In the past two decades, Salt Plains National Wildlife Refuge has been increasingly recognized as important habitat for both breeding and migratory shorebirds. North American snowy plovers Charadrius nivosus in particular rely on the nearly 5,000-ha salt flat at Salt Plains National Wildlife Refuge, which thousands use as breeding and stopover habitat. Elsewhere on the Southern Great Plains, decadal declines up to 75% within snowy plover subpopulations have been documented and attributed to vegetation encroachment, increased rates of nest predation, and decreased availability of fresh surface water. Despite many attempts to estimate this species' abundance across the continent, to date, no known attempt at distance sampling of snowy plovers has occurred. To address this paucity of data, we assessed feasibility of distance sampling methods to accurately estimate snowy plover abundance and detectability. Distance sampling surveys (2017–2018) indicated high detection probability (P = 0.80) and the population abundance estimate across the salt flat extrapolated to 3,307 individuals. The distance-sampling population abundance estimate is lower than population abundance estimates determined by two previous studies within the past decade but far greater than 2,105 estimated for a study in 2006. Overall, distance sampling snowy plovers at Salt Plains National Wildlife Refuge proved to be an effective addition to pre-established survey protocols but further investigation is needed to compare accuracy and precision of methods used in this study, annual surveys conducted by Salt Plains National Wildlife Refuge, and other potential snowy plover surveys.
Worldwide shorebird populations have declined 70% in the past 40 y (NABCI 2016) and some regional populations of snowy plovers Charadrius nivosus on the Southern Great Plains (SGP) mirror these declines (Saalfeld et al. 2013; Heath 2019). The Pacific coastal population of snowy plovers has been similarly, but perhaps more severely affected (Page et al. 1991, 2009), resulting in their listing as threatened in accordance with the Endangered Species Act (ESA 1973, as amended; USFWS 1993). Both western and SGP population declines are largely attributed to habitat degradation caused by vegetation encroachment (Winton et al. 2000; USFWS 2007; Saalfeld et al. 2012), disturbance (Wilson-Jacobs and Meslow 1984; Lafferty 2001; Page et al. 2009; Hardy and Colwell 2012; Saalfeld et al. 2013), increased rates of nest predation (Saalfeld et al. 2011, 2013; Hardy and Colwell 2012), and loss of reliable sources of fresh surface water (Andrei et al. 2009; Saalfeld et al. 2012, 2013; Heintzman et al. 2017). Some success in slowing, or reversing, these declines has been observed among the Pacific coastal population as a result of predator management and efforts to mitigate disturbance (Lafferty et al. 2006; Hardy and Colwell 2008; Mullin et al. 2010). The success of these management efforts highlights the necessity to monitor and accurately assess at-risk populations to better inform conservation and management objectives.
Snowy plover habitat includes sparsely vegetated coastal beaches, riparian sandbars, saline lake shorelines, and alkaline salt flats (Page et al. 2009). Salt Plains National Wildlife Refuge (NWR; also referred to here as the “Refuge”) in Oklahoma is a continentally significant stopover and breeding site for numerous species of shorebirds (Morrison et al. 2006; Thomas et al. 2012). In 1994, the Western Hemisphere Shorebird Reserve Network listed Salt Plains NWR as a Site of Regional Importance. Twenty-five years later in 2019, the designation was upgraded to a Site of International Importance as a result of greater understanding of the continental proportion and abundance of snowy plovers that breed at Salt Plains NWR (https://whsrn.org/whsrn_sites/salt-plains-nwr/, October 2020). Hosting perhaps 20% of the breeding population of snowy plovers in North America, Salt Plains NWR is second only to Great Salt Lake in Utah with respect to overall abundance of breeding snowy plovers (Thomas et al. 2012). However, with respect to the size of the Refuge, Salt Plains NWR has roughly 98% less available habitat area than Great Salt Lake and thus supports the greatest density of breeding snowy plovers (Thomas et al. 2012) and more migratory individuals (rather than simply breeding individuals) than anywhere in North America (Skagen et al. 1999). Known and potential threats to snowy plovers at Salt Plains NWR include saltcedar invasion (Tamarisk spp.; Hensley et al. 2014), heavy metal contamination (Ashbaugh et al. 2018), increased frequency and intensity of flood events (Karl and Knight 1997; USGCRP 2009; Villarini et al. 2013), and local habitat degradation due to oil and gas extraction activities (Hensley et al. 2014). Saltcedar invades open salt flat areas used by nesting snowy plovers. Conversion of open salt flats to areas characterized by saltcedar brush and forest fragments diminishes snowy plover nesting habitat, reduces freshwater sources required for snowy plover prey (invertebrates), and reduces the ability of snowy plovers to successfully thermoregulate their nests (Andrei et al. 2008; Saalfeld et al. 2012; Hensley et al. 2014; Heath 2019). Saltcedar treatment and removal from salt flats is a primary focus of current habitat management activities at Salt Plains NWR (Hensley et al. 2014; USFWS 2017).
