Ecological and Economic Benefit-Cost Comparison of Grazed and Ungrazed Prairie Land for Critical Species Protection in Western Washington

Seeded Species Establishment

Out of the 10 species we seeded into CGP treatments, 5 successfully established in at least one CGP site by spring 2020 (as captured in our 1m² quadrats): Castilleja levisecta (golden paintbrush), Collinsia parviflora (maiden blue-eyed Mary), Lupinus bicolor (bicolor lupine), Plectritis congesta (sea blush), Ranunculus occidentalis (western buttercup) (Figure 14). Presence of Collinsia parviflora and Lupinus bicolor in the BAU treatments was due to the fact they were already established at sites before seeding occurred (Table 6). The persistence of the annual species 2 years post-seeding shows that they are reproducing on their own in the CGP units. It is also worth noting that due to the patchy nature of establishment for some of the seeded species at the grazed sites, our monitoring quadrats did not capture their presence, despite large established patches in the paddocks. This is especially true for Cerastium arvense (field chickweed), which we found at both Fisher Ranch and Colvin Ranch (Figure 14).

Table 6. Mean abundance of five of the native seeded species (± 1SD) across the different sites and the three treatments (n=6).  Absolute abundance is quantified as the number of individuals per 1m² monitoring quadrat. R. occidentalis was already present in large quantities at Colvin Ranch so it was not seeded, nor was it monitored there. This is represented by ‘N/A’.

Native Species Richness

While the mean native species richness is still 4-5 times greater in the NUP than in either grazing treatment, the CGP treatments significantly increased native species richness over the BAU treatment within the first year of implementation (Figure 15, Table 7). Native species richness remained higher in the CGP treatments through 2020, despite no additional seeding in fall 2019.  Sowing over multiple years and increasing seeding rates would likely further boost the establishment and persistence of native species.   

Increased native species richness was observed within the CGP treatments at all ranch sites: Colvin gained five species, Fisher gained three species, while Riverbend gained one species (Figure 16). The increase in richness was due to the seeding of native species, in particular Plectritis congesta, Collinsia grandiflora, and Ranunculus occidentalis (Table 6).  Compared to native ungrazed prairies, native species richness at ranch sites was much lower (2-12 species on average at ranch sites compared to 15-23 species on average in NUPs).

Table 7. Test statistics and p-values calculated from main effects and interactions within fitted models just between CGP and BAU treatments. Wald tests were used to calculate chi-square (c²) statistics and p-values for main effects. Interactions between year and treatment were calculated from Wald Z-tests for native forb cover and native species richness (GLMMs) and Wald t-test for non-native species richness (standard LMM). P-values in bold are considered significant at α=0.05.

Percent cover of native forbs

Two years after the implementation of conservation grazing practices (CGP), native forb cover showed minimal change within the CGP treatment and decreased slightly in the BAU treatment (dropping from 4% to 1%), though the effect of treatment was not significant (Figure 17, Table 7). Higher seeding rates of species that successfully established (Table 4) would likely have increased native forb cover in CGP treatments. Ranch sites varied in native forb cover, but all sites were <5% (Figure 18). In contrast, native forb cover was 2-3 times higher in native ungrazed prairies and fluctuated more over time.

Non-native Species Richness

Non-native species richness showed minimal change from 2018 to 2019 but increased significantly in the CGP treatment in 2020 (Figure 19, Table 7). Both BAU and NUP treatments showed a slight increase in non-native species richness over time (~1.7 species on average for each treatment), but the CGP treatment gained an average of 6.2 species over the three years of the study. This surge was largely driven by increased richness at Fisher Ranch and Riverbend Farm (Figure 20), due to a mix of annual forbs and perennial grasses: Holcus lanatus, Medicago lupulina, Agrostis capillaris, Crepis capillaris, and Elymus repens.

Plant community composition over time

To visualize changes in plant community composition over time, we used non-metric multidimensional scaling (NMDS) ordination. This method clusters communities based on similarity so that assemblages that are more similar in species composition are closer together while those with disparate compositions are farther apart.

Overall, species composition across all plots became more similar from 2018 to 2020, as indicated by tightening of the plots across both Axis 1 and Axis 2 (Figure 21). Subsequent similarity percentage analysis (SIMPER) showed that Hypochaeris radicata, Plantago lanceolata, and Rumex acetosella increased in frequency across monitoring plots between 2018 and 2020, leading to increased similarity in composition. The native ungrazed prairie (NUP) sites all clustered together, reflecting similarity in composition. The ranch sites also clustered together more over time, with Fisher Ranch and Colvin Ranch hosting several plots with similar plant community compositions. Additionally, plant communities in the CGPs (blue shapes) have tightened along Axis 2, representing the native seeded species that have become established there.

