We investigated the economic and wildlife implications of converting exotic forages to native warm-season grass (NWSG). In spite of unfavorable growing conditions due to drought, cattle grazing NWSG consistently outperformed conspecifics on exotic grass forages. Partial budget analysis indicated that NWSG pastures yielded up to 36% marginal rates of return despite establishment costs. Fire ant indices did not differ among treatments, though NWSG increased Dickcissel productivity (fledglings/ha) through greater nest site availability. These results suggest NWSG can benefit avian species such as Dickcissels while offering a competitive alternative to exotic forages, resulting in net benefits for both conservationists and producers.
The purpose of this project was to evaluate the costs and returns of converting exotic grass pastures currently used in production (bermudagrass [Cynodon dactylon] and tall fescue [Schedonorus arudinaceus]) to native warm-season grass (NWSG) pasture and to characterize the response of wildlife and imported fire ants (Solenopsis spp.). Converting exotic grass pastures to NWSG has the potential to restore the natural functions of pastures by simultaneously providing excellent forage for cattle, nesting structure for birds, and excluding imported fire ants (IFA). For example, the structure of NWSG may be more amenable for livestock foraging and performance than exotic forages (Burns et al. 1984), and varieties of NWSG can be adapted to local growing conditions, tolerating poor soil conditions (Jung et al. 1988) and requiring few nutrient inputs (Brejda et al. 1995). Preliminary results of meat quality studies at Mississippi State University also indicate that meat from cattle grazed on NWSG has similar consumer acceptability; but better shelf life, higher protein, and lower fat content compared with meat from bermudagrass (M. W. Schilling, pers. comm.). In addition, the tall and clustered growth of bunchgrass provides both suitable nesting structure for grassland birds and interspaces for wildlife movement (Harper et al. 2007). Exotic grasses are ubiquitous throughout the Southern United States as pasture and forage for cattle (Barnes et al. 2013). However, these forages are highly unsuitable for wildlife such as Northern Bobwhite (Colinus virginianus) and grassland songbirds because they grow in dense mats and outcompete other native plant species (Hays et al. 2005). Thus, widespread establishment of exotic forages has likely contributed to declines in many grassland songbird populations (Sauer and Link 2011). Specifically, NWSG should be preferable over exotic forages for the Dickcissel (Spiza americana), a bird species that selects grasslands with tall and dense vegetation to establish territories and nest (Zimmerman 1971). Disturbance from heavy grazing may also maintain early successional structure and composition preferred by IFA (Tucker et al. 2010), an important agricultural pest (Flanders and Brees 2009).
Incorporating NWSG in cattle production has potential to substantially improve sustainability and wildlife conservation on private lands throughout the Southeast. However, NWSG conversion incurs costs from establishment and loss of revenue while pastures are taken out of production for two years. The possibility of not recovering these losses due to establishment failure, variation in market conditions, or weather therefore presents substantial risks for producers. Despite potential economic and environmental benefits of NWSG, uncertainty regarding establishment and risk remains a significant barrier to incorporating these grasses in livestock operations (Taylor 2000, Doll and Jackson 2009). Information is also needed on the costs and benefits of NWSG conversion to assist in distributing cost-share incentives to producers (Claassen et al. 2008). I therefore evaluated costs and benefits of NWSG conversion using an operational, experimental grazing system, and measured the response of a breeding grassland birds and IFA.
- Conduct a partial budget analysis by creating enterprise budgets for NWSG establishment and grazing exotic and NWSG pastures;
- Characterize response of Dickcissels (Spiza americana) to NWSG conversion by quantifying nest density, daily nest survival rates, and number of fledglings produced per nest among grazing systems. I will focus on Dickcissels because they are common in Mississippi, their populations are declining, and they are a suitable model organism for grassland avifauna in general because they are obligate grassland nesters;
- Quantify IFA response to NWSG conversion, and if necessary include any costs for eradication in enterprise budgets.
- Use sensitivity analyses on costs and cattle weight gain parameters to identify effective targets for cost-share or subsidies to make NWSG conversion viable.
