfowl hunting. Dabbling duck species like Spatula spp. (teal), Anas acutta (pintail), and Aix sponsa (wood ducks) were most important, followed closely by Anas platyrhynchos (mallard) and diving ducks. Sea duck species were reported as one of the least important species (fig. 6A–I). Similarly, geese species like Branta canadensis (Canada geese), Anser caerulescens (snow geese), and Branta bernicla (brant) were also less important, and 30 percent of all respondents reported that they were not at all important. Respondents in the Mississippi subflyway were asked about their preference of four particular species (diving ducks, geese, mallards, and other dabbling ducks); the results were similar with mallards and other dabbling ducks, like teal and pintails, being the most important species and diving ducks being the least important species (fig. 7A–D).

Barriers to Participation

Respondents were asked about the degree to which different barriers prevented them from participating in waterfowl hunting (fig. 8A–E). There was strong agreement between respondents in both subflyways among all barriers except for the lack of available public places for waterfowl hunting (table 6). Respondents in the Atlantic subflyway (36 percent) and Mississippi subflyway (25 percent) reported that available public places for hunting was extremely low or very limiting (fig. 8E). Respondents in both subflyways differed from the national survey in several key barriers and were significantly more concerned about crowding, hunting pressure, and interference with other respondents (fig. 8B–D). Respondents in the Atlantic subflyway (43 percent) and Mississippi subflway (49 percent) reported that overcrowding was extremely or very limiting. Respondents in the Mississippi subflyway (33 percent) and Atlantic subflyways (28 percent) reported that there was too much existing hunting pressure.

The least limiting barrier among all respondents was conflict with other respondents. Respondents in both subflyways (67 percent) reported that conflict with other hunters was slightly or not at all limiting, which was a much larger proportion in the national survey response (5 percent; fig. 8A). Although, interference from other respondents was a more pressing concern to respondents in the two subflyways than the national survey, at least half of the respondents reported interference was only moderately, or less, concerning.

Preferences for Regulations

When respondents were asked about their preferences for waterfowl hunting regulations and policy, there were substantial differences among the Mississippi and Atlantic subflyways and the national survey responses. Harvest and harvest-related regulations were viewed differently by the subflyway respondents. The Atlantic subflyway respondents reported a slightly lower priority on larger bag limits and longer season than the Mississippi subflyway respondents (table 7). Similarly, the Mississippi subflyway respondents placed a slightly higher emphasis on larger bag limits for mallard ducks. Although there were statistically significant differences among these regulation and policy priorities, the effect sizes were low.

The differences between respondents in the Mississippi and Atlantic subflyways were most pronounced (very high and very low differences) when they were asked about mallard-specific bag limits (fig. 9A–G). In the Mississippi subflyway, 12 percent of respondents reported that the largest drake mallard limits were a very high priority, and only 5 percent reported that it was a very low priority (fig. 9E). Compared with the Atlantic subflyway, only 6 percent of respondents reported that it was a very high priority, and 10 percent reported that it was very low. Among regulatory priorities where there was agreement between the subflyways, the largest possible waterfowl populations and simpler regulations were the highest priorities and ~80 percent (Mississippi subflyway) and ~65 percent (Atlantic subflyway) of respondents selected “high” or “very high.” Despite an emphasis on simpler regulations, most respondents in both subflyways moderately prioritized the largest possible bag limits and fewer species-specific limits (fig. 10A).

Identifying when regulatory changes might result in lower participation can be an important measure of regulatory preferences. Respondents were asked about the potential to not participate in waterfowl hunting based on changes to bag limits and hunting season length. Although there were subtle differences between the two subflyways, respondents’ preferences were still in line with the national survey responses (table 8; fig. 10A, B). Although the specific bag limit that respondents reported they would stop participating at differed between the two subflyways, Atlantic subflyway respondents (78 percent) and Mississippi subflyway respondents (80 percent) reported that they would stop participating in waterfowl hunting if bag limits decreased to three or fewer birds (fig. 10A).The same phenomena were observed when it came to season length and most respondents in either subflyway chose to stop participating if there were seven or fewer days in the season (fig. 10B). When respondents were asked about how challenging species-specific regulations were to comply with or understand, there were some differences in the magnitude of responses between the Mississippi and Atlantic subflyways. Overall, most respondents felt that the regulations were not difficult to understand or comply with in the field. More respondents in the Mississippi subflyway than in the Atlantic subflyway believed that species-specific regulations are not difficult to comply with or understand (table 9; fig. 11A, B).