Forage percent cover

Forage cover was high at all ranch sites (>85%, Figure 23) with no detectable difference between the CGP & BAU treatments (p=0.9; Figure 24) two years after conservation grazing practices were implemented. Not surprisingly, forage cover was significantly lower on native ungrazed prairies (p<0.001; Figure 24).   

Forage Cover Species Response to Conservation Grazing

Implementation of a rotational grazing system at one grazing site (Riverbend Ranch) afforded the opportunity to monitor forage cover responses to an introduced planned grazing regime. Cattle provided free and continuous access to fields are reputed to consume more tender, palatable species over time, resulting in stands dominated by short-lived, less palatable, and quick-reproducing species. These are presumed to consist of weedy annuals and perennials, and undesirable perennials such as Canada thistle (unpalatable), sub-terranean clover (smothers other desirable species), or other. By contrast, generating higher stocking rates on limited-area paddocks for short periods of time is understood to modify behavior by pressuring grazing livestock to forage less discriminately out of a degree of competition with the rest of the herd. In this forage cover sub-trial we evaluated the effect of CGP and BAU treatments on a variety of forage cover metrics including perennial cover and percent grass cover.

Inherent pasture heterogeneity across the grazing landscape reflected considerable micro-scale variability with the result that random effects from replication in many cases masked potential fixed effects of treatments. When all six replications for the experiment was analyzed, vegetation differences (percent grass cover and percent preferred grasses versus less preferred grasses) between the CGP and BAU treatments failed to rise to the level of significance (Table 8). Yet amidst significant vegetative variability between replications, emergent differences appeared to barely fail to rise to the level of significance, with the suggestion that vegetation dynamics appeared to be shifting to more preferred grass species after three years in an altered grazing regime in CGP.

To further explore this effect, several subsets of the full set of six replications were analyzed in an attempt to sift out treatment effects from swings in landscape patchiness (Table 9). This was done by running Student’s t-tests on uniform subsets of replications (i.e. 1-4 or 1-3), or by excluding the two replications with the highest percent cover preferred species from each treatment (which were not the same rep for each treatment). This amounted to excluding replication data where a handful of individual large plants dominated the count (i.e. high microsite variability). This patchiness on the landscape resulted in large swings in both CGP and BAU (note standard errors, Table 9). Excluding an even number of reps in each treatment where these large variations were observed (i.e. two BAU and two CGP treatments) allowed for a comparison of the remaining four reps with fewer outlying values.

We report this data because the emergence of trends in forage cover amidst considerable landscape heterogeneity appeared indicative of developing shifts in the forage community that were visibly apparent, particularly in persistent green cover in CGP treatments longer into dry summer months.

Table 8. Effect of Treatment on Forage Cover Metrics, All Reps

Alpha values of less than 0.05 for Prob > t indicates probability that mean value of CGP is greater than BAU. Italicized bolded p-value denotes narrowly insignificant values.

Table 9. Effect of Treatment and Replication Variability on More Preferred Grass Cover (%)

Alpha values of less than 0.05 for Prob > t indicates probability that mean value of CGP is greater than BAU.

Gopher Occupancy

Gopher occupancy, measured as the proportion of plots with fresh gopher mounds present, increased over the project period at all sites except Johnson prairie. The greatest increase occurred in the CGP treatment (56% occupied in 2018 to 83% occupied in 2020; Figure 22). Interestingly, treatment was not a significant factor for predicting gopher occupancy in the fitted model (X²=1.8; p=0.41), suggesting that gophers can use habitat on grazed working lands at levels comparable to protected prairie preserves. 

Forage biomass

Biomass production ranged from 0.49 to 1.76 (2019) and 1.54 to 2.5 (2020) tons per acre for ranch sites, and from 0.53 to 0.61 (2019) and 0.48 to 0.58 (2020) tons per acre for upland prairie (Figures 25 and 26). The highlights of this work indicate that:

• Biomass production in all years was generally the highest and not significantly different across CGP and BAU treatments at the Colvin and Riverbend Ranch sites (Figures 25 and 26).
• CGP practices did not depress overall forage production and thereby native seeding is unlikely to negatively affect forage production in grazed systems.
• Significantly lower forage production was observed at Fisher Ranch, likely due to a combination of high stocking rate, shorter and inconsistent rest periods (data not shown), and lower soil fertility (Tables 15, 16, and 17).
• Taller stubble height, and more regular rest periods resulted in greater forage production longer into the summer forage dormancy period in CGP as compared to BAU treatments (Figure 27).
• Early spring grazing prior to the deferment period effectively utilized spring forage availability in paddocks at Colvin Ranch, while lack of early spring grazing at Riverbend Ranch lowered overall utilization (wasted forage, Figure 28) by allowing grasses to become overly tall and mature during the grazing deferment.