This study was conducted at Mississippi State University’s (MSU) Prairie Research Unit in Monroe Co., Mississippi (N 33?47′, W 88?38′), located in the Black Belt Prairie region. Nine pastures (6.4-10.5 ha in size) were assigned to one of three treatments, each replicated three times. Treatments included a grazed mix of bermudagrass and tall fescue (hereafter, grazed mixed exotic pasture, or GMEP); Indian grass monoculture (Sorghastrum nutans, “Kentucky ecotype”; hereafter, grazed Indian grass native pasture, or GINP); and grazed mixed native pasture (GMNP) planted with big bluestem (Andropogon gerardii, “Kaw”), little bluestem (Schizachyrium scoparium, “Aldous”), and Indian grass. After two years of NWSG establishment, pastures were stocked with fall-born Angus ´ Hereford steer calves (average initial body weight: 250 kg) from mid-May through August 2011-2012 at a rate of 2.7 steer ha–1. All calves were weighed when initially stocked and again when removed in late summer. Prior to each summer grazing period soil samples from pastures were tested to determine fertilizer and lime application rates according to best management practices for exotic pastures and NWSG.
Total growing precipitation (April–October, recorded by a weather monitoring station in Aberdeen, MS; NCDC 2014) was 711.7 mm in 2011 and 808.4 mm in 2012, whereas the 30-y average (1983–2012) was 744.7 mm. Total precipitation in 2011 was unusually low in May (35.6 mm) and July (33.2 mm), but increased in June (100.4 mm) and August (92.3 mm). In 2012 total precipitation in May (66.4 mm) and June (59.0 mm) were well below the 30-y averages (129.4 and 114.4 mm for May and June, respectively). Precipitation subsequently increased in July (129.6 mm) and August (277.5 mm).
I used a partial budget approach to compare the relative net benefit in converting exotic grass pastures to NWSG. I constructed partial enterprise budgets for each treatment by only including costs that varied among treatments, thus permitting calculation of the marginal rate of return (MRR) for converting GMEP to native grass pasture (CIMMYT 1988). Establishment costs included herbicide purchase and costs associated with two applications for bermudagrass and tall fescue eradication in fall 2007. In spring 2008, costs included fire lane establishment, prescribed fire, pure live seed purchase, no-till planting, and one at-plant herbicide application of glyphosate + imazapic (Table 1). Budgets excluded costs for pasture rental, fertilizer application, procurement and marketing costs, and fixed costs such as depreciation, insurance, and taxes, because these should be identical across treatments. Establishment costs were then prorated over 10 years, and included as an annual cost during operation years. Instead of establishment costs, exotic grass pastures incurred maintenance costs from two herbicide applications for bermudagrass release. Additionally, converting exotic forages to NWSG incurs opportunity costs from taking pastures out of production for two years, so I summed cash rent reported by NASS (2013b) for pastureland in Monroe Co., Mississippi, during 2009–2010, adjusting for inflation, and included this value among establishment costs for each NWSG treatment.
During operation years (2011–2012), pastures were managed with regimes typically recommended for exotic and native grass pastures, including fire lane establishment and prescribed fire for NWSG, and fertilizer application informed by soil tests. I used price and rates for prescribed fire from my study at the PRU. Fertilizer rates were based on actual use during the study, and I used annual prices of ammonium nitrate reported by NASS (2013a) for East South Central United States. Interest on operating capital was also included. Inputs of lime and phosphorous may incur additional costs to a grazing enterprise, though in my study application rates for lime and phosphorous related to individual pasture condition rather than stand type. As such, these inputs were equivalent to other pasture maintenance items that did not vary by treatment, such as fencing, and therefore were not included in the partial budget analysis. I calculated cost from steer purchase by multiplying treatment-specific mean starting weights by stocking rate and price paid for steers and heifers in May ($ kg–1) reported by NASS (2013a). Cost of capitalization of steer and other operation costs were calculated over the length of the grazing season and finishing period each year. I also included costs from death loss during operation, and death loss and shrink during procurement (Rhea et al. 2007).
I used linear mixed models in R (R Development Core Team 2014) to estimate ADG (kg d–1) by treatment and year. I computed ADG for each animal by subtracting starting weights in May from end weights measured when steers were removed in September, then dividing by the number of days grazed in each season (111 and 113 days in 2011 and 2012, respectively). I also included a random effect of pasture. After the summer grazing period, I assumed all steers were finished on shelled corn and soybean meal for 130 days, yielding an average daily gain (ADG) of 1.32 kg day–1 (Rhea et al. 2007). I then estimated final weights using treatment- and year-specific ADG estimates multiplied by number of summer grazing days in a season, adding the May starting weights and weight from finishing. I computed annual gross income as the product of final weight, price received for steers the following January reported by NASS (2013a), and the stocking rate.