Prior population estimates of snowy plovers at Salt Plains NWR are either based on (unspecified) extrapolations, random sampling, or stratified random sampling (Morrison et al. 2001, 2006; Thomas et al. 2012; Heath 2019), all of which have their own assumptions and varying levels of reliability. Accuracy and precision of population monitoring are critical components of avian conservation and depend largely on consistency (Sutherland et al. 2004). Thus, it is difficult to extrapolate population trends for snowy plovers at Salt Plains NWR beyond 2013–2017 when annual surveys were conducted following a standardized protocol (Hensley et al. 2014; Heath 2019). Exacerbating the issue of consistent monitoring is detectability—it is accepted that habitat characteristics, crypsis, behavior, and other elements of sampling and survey variability prevent detection of all animals present during a survey (Thompson et al. 1998; Buckland et al. 2001). Therefore, nearly all surveys will underestimate how many birds are present (Thompson et al. 1998; Bibby et al. 2000). If the proper detectability adjustments are made, issues with underestimating populations can be minimized. Incorporating detection probability into annual surveys of snowy plovers at Salt Plains NWR may help improve accuracy and precision of estimates and facilitate comparison of estimates between years (Buckland et al. 2001; Sutherland et al. 2004).
The status of Salt Plains NWR as important stopover and breeding habitat for snowy plovers (Skagen et al. 1999; Morrison et al. 2006; Thomas et al. 2012), combined with documented declines of other SGP populations (Saalfeld et al. 2013; Heath 2019) and possible connectivity with other snowy plover populations in the United States (Heath 2019), accentuates the need to regularly monitor this potentially at-risk population more closely (Hensley et al. 2014). Considering declines of neighboring snowy plover populations, and Salt Plains NWR's continental significance to breeding snowy plovers (Saalfeld et al. 2013; Heath 2019), estimates of snowy plover abundance on the Refuge require consideration of known biases. Further, monitoring annual changes in snowy plover abundance when evaluating the effectiveness of management efforts such as saltcedar removal requires precise estimates. Several studies on the SGP calculated or assumed detection probabilities of inland snowy plovers (Morrison et al. 2001, 2006; Pearson et al. 2007; Brindock 2009; Thomas et al. 2012; Saalfeld et al. 2013; Ellis et al. 2014), but no studies have incorporated distance sampling to verify the species' detectability. We predicted high detectability of snowy plovers because of findings of multiple studies, and previous experience surveying for the species. We also expected consistency between years in detectability and relative density. Our objective was to perform a preliminary study of distance sampling of snowy plovers at Salt Plains NWR to assess feasibility of survey methods, detectability of snowy plovers, and precision of distance sampling estimates.
Salt Plains NWR was established in 1930 and is located in north-central Oklahoma (USFWS 2002; Figure 1). The Refuge currently encompasses ∼12,900 ha, including nearly 5,000 ha of salt flat, salt marsh, ephemeral streams, some agriculture, a 4,000-ha reservoir, and a portion of the Salt Fork of the Arkansas River (USFWS 2002; Bonner 2008). The Great Salt Plains Reservoir, created in 1941, is shallow (average depth of 1.2 m) and slowly filling with sediment from the Salt Fork (USFWS 2002; Bonner 2008). Salt Plains NWR is an important stopover site for migrating shorebirds, waterfowl, and sandhill cranes Antigone canadensis, and an important breeding area for snowy plovers, American avocets Recurvirostra americana, and interior least terns Sternula antillarum (USFWS 2006; Thomas et al. 2012).