Forage production at Riverbend and Colvin ranches were not significantly different. Due to higher stocking rates, Fisher Ranch was generally utilized fewer and shorter rest periods between grazing rotations, detected in significant cage/no-cage grazing differences during rest periods (data not shown). This resulted in lower biomass production (Figures 25 and 26). Additionally, low productivity at Fisher Farm is very likely also linked to lower phosphorus, potassium, calcium, magnesium, cation exchange capacity, and pH levels at this site (Tables 15, 16, and 17).

A notable dynamic in CGP treatments was the importance of utilizing spring forage prior to deferment. Missing this utilization as occurred at Riverbend depressed overall forage production below BAU at Riverbend, and lower than overall production as compared to Colvin Ranch generally. Nevertheless, CGP biomass totals were not statistically less than BAU totals, indicating no detrimental effect of seeding native species into pastures. In the future, improved spring biomass utilization in rotationally grazed fields at Riverbend combined with extended forage production into the summer may eventually yield greater forage productivity than continuously grazed fields.

Lower biomass production at the ungrazed prairie sites (Johnson West Rocky, and Wolf Haven) in relation to Riverbend, Colvin and Fisher may be due to lower nutrient levels at these sites, in particular phosphorus, potassium, calcium, iron, and cation exchange capacity (Tables 15, 16, and 17). Another factor may be soil moisture. Forage and soils work in 2020 included forage quality assessment, soil compaction and soil temperatures, to evaluate the potential impacts of grazing generally, conservation grazing in particular, and lack of grazing on these forage quality and soil health parameters (see below).

In 2020 an expanded forage evaluation was developed to investigate the relationships between forage height, soil temperatures, and the potential to extend forage production further into the dry summer period when forage typically enters dormancy. Forage biomass aspects of this work are reported here and below under the section titled “Soil Quality Parameters”. Generally, CGP paddocks did exhibit the capacity to produce more, longer into the summer than BAU, as illustrated in significant differences in July biomass measured under exclusion cages in the two treatments (Figure 27).

Forage Utilization

A take-half/leave-half approach to forage utilization is generally encouraged in rotational grazing systems. While it requires ranchers to forego usage, leaving forage allows for greater biomass productivity as a result. Percent forage utilization provides an indication of the extent to which this strategy is implemented by different ranch operations. As noted in the Methods section, utilization estimates were difficult to obtain due to considerable variability and differences in utilization data collection methods and calculations between rotationally and continuously grazed sites.

In 2020 forage utilization was apparently highest at Fisher and in BAU at Riverbend (Figure 28). Riverbend BAU and both Fisher treatments tended to exhibit very low stubble height (data not shown), high grazing pressure, and consequently high utilization rates. The 70-84% utilization rates in these systems effectively prevent robust root establishment and elevate daily average soil temperatures (Figures 35 and 36), which can delay regrowth in the spring and fall and lead to earlier forage dormancy in the summer dry season (Figure 27).

Relatively low utilization (47-50%) in Riverbend CGP reflected a better approach to conserving stubble, but also June/July forage trampling due to missed grazing prior to deferment. Given similar forage trampling was observed at Colvin Ranch, the higher utilization (62-73%) there in 2019 was unexpected. Utilization at Colvin in 2020 was only obtained for the spring rest period due to missing post-graze data. The lower March forage utilization rates (34-51%) reflect a modest and strategic use going into deferment.

Butterfly Movement and Behavior 

In Year 1, we observed flight paths for 11 female and 27 male silvery blue butterflies. We observed six female and 23 male ochre ringlet flight paths. Year 1 silvery blue female diffusion rates were lower in native habitat than conservation or conventional grazing but male movement behavior did not show a detectable difference (Figure 29a). Female silvery blue butterfly diffusion rates indicated concentrated search patterns exhibited in areas with high reward. Ochre ringlets did not exhibit a trend in diffusion rates across management types, regardless of sex (Figure 29b).

In Year 2, we observed 122 butterfly flight paths throughout the season (Table 10). Female ochre ringlets are more difficult to observe as they are skittish and sedentary. This limited the number of observations that could be completed. In the late season, weather limited progress, as it was unusually cloudy and extremely windy in the afternoons.

Table 10. The total number of observations in Year 2 per site separated by species and sex. Ochre ringlets are further separated by the flight period in which the observation was collected. The early flight ran May-July and the late flight ran July-September. Silvery blue butterflies have only one flight period (May), so the data are not separated by flight period.

Question 1

We did not detect significant differences in silvery blue butterfly diffusion rates between management types with a GLMM (Table 11; Figure 30). We detected a nearly significant effect of the Native Upland Prairie management type on silvery blue butterfly diffusion rates with a GLM (Table 11; Figure 30). The estimated regression coefficient for Native Upland Prairie was negative, indicating that diffusion rates are lower in prairie habitat.

Question 2

We observed 40 total flowering plant species in bloom across the six sites throughout the silvery blue butterfly flight season (Appendix C). For ease of analysis, here we include only the 11 flowering plants that we observed silvery blue butterflies to nectar on (Figure 31a). We observed four host plant species across the six sites (Figure 31b).