I calculated MRR for NWSG conversion (GINP or GMNP) by dividing the marginal net benefit (difference in net benefit between NWSG treatment and GMEP) by the total marginal cost (difference in total cost between NWSG treatment and GMEP) and multiplied the result by 100% to convert to percentage. The resulting value is interpreted as the additional benefit (after investment in NWSG conversion) that the enterprise would receive relative to GMEP (CIMMYT 1988). Finally, cattle prices may have a large influence on profitability of a grazing enterprise (Manley et al. 1997), and it is useful to examine how changes in price would affect MRR beyond the two years of my study. Therefore, I calculated MRR from NWSG conversion for the previous 15 years using my 2011 partial budget as a baseline and May price paid (1999–2013) and January price received (2000–2014; NASS 2013a). I then examined the number of years that NWSG conversion yielded a positive rate of return, and calculated the breakeven value for price received.
From early-May through July 2011–2012, teams of three technicians searched for nests in each pasture using a rope-drag method once every two weeks. Nests were monitored every 2-4 days, and daily within 2-3 days of anticipated fledge date to accurately determine nest fate. I estimated nest density in each pasture using a multi-state Jolly-Seber model to account for nest stage- and treatment-specific differences in detectability. I also estimated mean number of fledglings produced per nest of each treatment, and used a Bayesian approach to estimate nest daily survival rates for each treatment (Royle and Dorazio 2008) in program R (R Core Development Team 2014). I then used estimates of nest density, brood size, and nest survivorship to compute productivity estimates (fledglings ha–1) for each treatment and year. I also used a partial budget approach to calculate MRR from investment in terms of Dickcissel productivity. In this case, I replaced net benefit with treatment-specific productivity estimates, and divided the marginal net benefit (difference in productivity between NWSG and GMEP) by the marginal cost to yield a marginal rate of return in productivity. I interpreted this value as the change in productivity for each $100 invested in NWSG conversion.
During 2011-2012, I conducted surveys for IFA mound density and activity each spring prior to cattle stocking, and again in the fall following cattle removal. I surveyed along four 100-m transects established randomly in each pasture, and I used a Bayesian approach to distance sampling (Royle and Dorazio 2008) to estimate mound densities (active mounds ha–1). I measured visual obstruction from vegetation using a Robel pole (Robel et al. 1970) at five points at 15-m intervals along each transect and include this as a covariate of mound detection probability because tall vegetation may obscure visibility of mounds. I measured IFA activity by placing 9 bait cups at 10-m intervals along each transect, and collected these after 30 minutes (Hill et al. 2008). I then compared IFA occurrence among vials in GMEP and NWSG treatments using linear mixed models in R (R Development Core Team 2014).
In each enterprise budget, an annual cost of pesticides would be incurred if IFA mound density estimates exceeded 49 mounds ha–1 (20 mounds ac–1) in a treatment, as pesticide applications below this density may harm native ant communities more than can be justified for IFA control (Flanders and Brees 2009). I further assumed that pesticide applications would reduce densities below 49 mounds ha–1 and eliminate costs from IFA damage for that year. If mound density estimates were below 49 mounds ha–1, I assumed that no costs from pesticides or IFA damage were incurred.
Additionally, my study sites were located along the hybrid zone between S. invicta and black imported fire ant (S. richteri Forel) (Shoemaker et al. 1994, Streett et al. 2006), and two breeding systems are found among S. invicta including monogyne colonies with one fertile queen per colony and a polygyne system with multiple queens per colony. Polygynous mounds are often more numerous and closely spaced (Porter et al. 1991, Vogt et al. 2009), and ants from these colonies forage farther than their monogynous congeners (Martin et al. 1998). Identifying species and breeding system of mounds can account for mound distribution, response to management, and impacts on other arthropods and breeding birds (Forbes et al. 2002, Hale et al. 2011). In mid-April 2013, I randomly selected one mound from each transect for ant identification (n = 4 mounds per pasture). At each mound, a sample of 100 ants was extracted using an aspirator and placed in a vial with 5 mL of hexane. Individuals from each mound were then identified to species or hybrid by analyzing their cuticular hydrocarbons and venom alkaloids with gas chromatography/mass spectrometry (Vander Meer et al. 1985, Ross et al. 1987). Samples were analyzed at the USDA Agricultural Research Service in Stoneville, Mississippi, following Menzel et al. (2008). I collected an additional 20 workers from each mound to distinguish among monogyne, polygyne, and hybrid colonies, preserving them in 70% isopropanol for polymerase chain reaction (PCR). Polygyne ants possess both alleles for the Gp-9 gene (B and b) whereas ants from monogyne colonies possess only the B allele, so PCR offers an accurate and cost-effective method for discriminating between the two breeding systems (Valles and Porter 2003). DNA was amplified with PCR following the protocol of Valles and Porter (2003), except we used TaKaRa Taq (hot start version) for DNA polymerase, and 65–125 ng genomic DNA per 23.8 mL reaction (D. Cross, pers. comm.). We also used a MyCyclerTM thermal cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA), and electrophoresed 3 mL of amplified DNA on agarose gel with GelRed (Biotium) to detect and photograph DNA under UV light.