We conducted distance sampling surveys weekly at Salt Plains NWR from 18 July to 8 August in 2017 and 12 June to 31 July in 2018, which we designed to augment protocols used during previous annual snowy plover surveys (Hensley et al. 2014; Heath 2019). We conducted surveys during this time period because of observer availability. Prior to surveys, we overlaid the entire salt flat area of Salt Plains NWR with a 300 × 300 m2 grid using aerial imagery (Figure 2). The grid was instated by Thomas et al. (2012) and subsequently used for annual Refuge surveys (Hensley et al. 2014). We created three regions designated as “North,” “Middle,” and “South,” with each region consisting of 75 grid cells (Figure 2). We selected these three regions for their accessibility to all-terrain vehicles (certain areas are impassable because of the presence of creeks flowing into the reservoir) and known regular occurrence of snowy plovers. This also ensured that two observers would be able to access all grid cells on all-terrain vehicles before noon and repeat these surveys weekly during both survey years. We randomly selected only four grid cells in each region (n = 12) for use in this study because of logistical time-constraints. We surveyed each grid cell once weekly following procedures in Hensley et al. (2014) and Thomas et al. (2012). We created a weekly survey schedule a priori to avoid spatiotemporal biases associated with surveying the same regions and same grids at the same time and day each week. To achieve this, we randomly sampled regions as well as the grid cells within those regions each day.
Before we conducted surveys, we calibrated each observer's pace to improve consistency following survey protocols used in Thomas et al. (2012) and for annual Refuge surveys. Calibration involved each observer counting their steps for a 75-m distance, repeatedly, so that each observer had a good approximation of how many of their steps equated to 75 m. In addition, we calibrated observer distance estimation each year prior to initiating any surveys to minimize biases and improve accuracy. This involved each observer being tested repeatedly on the distance of small objects (up to 75 m) on salt flats until a reasonable level of accuracy was obtained. We estimated distance (m) of snowy plovers perpendicular from transects.
To begin a survey, two observers approached a grid cell without traveling through the cell to avoid displacing snowy plovers. Starting at the northwest corner, the two observers walked quickly due east. The first observer stopped after 75 m; the second observer stopped after walking another 150 m. Both observers then turned due south and began walking at the same time and at a steady pace, counting all snowy plovers within 75 m of either side of their (belt) transect with no overlap with the other observer. Also, observers made effort to record snowy plovers at their initial location because it is common for snowy plovers to be attracted to observers. This effort was aided by high visibility of snowy plovers on the salt flat. Observers paid specific attention to not count birds outside of this area, particularly in the other observers' area. Snowy plovers that moved from one observer's area to another were not counted by the second observer (this required extreme attention and verbal communication during and after the survey was completed). Observers stopped every 75 m to help calibrate the survey pace and scan the next survey section. After walking their 300-m transect, observers ended their surveys at approximately the same time. This survey design was intended to be identical to annual Refuge surveys with the exception of collecting distance data to make future incorporation of distance sampling easier.
We made an effort to avoid rounding or binning distance observations during the survey (a process known as “heaping”) to prevent the effect of rounding errors (Buckland et al. 2001). We did not use rangefinders because we assumed that taking individual readings on each bird would slow down the survey and prevent a “snapshot” view of the grid cell. This in turn would violate a basic assumption for distance sampling that objects are detected in their initial position (Buckland et al. 2001). One observer was constant among both seasons; the second observer differed between seasons.