There was clear disagreement between respondents in the Mississippi and Atlantic subflyways when it came to preferences for regulation intentions (table 9; fig. 11C). The Mississippi subflyway respondents showed a clear preference for regulations that maximize harvest opportunity by maintaining individual species regulations (62 percent) relative to the Atlantic subflyway respondents, who preferred simpler aggregate regulations even at the expense of harvest opportunity (54 percent). The subflyway respondents differed in these opinions from the national survey response, which was split equally between each preference (fig. 11C).

Identity and Connection to Waterfowl Hunting

Like avidity, asking recreation participants how strongly they identify with aspects of an activity can be an effective way to measure their involvement in that activity (Arlinghaus and others, 2020). Measuring connections to an activity or how central it is to the respondent’s sense of self describes how important it is to that person’s life or even how they view themselves and want others to perceive them. Respondents were asked to describe how strongly they identify with various aspects of waterfowl recreation and waterfowl hunting.

There was strong agreement among subflyway respondents when it came to identifying themselves with waterfowl and wetland-related activities (fig. 12A–E); ~65 percent of all respondents strongly or very strongly identified as duck hunters (fig. 12C). This result was also true for respondents who identified as “conservationists” and as “goose hunters” despite slight differences in the extremes (very low or very high). Few respondents identified as “birdwatchers” or “other hunter” and at least 22 percent of respondents reported that they do not identify with those activities at all. There was a key difference in respondents who identified as “other hunter.” Respondents in both subflyways were less likely to strongly identify as “other hunter” than a respondent in the national survey, and the Atlantic subflyway respondents were least likely to identify as “other hunter” (table 10).

Respondents were asked about how central or how much of a central role waterfowl hunting has in their lives (table 11; fig. 13A–C). Similarly, about 30 percent of respondents in both subflyways were neutral when asked whether waterfowl hunting was central to their life or if they organize their life around it. About half of the respondents in both subflyways disagreed or strongly disagreed that they would not know what to do if they could not go waterfowl hunting. Respondents in the Mississippi subflyway felt stronger than Atlantic subflyway respondents that waterfowl hunting played a central role in their lives ().

Participation and Trust in Organization

It is important to understand and describe participation in waterfowl and wetland-related organizations because it can be indicative of stronger social connections to an activity and provide information for the best avenues for communication and outreach. Respondents were asked how involved they were in common waterfowl hunting organizations, such as Ducks Unlimited and Delta Waterfowl (table 12; fig. 14A–C). Overall, respondents were not particularly active in any of the organizations and more than 70 percent of respondents in either subflyway suggesting that they are not involved in Delta Waterfowl or other regional waterfowl associations. Ducks Unlimited had the highest involvement for both subflyways and about half of all respondents had some involvement.

Describing the trust that respondents have in waterfowl-related organizations is imperative to understanding how communication and outreach are likely perceived (table 13; fig. 15A-G). Waterfowl hunting and conservation organizations were the most trusted organizations across both subflyways and about half of all respondents trusted them a lot or completely (fig. 15G). Similarly, Federal and State wildlife agencies and university researchers were also well trusted with 35–45 percent of respondents trusting them either a lot or completely; the only entity that respondents largely did not trust was elected officials (fig. 15C, E, F). Respondents in the Atlantic subflyway had less trust for elected officials than respondents in the Mississippi subflyway with 41 percent and 35 percent, respectively, stating that they did not trust them at all. A low effect size also confirmed differences between the Atlantic and Mississippi subflyways, which related to birding and bird conservation organizations and other conservation organizations.

Another aspect of involvement in a recreational activity beyond waterfowl or wetland-related organizations is how involved participants are in peripheral activities that relate to or support the activity. For waterfowl hunting, these activities are often conservation behaviors that are tied to protecting waterfowl or preserving waterfowl-related habitat. Respondents did not often participate in these conservation behaviors, and more than half of all respondents stated that they never participated in these activities (table 14; fig. 16A–F). The only exception was voting for candidates or ballot issues that support waterfowl and wetland conservation where only ~43 percent of all respondents stated they never participate. Similarly, the conservation behavior that was participated in the least was contacting elected officials or agencies with 74 percent of all respondents stating they never participate (fig. 16C). There were no differences among respondents in both Mississippi and Atlantic subflyways and the rest of the national response (table 8).