I conducted sensitivity analyses by varying individual parameters in each partial budget by ±20% and monitoring change in MRR for GINP and GMNP, thus evaluating the relative importance of each parameter. I used prices and rates from 2011 as a baseline scenario, and parameters to vary included January selling price, ADG, fertilizer price, seed prices, prescribed fire, and interest rates. I also tested the response of MRR to total establishment costs of GINP and GMNP (before being prorated) because variation in NWSG establishment success determines the final price paid for establishment.
I monitored 85 Dickcissel nests in 2011 and 123 nests in 2012. Over two years and among all nests, apparent fledging rate was 24% (n = 49). Causes of failure for unsuccessful nest attempts included 81% from predation (n = 128), 11% from abandonment (n = 17), 3% from trampling (n = 5), and 6% from other (n = 9). Four nests were trampled in grazed pastures, and one was crushed by farm equipment in a non-grazed pasture. Thirteen nests (6%) were found with IFA, though I was unable to determine whether ants were the direct source of nest failure. Among all nests, seventeen nests (8%) were parasitized by brown-headed cowbirds (Molothrus ater). Mean clutch size and number of fledglings produced per successful nest were similar among treatments (Table 2). In 2011 and 2012, the JS nest density model indicated densities were lower for GMEP, whereas densities among NWSG treatments tended to be higher, except for GINP in 2012 (Fig. 1).
Bayesian p-values suggested the nest survival model adequately fit the data (p = 0.44), when p < 0.05 or > 0.95 typically indicate poor fit (Gelman et al. 2014). Estimates of DSR were lowest among nests in GMEP (0.91, 95% CrI: 0.85, 0.95), and greater and similar for GINP (0.92, 95% CrI: 0.85, 0.96) and GMNP (0.93, 95% CrI: 0.88, 0.97), though credible intervals overlapped among all treatments. Over a 21-d nesting period, while accounting for effects of age and year, the model predicted 0.16 (95% CrI: 0.04, 0.34) and 0.08 (95% CrI: 0.01, 0.22) survivorship for GMEP nests in 2011 and 2012, respectively. Nest survivorship estimates were somewhat greater for grazed native grass treatments, with estimates of 0.25 (95% CrI: 0.08, 0.34) and 0.15 (95% CrI: 0.03, 0.35) for GMNP in 2011 and 2012, and 0.20 (95% CrI: 0.04, 0.42) and 0.11 (95% CrI: 0.01, 0.30) for GINP. Productivity was least for GMEP with estimates of –1 (Fig. 2). Productivity estimates among grazed native treatments were greater but GINP tended to be lower than GMNP, particularly in 2012.
These results suggest a generally positive response in Dickcissel productivity to NWSG establishment relative to GMEP, though with some uncertainty from overlapping credible intervals. Grazing may have a disproportionate effect on availability of nest sites in NWSG for this species, so if reduced stocking rates or resting is recommended for managing and maintaining native warm-season grass pastures (Mousel et al. 2003, Chamberlain et al. 2012), employing such regimes on native grass pastures may increase availability of nesting habitat relative to exotic grass pastures in this region.
Imported Fire Ant Response
I captured 27,163 IFA individuals over four sampling events. Native ants also captured included Monomorium minimum Buckley (n = 13,045), Solenopsis molesta Say (n = 119), and Nylanderia vividula Nylander (n = 4), which is similar to ant richness reported in other exotic pastures in the Black Belt (Hill et al. 2008). Of the 48 mounds sampled in 2013, amplification from PCR indicated that all belonged to the monogyne breeding system, none possessing the polygene Gp-9b allele. I also determined that 8 mounds were S. richteri (Alkaloid index [I] < 0.06, where 0.06 < I < 0.85 suggests the colony is hybrid; Ross et al. 1987), whereas the rest (n = 40) were S. invicta × richteri hybrids (I range: 0.06–0.54).
Modeling IFA occurrence suggested significant positive interactions for GINP × year (β = 1.14, SE = 0.32, P < 0.001), whereas the overall effect of year was negative (β = –1.66, SE = 0.25, P < 0.001). There was also a significant negative interaction for GINP × season (β = –0.81, SE = 0.31, P = 0.010), whereas the effect of season alone was positive (β = 0.65, SE = 0.24, P = 0.006). This is best illustrated with interaction plots (Fig. 3), where occurrence of IFA at vials among treatments generally decreased from 2011 to 2012, though to a lesser degree for GINP, and where occurrence was greater in Fall than in Spring for all treatments except GINP.