The a priori GLM with host plants and sex as explanatory variables found a significant effect of Lupinus spp. volume and Vicia hirsuta volume on silvery blue butterfly diffusion rates (Table 12). The estimated regression coefficients for both Lupinus spp. and Vicia hirsuta were negative, indicating that greater volumes of those plants are associated with lower diffusion rates. There were no significant effects of nectar availability, vegetation height, or sex, but there was a significant effect of total host plant volume (Table 13).

Female silvery blue butterflies only oviposited on Lupinus spp. host plants but did oviposit in a grazed site (MA; Figure 32). We observed only one instance, which was not during a formal observation, of a butterfly walking on a Vicia sativa plant, indicating that the individual was considering oviposition, but she left the plant before doing so. Our next steps are to continue these analyses with the ochre ringlet diffusion rates and behavior.

Table 12: An a priori model of mean host plant volume per path effects on silvery blue butterfly diffusion rates.

Table 13: An a priori model with total host and nectar availability and vegetation height on silvery blue butterfly diffusion rates.

Butterfly Preference

In 2020, we observed 52 total flight paths (Table 14). Chi-square tests indicated significant differences in the ending locations based on the starting location in silvery blue males and females and ochre ringlet males (p < 0.05; Figure 33). Post-hoc tests indicated that the significant differences were between the BAU and CGP release locations, which showed that individuals did not tend to move far enough to cross borders and were not attracted to CGP or BAU. There were no significant differences in path ending locations at the border release locations, indicating that individuals did not exhibit a preference for BAU or CGP.

Our next steps are to follow similar methods to the behavior/movement experiment to calculate the direction of movement, as well as flight path parameters (movement steps and turning angles) to quantify habitat-specific diffusion rates and compare them between BAU and CGP treatments, as well as to compare movement and movement directions at the border release locations.

Table 14: Number of successful behavior observations per release location by species and sex in 2020.

Butterfly results discussion

Greater host plant availability (especially large perennial Lupinus spp.) on a flight path level results in slower silvery blue butterfly diffusion rates, indicating that individuals perceive habitat with host plants as higher value habitat. Lupinus spp. host plants are often considered to be undesirable by farmers and ranchers as they are toxic to cattle and may cause pregnant cows to abort their fetuses (Panter et al. 2002). However, cattle generally will not eat Lupinus spp. unless they have no other choice (Stephen Bramwell, pers. comm), and we observed cattle to clip individual grass stems from under Lupinus plants without touching the plant itself (pers. observation). This observation fits with other studies that found cattle to promote greater plant community diversity and more nectar and host plant resources by reducing competitively dominant grass cover through foraging preferences (Wang et al. 2007; Zakkak et al. 2014; Delaney et al. 2016).

Silvery blue butterflies are common, relatively generalist species, so from their perspective, there may not be strong differences between the sites and management types apart from host plant availability. Even our “conventional” grazing pastures did not cover the full range of potentially negative grazing practices in the region, as producers willing to allow access to their land for butterfly experiments were already interested in conservation and were involved with the WSARE project. Two pastures in the behavior and movement experiment were certified organic farms (MA and MY), and the other two, regardless of grazing strategy, use little to no pesticides, which is not the case at all farms and ranches in the region. It is possible that we may have seen more of a contrast in behavior and movement rates if we had access to pastures that were more destructively grazed or used pesticides. In addition, silvery blue butterflies may be responding to an unmeasured or unanalyzed variable(s) in our system such as immediate weather conditions or microclimate variables.

Female silvery blue butterflies oviposited on lupines in grazed pastures, indicating that they recognize grazed pastures as habitat if their Lupinus spp. host plants are present. If butterflies will recognize grazed pastures as habitat if their host plants or other resources are present, then there is potential for grazed land to contribute to butterfly habitat in the landscape. Therefore, to support a wide range of butterfly species, grazing practices that support the greatest diversity of host plants should be encouraged, such as low intensity grazing with compensation to producers for lost income (Bussan 2022, Pöyry 2004, 2005, Weiss 1999). This conclusion does not apply to all butterfly species or grazed habitats. While we were fortunate to be allowed access to working farms and ranches, we were restricted to working with common generalist species, like silvery blue butterflies. Sensitive species may be more discerning in their behavior, movement, and habitat perceptions and may not respond to grazing in the same way as generalist species (Henry et al. 2019, Murphy et al. 2011).

Our butterfly preference experiment did not find evidence of preference for one treatment over another. Individuals of both species and of both sexes in silvery blue butterflies tended to stay within the treatment in which they were released. The slightly higher number of individuals who left CGP treatments may have been because of the smaller size of CGP paddocks; there was a higher chance of encountering a border. Additionally, there may not have been enough time since the beginning of the experiment to observe a strong enough signal from the plant community (Figure 15, 17) to be reflected in butterfly preferences and behavior in CGP and BAU treatments.