For weekly snowy plover surveys conducted in 2017 and 2018, we analyzed survey data using Program R (version 4.0.0, R Core Team 2020; Table S1, Supplemental Material). We fit detection functions using the ‘ds' function in the ‘Distance' package (version 7.2; Miller et al. 2019). We modeled detection functions for 2017, 2018, and both years combined using perpendicular distance of snowy plovers and covariates ‘duration' (length of individual surveys), ‘region' (North, Middle, or South), and ‘observer.' We also included Uniform, Hazard-rate, and Half-normal key functions in all model combinations for a total of 45 a priori models (Table 1). We ranked models by Akaike's Information Criterion (AIC) and evaluated using ΔAIC of <2 (Akaike 1987; Burnham and Anderson 2004). We used no adjustment terms (cosine, hermite, or polynomial) except in models using Uniform key functions, which required two cosine adjustment terms. Using the ‘dht' function, we produced abundance estimates of snowy plovers at Salt Plains NWR from the top-ranked model or model-averaged density-detection-function models where appropriate. We produced abundance estimates for each year and both years combined at the region level and for all three regions combined (pooled). For an extrapolated abundance estimate for the entire salt flat area, we multiplied the pooled density estimate by 6,012 (668 total grid cells, 9 ha/grid cell). We considered abundance estimates significantly different if 95% confidence intervals of model estimates did not overlap. We confirmed model fit using the Cramér-von Mises goodness-of-fit test with the ‘gof.ds' function (Stephens 1986; Table 2). We present the three top-ranked models for 2017, 2018, and both years combined.
Three grid cells out of the 12 originally chosen to sample occurred in areas dominated by surface water or in nonhabitat (e.g., extensive water coverage, large patches of saltcedar, or other vegetation) and were removed from the surveys. We did not replace these because it required ∼5 h for travel and actual survey time to perform surveys in the remaining 9 cells. There were two competing (ΔAIC <2) models for 2017 and 2018 combined, two competing models for 2017, and four for 2018 (Table 1). Confidence intervals for detection probability, density, and abundance in 2017 and 2018 overlapped (Tables 2, 3). Therefore, the following reported values of detection probability, density, and abundance estimates are model averages from the three top competing models for 2017 and 2018 combined. Mean snowy plover detection probability (± standard error; SE) was 0.80 ± 0.04 (Table 2, Figure 3). We calculated snowy plover density (± SE) and abundance (± SE) estimates individually for each survey region of the salt flat (Middle, North, and South), and then combined for all three regions, or “pooled” (Table 3). Point estimates of density (individuals/ha) were greatest in the Middle (0.69 ± 0.15) and North (0.50 ± 0.06) regions, with abundance estimates of 462 ± 100 and 336 ± 38 individuals, respectively (Table 3). Lowest density occurred in the South region (0.48 ± 0.16) with abundance estimates of 324 ± 105 (Table 3). Overall, density and abundance estimates for pooled regions for both years combined were 0.55 ± 0.08 and 1,123 ± 151 (Table 3). Abundance across the entire salt flat, extrapolated from pooled density of both 2017 and 2018 combined, was 3,307 ± 421. Beta parameter estimates revealed a slight positive effect of duration on detectability (0.05 ± 0.03), while birds were less detectable in the North (−0.49 ± 0.33) and South (−0.73 ± 0.35) regions. However, each confidence interval except South and Middle regions overlapped zero and thus are likely uninformative (Arnold 2010). Two observers had both negative (−0.35 ± 0.42) and positive (0.66 ± 0.33) effects, while two others had too few observations to have a discernible effect.
In the preceding two decades, there have been multiple efforts to either estimate snowy plover population size or determine long-term population trends on the SGP (Morrison et al. 2001, 2006; Gorman and Haig 2002; Andres et al. 2012; Thomas et al. 2012; Saalfeld et al. 2013; Heath 2019). To date, however, there have been no known attempts to directly quantify detectability of snowy plovers using distance sampling. On the Southern High Plains of Texas, both Saalfeld et al. (2013) and Thomas et al. (2012) reported high detection rates for snowy plovers. Saalfeld et al. (2013) attributed this to lack of vegetation and the species' behavior (easily spotted movements and displays in response to perceived danger to nests). These claims were also supported by consistent survey counts (±2 individuals) at saline lakes with low populations (∼20 individuals, Saalfeld et al. 2013). High detection rates at Salt Plains NWR (0.80 ± 0.04) during the 2017 and 2018 breeding seasons support the claims of prior snowy plover studies, which either assumed (Saalfeld et al. 2013; Brindock 2009) or calculated high (≥0.74, Pearson et al. 2007; Thomas et al. 2012; Ellis et al. 2014) probability of detection of snowy plovers. One exception, a study by Hood and Dinsmore (2007) in the southern Laguna Madre region of Texas, reported a detection probability of 0.58, markedly less than this study and others (see above). Our study corroborates, at least in part, precision of estimates, inferences made, and methods used by previous studies and also supports long-term population trend data reported for snowy plovers in the SGP of Texas (Saalfeld et al. 2013; Heath 2019). As predicted, detection probability did not vary between years.