Demographics

The demographics of respondents in the Southeast fit closely with those demographics that are observed nationally in waterfowl hunting (Patton, 2018). The median age for respondents from both subflyways was 45 years old, and most respondents were between 35 and 55 years old (fig. 17A). Similarly, almost all respondents were male and only 4 percent female in the Atlantic subflyway and 2 percent in the Mississippi subflyway (table 15; fig. 17B). Slightly more respondents in the Mississippi subflyway achieved bachelor’s and graduate degrees than those in the Atlantic subflyway but the effect size was low (table 16; fig. 18A). The study did not identify a significant difference in the income of respondents between the two subflyways, and most respondents reported making less than $75,000 (fig. 18B). About 70 percent of respondents reported that their primary source of personal income did not come from a nature-related profession (fig. 18C; table 15). When assessing differences between the urban–rural gradient, where a person currently resides may be different from where they were raised. For that reason, respondents were asked where their childhood and current residence are on the urban–rural gradient. Respondents in the Mississippi subflyway were slightly more rural than respondents in the Atlantic subflyway, although the effect size was low (fig. 19A). For childhood residence, there were no statistically significant difference between the two subflyways, but the effect size was negligible, which indicates that they should be considered the same (fig. 19B).

Waterfowl Hunting Trip Preferences

Understanding preferences for waterfowl hunting trips is useful for managers who are considering options for allocating management resources and making regulatory decisions. One of the most effective methods for describing preferences for any outdoor recreational activity is discrete choice experiments (DCE; Ryan and others, 2008). A DCE is a form of statistical modeling framework called choice models that originates from an economics theory that assumes people will seek to maximize their own enjoyment (in other words, utility; McFadden, 1972). These types of models use hypothetical scenarios with strategically designated attributes and levels that are presented in a way that preference for each attribute can be predicted relative to all others included in the experiment. Respondents are asked to select their most preferred scenarios relative to other combinations of attributes across many different iterations, which allows researchers to estimate the importance and utility of each attribute and its levels.

Each respondent was asked 10 individual DCE questions with predetermined attributes and levels so that each combination was asked equally across all respondents. The DCE in this survey was given as a two-option experiment with an opt-out. Each respondent received and was asked to select among three options: an “a” scenario, a “b” scenario, and an “opt-out” (neither “a” nor “b”). When respondents selected the opt-out, it implied that they would rather not participate if “a” and “b” were their only two options. The attributes selected for this DCE were harvest, access difficulty, travel time, waterfowl abundance, and competition with other respondents (table 17). These attributes and their associated levels were selected based on input from respondents in 13 different qualitative workshops that were administered across the country.

There are several ways to estimate preference using a DCE, and most ways estimate a central preference that represents all respondents included in the experiment. The issue with using these central tendency-based solutions is the underlying assumption that the sample population is homogenous. There is substantial evidence that this is a faulty assumption and that most outdoor recreator populations are heterogenous with clearly defined groups based on circumstances, backgrounds, and motivations.

In recognition of limitations of assuming a homogenous population, we estimated waterfowl hunter preference using a method called “latent class analysis” (LCA). This method is a likelihood-based estimation method (probabilistic) that hypothesizes a certain number of groups within the overall sample population that have unique central tendencies relative to other groups (Swait, 1994). Using the hypothesized number of groups, individual estimates of preference among the included attributes and levels are obtained for each. Based on each respondents’ selections, a probability of membership to each group can be calculated.

Latent class analysis is a model-based method where standard model selection methods are applied, selecting the number of groups that account for the most uncertainty. Instead of model selection, we chose to use a more limited number of groups that are more appropriate to meet the realities of management given input from U.S. Fish and Wildlife Service decision making. We chose to use a three-group estimation that was conducted using the Sawtooth Software LCA package (Sawtooth Software, 2019). Attribute importance for each group was calculated based on the preferred attributes—attributes that had the greatest effect on a respondent’s enjoyment—that contributed most to determining which respondents were most likely in that group.