Soil Quality Parameters

Forage Heights, Soil Parameters and Soil Taxonomic Work

Soil nutrient status varied considerably between sites (Table 15), as well as by management type, at the outset of the study (Fall 2018). Soil nutrient levels tended to be generally higher under grazed as compared to ungrazed management, as observed in significantly (p < 0.05, Students t-test) higher levels of phosphorus, potassium, calcium, and magnesium in grazed systems (data analysis not shown). Sulfur levels were higher under ungrazed management, and pH, cation exchange capacity levels were higher under grazed management. No differences were observed in N0₃-N, or organic matter levels.

Table 15. Soils Nutrient Data Collection in 2018 as Expressed as the Sample Mean ± 1 Standard Deviation.

Table 16. Effect of Management and Experimental Treatments on Soil Fertility Parameters at Riverbend Ranch, Colvin Ranch, Fisher Ranch, Johnson Prairie, and Wolf Haven Prairie..

*Bold font indicates significantly different pairs (p < 0.05, t-test).

Differences soil nutrient parameters also varied significantly by site (Table 17). Nutrient levels were generally highest at Riverbend and Colvin Ranch, with soil nutrient parameters for Fisher Ranch often clustered with the ungrazed prairie sites (Johnson Prairie and Wolf Haven prairie).

Table 17. Effect of Site on Soil Fertility Parameters at Riverbend Ranch, Colvin Ranch, Fisher Ranch, Johnson Prairie, and Wolf Haven Prairie.

RB = Riverbend, CR = Colvin Ranch, FR = Fisher Ranch, JP = Johnson Prairie, WH = Wolf Haven Prairie

Non-overlapping letters denote significant differences between site (p < 0.05, Fisher’s LSD). 

Soil Temperature, Bulk Density and Forage Dynamics

Grazing treatments (CGP and BAU) and site management (grazing as compared to no grazing) were evaluated for their effect on residual biomass (i.e. stubble height) and soil parameters including soil bulk density and soil temperature. Stubble height and soil temperatures are illustrated in Figures 33 and 34. The goal of this evaluation was to understand potential impacts of stubble height on soil temperature, both from the point of view of climate resilient grazing systems, and for the potential of conservation grazing management to extend forage production into the warm dormant season by means of cooler soil temperature. Significantly different stubble heights were observed between sites (West Rocky, Riverbend) and grazing treatments (CGP = 5.72 in, BAU = 2.00 in; p < 0.05, Student’s t-test).

Significant differences in soil temperatures were also observed (Figures 35 and 36) between sites and treatments. Soil temperatures were highest at West Rocky Prairie and lowest in Riverbend CGP plots (highest stubble height), with soil temperatures in Riverbend BAU plots in the middle. Mean soil temperatures were 69.5 F (West Rocky), 67.4 F (Riverbend BAU), and 65.1 F (Riverbend CGP). Grazing events in which forage height was reduced appear to have led to temperature increases in CGP plots in relation to West Rocky.

Higher stubble height management, and consequent shading effects, may have a beneficial impact on forage growth by decreasing summer average daily high soil temperatures. Significantly greater July biomass observed in CGP as compared to BAU cage samples (see Figure 26) illustrate the mitigating effect of higher forage heights on soil temperatures under CGP management.

Soil Bulk Density

Soil bulk density vales were obtained from CGP and BAU treatments at Riverbend Ranch. Mean bulk density in CGP and BAU were 1.14 g/cm³ and 1.15 g/cm³, respectively, and were not significantly different (data not shown). Typical bulk density values for moderately grazed sites are reported to be 0.9 g/cm³ (Engels n.d), indicating Riverbend sites (BAU as well as CGP) with a history of continuous grazing exhibit soil compaction. While we were not able to demonstrate that pasture rest associated with the CGP treatment decreased bulk density, it is possible that the duration of the study was insufficient for shifts in soil bulk density to occur.

Soil taxonomic work

Regarding soil taxonomy, the majority of soil profiles at ranch and prairie preserve sites classified as either a Spanaway (Typic Humixercept) or Nisqually (Pachic Humixerept) series, or a higher-taxa of Pachic Humixercept (i.e., there is no existing soil series that would be a good fit). In general, the Spanaway-like soils were found in intermound areas, and the Nisqually-like soils were found on the mound sites that had a deeper, darker surface horizon(s). The higher-taxa Pachic Humixerepts, were described as Loamy-skeletal as compared to the sandy Nisqually, or sandy-skeletal Spanaway, meaning it had slightly more clay content within the control section. One outlier was a poorly drained, Norma soil found near a drainage ditch on one of the ranch research sites.

Based on these sites, a fairly clear pattern of deeper, more organic rich surface A horizons found on mounded areas and thinner A horizons found in the intermound position emerged. In addition, the mounded areas generally had either less rock fragment content by volume and/or smaller rock fragment diameter within the surface A horizons. In general, soil pit descriptions mostly conformed (though did not capture the higher taxa) to soil units mapped on NRCS Web Soil Survey (Table 18). Full soil pit descriptions are available upon request.