Density and abundance estimates did not vary significantly between years; therefore, we focused on model-averaged estimates created from the top models for both years combined. The snowy plover density estimate (0.55 individuals/ha) is fewer than expected when compared with 0.76 individuals/ha suggested by Refuge surveys (Heath 2019) and 0.88 individuals/ha calculated from Thomas et al. (2012) abundance estimates. If the global estimate for 2017 and 2018 combined (0.55 individuals/ha) is extrapolated to the entire salt flat, represented by 668 grid cells (not just the three regions sampled, a total of 225 grid cells), the estimated abundance of snowy plovers is 3,307 ± 421 individuals. The estimated abundance in this study is less than the 5,280 individuals estimated by Thomas et al. (2012) and most annual estimates calculated by annual Refuge surveys (3,086–5,280 individuals; Heath 2019) but more than 1,964 individuals by Gorman and Haig (2002), and 2,105 individuals for all of the Great Plains (excluding Great Salt Lake) by Morrison et al. (2006). We do not believe a true comparison of abundance over time is warranted between these studies and ours because of disparities in methodological techniques and difficulties extrapolating density estimates across a heterogeneous landscape (i.e., the sampled areas in this study likely do not accurately represent a true heterogeneous distribution because sampled regions were chosen for known occurrence of snowy plovers); however, this abundance estimate remains noteworthy because it indicates a similar magnitude of snowy plover individuals at Salt Plains NWR to recent studies (Thomas et al. 2012; Heath 2019) and lends credence to efforts to protect and improve this habitat. It may also indicate a decrease in the snowy plover population, which warrants further and more robust scrutiny.
In addition to a positive sampling bias and difficulties extrapolating across a heterogeneous landscape, disparities between estimates may also be due to study design, timing of surveys, and analysis of this study and previous ones. Salt Plains NWR annual survey estimates extrapolate a mean number of birds observed per grid cell to the total number of grid cells, a commonly used method of random sampling (Thompson et al. 1998; Bibby et al. 2000; Sutherland et al. 2004), resulting in estimates of 3,086–5,280 individuals (Heath 2019). Consequently, this method did not fully consider distribution, detectability, or stratification of habitat. Thomas et al. (2012) used a stratified random sampling method and N-mixture modeling to estimate detectability and abundance, but based stratification on assumptions and expert opinion rather than known distribution or data. Ideally, a combination of these methods would result in greater accuracy and precision of annual surveys. For example, annual Refuge surveys may play an important role in future models by linking distribution data to high–medium–low habitat quality. The addition of distance sampling could further enhance accuracy.
We found distance sampling to be an effective tool to estimate density and abundance of snowy plovers at Salt Plains NWR. With appropriate training and distance calibration prior to surveys, estimating perpendicular distance of individual snowy plovers to transects was a reasonable addition to the standard methodology used in Refuge surveys. Further, our results suggest that incorporating distance sampling into annual surveys at Salt Plains NWR may improve accuracy and precision of abundance estimates given a detection probability of 0.80 ± 0.04 and a relatively low SE of abundance (±421 individuals). A detection probability of 0.80, in theory, suggests there were 20% more birds present than were observed and that accuracy of abundance estimates may be increased accordingly, although it is important to note that because of the timing of our surveys, this detectability may not be an accurate representation of detectability earlier in the breeding season when annual surveys are conducted. However, considering this and considering a suitably low SE, distance sampling may be a method to preserve accuracy and precision while requiring fewer observers during annual surveys (i.e., if fewer samples are required for the desired accuracy and precision), although further investigation is warranted.