Attribute importance is best used to determine what drives heterogeneity within the population and what is effective for naming each group (fig. 20A–C). For group 1, we found that 31 percent of importance was based on lower travel time, 24 percent on increased harvest, and 22 percent on lower competition. Respondents in group 2 placed slightly less importance on travel time (25 percent) but much more importance on decreased competition (40 percent). Respondents in group 3 placed much more importance on increased harvest (43 percent) and less importance on travel time (21 percent) and competition than respondents in groups 1 and 2. Based on these results, group 1 is more of a generalist group that places similarly high importance on multiple attributes (fig. 20A). Group 2 is characterized by being averse to competition and is likely to value seclusion more than other attributes (fig. 20B). Group 3 is comprised of harvest-driven individuals whose enjoyment is largely based on the number of birds they can harvest (fig. 20C). Once groups are determined, positive or negative changes in respondent enjoyment in each group can be visualized (fig. 21A–C). Enjoyment is visualized using part-worth utility, which is a hypothetical, unitless measure that expresses enjoyment such that as values increase, so does enjoyment of individuals in the group. The enjoyment values are additive, which means that a cumulative utility value can be calculated by measuring the part-worth utility values for each attribute that respondents are likely to experience and adding them together. Therefore, a key aspect of the utility for each attribute is the level where it changes to a negative value because this level is where enjoyment begins to decrease for group members.

For the generalists (group 1), travel time was the most important attribute and distance traveled more than 2 hours is where generalists begin to experience less enjoyment for the trip (fig. 21A). The same was true for respondents in the seclusionists (group 2; fig. 21B). The harvest-oriented respondents (group 3) did not experience negative utility until 3 hours, which means they can be expected to go farther more consistently (fig. 21C). This information helps decisionmakers because they can expect that many respondents are unlikely to travel farther than 2 hours for a single trip.

The range of part-worth utility can be evaluated between the lowest level and the highest level for each group. For example, evaluating the enjoyment that members of each group obtain from increasing harvest shows the seclusionist group experiences relatively little change in utility in comparison to the harvest-oriented group who experience a drastic increase in enjoyment for each additional bird harvested (fig. 21B, C). This is an important finding given that levels of enjoyment were not strongly positive or negative among any of the groups for two related attributes—bird abundance and access. This finding would imply that these attributes are less important than others attributes in the DCE because they contribute little to overall enjoyment for respondents.

Finally, “None” represented the choice each respondent had to opt-out rather than select one of the two scenarios that were presented (fig. 21). The choice represents the change in enjoyment that each group would experience to choose not to participate. The generalist group was the only group that had a positive utility for not participating, which implies that they are also more likely to not participate if the other attributes that are available are suboptimal. The seclusionist group expressed a substantial negative utility by choosing not to participate, which indicates that a substantial part of their total utility is derived just from participating and are likely to participate no matter what other attributes they perceive to be available (fig. 21B).

Respondents in both subflyways and all three latent groups were most likely currently living in a rural area (fig. 22A–C). The urban–rural gradient of current residence was not predictive of preference among the surveyed respondents and was similar across the three latent groups; rural seclusionists were the most prevalent group in both subflyways (fig. 22B). The harvest-oriented group was the least prevalent in both subflyways, but it is important to recognize that increasing harvest was important to all groups of respondents, as shown by the increasing part-worth utility for the increased harvest attribute in figure 21. The harvest-oriented group placed a larger importance on increased harvest relative to other attributes and will benefit more by achieving increased harvest (fig. 21C).

Social Connections to Waterfowl and Wetland Conservation

The number of social ties to an outdoor recreational activity a person possesses measures how important that activity is to a respondent and their overall involvement. Social connections correspond to concepts around social capital and networks. Social capital can refer to a variety of functions, but all incorporate some aspect of the respondent’s social network and the corresponding actions of individuals within that network. Although social capital is often represented as social goods (for example, relevant information, education, or training related to a subject), it can also represent the positive value or status gained within one’s social network from showing behaviors that are positively associated with certain activities, professions, or organizations related to that activity. As the number of social ties increases, the more social value an individual would be expected to gain from showing behaviors closely connected to the activity being considered. There are different forms of social ties, such as relatives, close friends, or acquaintances. People are generally more emotionally invested in family and close friends, so these social ties are considered stronger than ties like acquaintances.

Respondents were asked to report whether they had at least one acquaintance, close friend, or relative who participated in a range of 24 activities, professions, or organizations related to waterfowl hunting. The following options were provided for activities:

Respondents in the Mississippi subflyway had more social ties overall to waterfowl hunting-related categories than respondents in the Atlantic subflyway (figs. 23, 24). In the Mississippi subflyway, 43 percent of respondents had between 6 and 13 strong social ties as compared to only 33 percent of respondents in the Atlantic subflyway. Mentorship and facilitating hunting for others were additional aspects of social connections that respondents were asked about. Respondents were asked if they had taken anyone who had not hunted waterfowl in the previous year. Responses were similar across both subflyways and just greater than 50 percent of respondents reported that they had not done so (fig. 25).