Table 18. Mapped soil series and soil series as verified by soil pit descriptions at ranch and prairie preserve research sites.

Survey of Grazing Practices in Southwest Washington

We received a total of 133 responses, 95 of which were from the emailed link and 38 from the printed surveys. Eighty-nine of the responses were from producers who have livestock and 39 were from producers who do not. Five respondents did not answer the question about whether they had livestock. As respondents could choose not to answer individual questions, response numbers per question varied. Twenty-two respondents owned or leased grazing land in Thurston County, 28 in Pierce County, 23 in Lewis County, and 13 in other counties. Full questionnaire results are available in the supplementary Survey Report.

Question 1: Producer perception of incentives and barriers to implementing conservation

Tax incentives and on-farm cost-share programs were the most preferred incentives (chi-squared = 64.024, p << 0.05), followed by conservation easements and transfer/purchase of development rights (Figure 1). Selling property for restoration/conservation or registry/branding certification programs were the least preferred (Figure 37).

“Loss of development rights if listed species found” and “Not enough financial assistance available” were the most highly rated barriers (Figure 36), followed closely by “Not enough technical assistance available” and “Conservation is too expensive” (chi-squared = 63.369, p-value << 0.05) Social stigma was the lowest rated barrier to implementing conservation grazing (Figure 38).

Question 2: Demographic and economic effects on producer perceptions of incentives

Respondent age and education level consistently played a strong role in determining incentive perception (Figure 39). Respondents in the oldest age category (70+ years old) tended to be less interested in all of the incentives than respondents in the 30-49 and 50-69 age categories. Respondents with associate, bachelor’s, or advanced degrees tended to be more interested in the incentives than those without education beyond high school. Farm size was important in perceptions of on-farm cost-share programs. Respondents in the largest farm category (300+ acres) were not interested in cost-share programs, while respondents in the next largest (100-299 acres) were interested (Figure 39).

Cow-Calf, Grass Finished Beef, and Restoration Enterprise Budgets

Completed enterprise budgets for both grazing operations and ungrazed restoration operations are available on the project website: https://extension.wsu.edu/thurston/agriculture/on-farm-conservation/prairie/. Annual operational costs on a 50-acre basis for habitat establishment on abandoned cropland, abandoned rangeland, and Scotch broom infested land were $15,399, $15,818, and $21,339, respectively. No income was derived from these operations so no profit-loss figures were generated.

Annual net returns over total costs for a 50-head cow-calf operation in South Puget Sound were ($29,737) for the herd and ($595) per head. Discounting labor and fixed cost recovery, as is not uncommon in the region resulted in annual net returns over total costs of $3,880 for the herd, and $78 per head. Operations assumed 2 acres per head. Growing steers out over the approximately 30-month period to produce grass finished beef resulted in annual net returns over total costs for a 50-head operation of $5,151 for the herd, and $215 per head when marketing as quarter, half, or whole through WSDA custom exempt slaughter and processing. Marketing as USDA-certified generated annual net returns over total costs of $32,497 for the herd, and $1,354 per head.

Economic Impact Analysis

This study quantified the economic impacts of five scenarios (Table 19) that represent varying utilization of working lands (as either cow-calf or grass-finished beef operations) for species protection in the Thurston County Habitat Conservation Plan (HCP).

Table 19. Model Scenarios Identifying Acreage Mixes of New Reserves and Working Lands Easements to Protect, Enhance and Manage Habitat for the Covered Species

¹New reserves starting point split evenly as “abandoned cropland (Abn-Crp), Abandoned rangeland (Abn-Rgn), and and Scotch broom (Sct-Brm)

²Working lands split 80:20 cow-calf and grass finished steer                                                

³Easements set up using grazing budget

Sales from the livestock sectors introduced ‘new’ dollars into the region through the sale of product outside of the region or substitution of locally produced product (like beef) for imported product. New Reserve (NR) sectors in the baseline model run did not introduce new dollars into the region because all funding for these sectors (new prairie preserves established on abandoned range, and abandoned cropland, or Scotch broom) were projected by Thurston County to be funded locally from mitigation fees (on developers) and other sources (i.e. Conservation Futures taxes). As a result, total economic impacts increased when more working lands were recruited into the program ($0 with no working lands in Scenario 1 to $2.09 million with 400 ac working lands in Scenario 2, and $7.83 million with 1,500 ac of working lands in Scenario 4; Table 20).

Table 20. Economic Impacts‡ of the Respective Scenarios with Baseline Assumptions

‡Figures reported in 1,000s

Assuming 25% of NR funding could somehow be sourced externally increased total economic impacts ($688 thousand in Scenario 1, $3.60 million in Scenario 2, and $11.65 million in Scenario 4; Table 21). Generally, we found that total economic impacts uniformly increased, and costs decreased, in scenarios where greater proportions of working lands easements (WLE) were engaged.