For future estimations of snowy plover abundance at Salt Plains NWR and subsequent population trend analysis, we recommend that annual surveys incorporate distance sampling to test its effect on accuracy and precision of annual estimates. We also recommend that future analyses consider the nonindependence of simultaneous surveys used in annual Refuge surveys. Outside of distance sampling, stratification of habitat quality (high–medium–low) using annual survey data may improve future N-mixture models like those in Thomas et al. (2012). A comparison of results from the different methods (stratified random sampling with N-mixture models versus distance sampling) might also clarify their suitability to precisely and efficiently estimate snowy plover abundance at Salt Plains NWR. This in turn will aid in monitoring the population's response to conservation management efforts such as saltcedar removal as well as potentially aid the management of snowy plover habitat elsewhere in North America
Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.
Table S1. Distance sampling observations of snowy plovers Charadrius nivosus in 2017 and 2018 at Salt Plains National Wildlife Refuge. Each line represents a snowy plover observed at a perpendicular distance from the transect, where “NA”s recorded in the “Distance” column indicate that no snowy plover individuals were seen at that transect on that day.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S1 (92 KB DOCX).
Reference S1. Hensley G, Schmidt P, Johnson B, Smart B, McDowell K, Metzger K, Harris G, Archibeque A. 2014. Inventory and monitoring plan for Salt Plains National Wildlife Refuge. Jet, Oklahoma.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S2 (535 KB PDF).
Reference S2. Pearson SF, Sundstrom C, Brennan K, Fernandez M. 2007. Snowy plover distribution, abundance, and reproductive success: 2006 research progress report. Olympia: Washington Department of Fish and Wildlife, Wildlife Science Division.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S3 (814 KB PDF).
Reference S3. Skagen SK, Sharpe PB, Walermire RG, Dillon, MB. 1999. Biogeographical profiles of shorebird migration in midcontinental North America. Biological Science Report. USGS/BRD/BSR–2000-0003. Denver: U.S. Government Printing Office.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S4 (5.37 MB PDF).
Reference S4.[USFWS] United States Fish and Wildlife Service. 2002. Contaminant investigations at Salt Plains National Wildlife Refuge including an assessment of confined animal feeding operations. Tulsa, Oklahoma: U.S. Department of the Interior, Fish and Wildlife Service, Region 2.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S5 (6.44 MB PDF).
Reference S5.[USFWS] United States Fish and Wildlife Service. 2006. Salt Plains National Wildlife Refuge: comprehensive conservation plan. Jet, Oklahoma.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S6 (39.65 MB PDF).
Reference S6.[USFWS] United States Fish and Wildlife Service. 2007. Western snowy plover (Charadrius alexandrinus nivosus) Pacific coast population recovery plan. In 2 volumes. Sacramento, California.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S7 (21.85 MB PDF).
Reference S7.[USFWS] United States Fish and Wildlife Service. 2017. Region 2 Southwest biological priorities refuges. Albuquerque, New Mexico: U.S. Department of the Interior, Fish and Wildlife Service.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S8 (3.2 MB PDF).
Reference S8.[USGCRP] United States Global Change Research Program. 2009. Global climate change impacts in the United States. New York: Cambridge University Press.
Found at DOI: https://doi.org/10.3996/JFWM-20-041.S9 (13.41 MB PDF).
We thank all Refuge staff at Salt Plains NWR for assistance, shared equipment, and lodging. Thanks to all technicians (C. Olivas, D. Spadafora, S. Melendez, M. Rainey, R. Bracken), students, private landowners, and collaborators that have helped with field and lab work during the past 10 y. Thank you, M. Acre, for advice and edits. Also, we would like to thank the reviewers and Journal Editors for their thoughtful edits and recommendations. Financial and logistical support was provided by the U.S. Fish and Wildlife Service, Texas Parks and Wildlife Department, the U.S. Geological Survey, The Rumsey Research and Development Fund, and the Bricker Endowment for Wildlife Management.
Any use of trade, product, website, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.
Citation: Heath-Acre KM, Conway WC, Boal CW, Collins DP, Hensley G, Johnson WP, Schmidt PM. 2021. Detectability and abundance of snowy plovers at Salt Plains National Wildlife Refuge, Oklahoma. Journal of Fish and Wildlife Management 12(1):50–60; e1944-687X. https://doi.org/10.3996/JFWM-20-041