Table 21. Economic impacts‡ the respective scenarios with the 25% exogenous restoration funding assumptions

‡Figures reported in 1,000s

Regarding restoration costs, these uniformly decreased in scenarios where greater proportions of WLE were engaged ($1.43 million in Scenario 1 with no working lands; $812 thousand in Scenario 4 with 1,500 acres WLE; Table 22).

Table 22. Other Economic Implications of Different Model Scenarios

While WLE acquisition and restoration costs were lower than for NR acres, habitat value between WLE and NR are not equivalent (Bramwell et al. 2021). Higher NR acquisition and restoration costs may be justified in the Thurston County HCP by higher habitat values designed for and needed from NR acres. This economic data is intended only to help optimize the balance of WLE and NR acreage used in the HCP by weighing economic costs and benefits of utilizing each land type with respective habitat value costs and benefits.

Private tax revenue generated by working lands retained on the tax base while participating in the County HCP increased when more working lands acreage was recruited ($0 in Scenario 1 and $390 thousand in Scenario 4). Economic multipliers for cow-calf (2.05) and grass-finished beef sectors (2.51) were higher than economic multipliers for New Reserves established on Scotch broom (1.80), abandoned range (1.92), or abandoned cropland (1.90) due to higher rates of local re-spending in the livestock sectors. Scenarios overall exerted minimal impact on job creation due to the efficiency of managing large acreages, whether as New Reserves or Working Lands. Total program costs were optimized as greater acreages of working lands were utilized in the program ($99.68 million in Scenario 1 with no working lands; $92.18 million in Scenario 2 with 400 ac working lands; $71.58 million in Scenario 4 with 1,500 ac working lands; Table 23).

Table 23. Total Estimated Acquisition Costs for NR and WLE

*Figures here were from the HCP report submitted to US Fish and Wildlife in summer 2021

Generally, acquisition and restoration cost savings are possible by recruiting WLE acres, to the extent that the use of WLE still allows Thurston County to meet habitat protection requirements.

References

Brown LM, Fuda RK, Schtickzelle N, Coffman H, Jost A, Kazberouk A, Kemper E, Sass E, Crone EE. 2017. Using animal movement behavior to categorize land cover and predict consequences for connectivity and patch residence times. Landscape Ecology 32:1657–1670.

Bramwell, S.G., S. Hamman, C. Schultz, and S. Bussan. 2021. Evaluation of Grazed Prairie Land for Critical Species Protection in Western Washington; 2018-2020. https://s3.wp.wsu.edu/uploads/sites/2056/2021/04/Prairie-Grazing-Trials-Partner-Summary_FINAL.pdf.

Bussan S. 2022. Can cattle grazing benefit grassland butterflies? Journal of Insect Conservation. Published online 31 January 2022.

Cooke, S., & Watson, P. (2011). A comparison of regional export enhancement and import substitution economic development strategies. Journal of Regional Analysis & Policy, 41(1), 1-15.

Daniel Ebbert (2019). chisq.posthoc.test: A Post Hoc Analysis for Pearson's Chi-Squared Test for Count Data. R package version 0.1.2. https://CRAN.R-project.org/package=chisq.posthoc.test

Delaney J.T., Moranz R.A., Debinski D.M., Engle D.M., Miller J.R. 2016. Exotic-Dominated Grasslands Show Signs of Recovery with Cattle Grazing and Fire. PLOS ONE:1–15.

Douglas Bates, Martin Maechler, Ben Bolker, Steve Walker (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1-48. doi:10.18637/jss.v067.i01.

Henry, E., E. Brammer-Robbins, E. Aschehoug, and N. Haddad. 2019. Do substitute species help or hinder endangered species management? Biological Conservation 232:127–130.

Jerrentrup JS, Klimek S, Marchiori E, Bittante G, Ramanzin M, Sturaro E, Marini L. 2016. Impact of dairy farming on butterfly diversity in Alpine summer pastures. Agriculture, Ecosystems and Environment 232:38–45.

Kareiva P, and Odell G. 1987. Swarms of predators exhibit “preytaxis” if individual predators use area-restricted search. The American Naturalist 130(2):233-270.

Kareiva PM, Shigesada N. 1983. Analyzing Insect Movement as a Correlated Random Walk. Oecologia 56:234–238.

Kerr J.T., Sugar A., Packer L. 2000. Indicator Taxa, Rapid Biodiversity Assessment, and Nestedness in an Endangered Ecosystem. Conservation Biology 14:1726–1734.

Kristian Hovde Liland, Bjørn-Helge Mevik and Ron Wehrens (2021). pls: Partial Least Squares and Principal Component Regression. R package version 2.8-0. https://CRAN.R-project.org/package=pls

Kruess A, Tscharntke T. 2002. Contrasting responses of plant and insect diversity to variation in grazing intensity. Biological Conservation 106:293–302.

Leonard, J. L., & Watson, P. P. S. (2011). Description of the input-output model for Pacific coast fisheries. NOAA Technical Memorandum NMFS-NWFSC-111.

Linstone, H.A. and Turoff, M. 1975. The Delphi Method. Reading, Massachusetts: Addison-Wesley. 620 pp.

Michelot, T., Langrock, R., Patterson, T.A. (2016). moveHMM: an R package for the statistical modelling of animal movement data using hidden Markov models. Methods in Ecology and Evolution, 7 (11), pp. 1308-1315.

Murphy, D. D., P. S. Weiland, and K. W. Cummins. 2011. A critical assessment of the use of surrogate species in conservation planning in the Sacramento-San Joaquin Delta, California (U.S.A.). Conservation Biology 25:873–878.

Panter, K. E., L. F. James, D. R. Gardner, and Society for Range Management. 2002. Reproductive losses to poisonous plants: Influence of management strategies. Reviews in Endocrinology and Metabolism 55.

Pöyry, J., Lindgren, S., Salminen, J., Kuussaari, M. 2004. Restoration of Butterfly and Moth Communities in Semi-Natural Grasslands by Cattle Grazing. Ecological Applications 14(6):1656–1670

Pulliam RH. 1988. Sources , Sinks , and Population Regulation. The American Naturalist 132:652–661.

R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.

Robel RJ, Briggs JN, Dayton AD, Hulbert LC. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management. 23:295

Ross, J. A., S. F. Matter, and J. Roland. 2005. Edge avoidance and movement of the butterfly Parnassius smintheus in matrix and non-matrix habitat. Landscape Ecology 20:127–135.

Schultz CB, Haddad NM, Henry EH, Crone EE. 2019. Movement and Demography of At-Risk Butterflies: Building Blocks for Conservation. Annual Review of Entomology 64:167–184.

Schultz, C. B., A. M. A. Franco, and E. E. Crone. 2012. Response of butterflies to structural and resource boundaries. The Journal of Animal Ecology 81:724–734.

Schultz, C: B. 1998. Dispersal and its implications for reserve design in a rare Oregon butterfly. Conservation Biology 12:284-292

Steinback, S. R., & Thunberg, E. M. (2006). Northeast region commercial fishing input-output model. NOAA Technical Memorandum NMFS-NE-188.https://repository.library.noaa.gov/view/noaa/4029.

Stevens VM, Polus E, Wesselingh RA, Schtickzelle N, Baguette M. 2005. Quantifying functional connectivity: Experimental evidence for patch-specific resistance in the Natterjack toad (Bufo calamita). Landscape Ecology 19:829–842.

Thompson JN, Pellmyr O. 1991. Evolution of Oviposition Behavior and Host Preference in Lepidoptera. Annual Review Entomology 36:65–89.

Turchin, P. 1998. Quantitative analysis of movement: measuring and modeling population redistribution in animals and plants. Reprint: Beresta. Storrs, Connecticut. 2015

Wallis De Vries MF, Noordijk J, Colijn EO, Smit JT, Veling K. 2016. Contrasting responses of insect communities to grazing intensity in lowland heathlands. Agriculture, Ecosystems and Environment 234:72–80. Elsevier B.V.

Wang Y, Chen J, Liu W, Xu R (2007) Efect of cultivating croplands and grazing in aridgrassland habitats on the conservation of melitaeinebutterfies in a mountainous area in Northern China. Sci China, Ser C: Life Sci 50:40–46

Webler, T., D. Levine, H. Rakel, and O. Renn. 1991. A novel approach to reducing uncertainty: the group Delphi. Technological forecasting and social change, 39(3): 253-263.

Weiss SB. 1999. Cars, cows, and checkerspot butterflies: Nitrogen deposition and management of nutrient-poor grasslands for a threatened species. Conservation Biology 13:1476–1486.

Zakkak S, Chatzaki M, Karamalis N, Kati V (2014) Spiders in the context of agricultural land abandonment in Greek Mountains: species responses, community structure and the need to preserve traditional agricultural landscapes. J Insect Conserv 18:599–611.

Appendix A. 

Appendix B.  Species seeded into farm sites, along with life history and seeding rates

PLS = Pure Live Seed

* Only seeded CASLEV into 3 of the paddocks at Colvin and into 6 plots (1 per paddock) at Riverbend due to limited seed availability. Seeding rate of CASLEV at Riverbend was slightly higher than other sites: 309 PLS/m²;

† Only seeded into plots, not the entire paddock, at Fisher due to limited seed availability

Appendix C. 

Actively flowering plant or potential host species found within plant plots along silvery blue butterfly paths. The category refers to how the species was classified in our analyses. If no category is listed, the species was not included in the analyses.