Over two years and four surveys I detected 1,351 IFA mounds across all study sites, and I estimated 4,064 mounds in the superpopulation (95% CrI: 3,801–4,340). These estimates indicate that ignoring variation in detectability would have substantially underestimated and biased mound density at my study sites. For the process part of the model, mean estimates for effects of year (β = 0.37, SD = 0.26) and of season (β = –0.22, SD = 0.23) were positive and negative, respectively, though credible intervals of both parameters overlapped zero. Credible intervals from all interaction terms also overlapped zero. Mean parameter estimate for treatment was higher for GINP (β = 0.63, SD = 0.36) whereas GMEP (β = 0.28, SD = 0.36) and GMNP (β = 0.03, SD = 0.33) were more similar, though there was broad overlap among credible intervals. These parameter estimates are reflected in derived treatment estimates of mound density (Fig. 4), with slight intra-annual decreases and inter-annual increases in mound density, but overlapping credible intervals among treatments. However, mound density estimates among all treatments largely exceeded the 49 mounds ha–1 threshold requiring pesticide application, except for GMNP in fall 2011 when credible intervals overlapped 49 mounds ha–1.
My results indicate there was seasonal and annual variation in IFA occurrence and mound density unrelated to treatment, but in general there was no appreciable response to NWSG conversion during the two years of my study. This may reflect seasonal fluctuations in IFA populations, which often decline from spring until mid-summer due to allocation of resources towards production of sexuals rather than workers, then recover in the fall through winter (Tschinkel 1993). Similarly, mound density decreased from February–August in nearby Clay Co., MS, before increasing again in the fall (Vogt et al. 2004). Additionally, soil temperature and insolation that may affect mound density and foraging activity of IFA (Porter and Tschinkel 1987, Vogt et al. 2003, Vogt et al. 2009), and disturbance may facilitate establishment and maintenance of IFA colonies while reducing richness and diversity of native ants (Stiles and Jones 1998, King and Tschinkel 2008, Stuble et al. 2011). Disturbance from recent NWSG establishment, and grazing in the two NWSG treatments, may have precluded observing conditions similar to other native prairie in the region. Given that my estimates of IFA mound density were above the 49 mounds ha–1 threshold requiring pesticide application, irrespective of treatment, I excluded this cost from the partial budget analysis while acknowledging this may present a substantial cost to producers.
For modeling ADG, inclusion of a Treatment × Year interaction was not supported (L = 3.60, df = 2, P = 0.17), but dropping Treatment (L = 13.07, df = 2, P = 0.002) or Year (L = 51.58, df = 1, P < 0.001) effects did not improve fit. Overall, ADG was lower in 2012 than 2011 (b = -0.29, SE = 0.04, P < 0.001), and ADG for GMEP was 33% and 31% lower than GINP and GMNP in 2011, respectively, and 42% and 40% lower in 2012 (Table 3). Confidence intervals for ADG overlapped broadly between GINP and GMNP estimates.
Nutritional analysis of forages among my pastures indicated crude protein and digestibility were highest among GMEP pastures, likely a response to greater fertilizer rates (Oloyede 2013). However, dry matter yield was higher among native grass pastures, especially for GINP during peak production (July), and therefore forage availability rather than quality may explain observed differences in ADG among treatments. Average daily gain estimated for GMEP in 2011 was comparable with gains reported previously for bermudagrass with higher stocking rates but also with greater nutrient inputs (Utley et al. 1976, Burns et al. 1984, Burns and Fisher 2008, Burns et al. 2009, Burns and Fisher 2013). Higher ADG for NWSG is also consistent with previous studies (Coleman et al. 2001, Gillen and Berg 2001, Burns and Fisher 2013), although increased stocking rates in exotic grass treatments often produced total gain ha–1 equivalent to native grass pasture. However, higher stocking rates in these intensive systems also incurred higher production costs, making native grass systems more profitable overall (Phillips and Coleman 1995, Coleman et al. 2001, Gillen and Berg 2001). Higher stocking rates also exposes producers to greater risk from variability in weather (Parsch et al. 1997) and cattle price (Manley et al. 1997).
Importance of forage availability is also illustrated by the reduction in ADG that coincided with drought in 2012. Drought reduces profitability of grazing enterprises (Dunn et al. 2010), and all treatments yielded lower net revenues in 2012 than 2011. Drought resistance is touted as a potential benefit for native warm-season grasses due to their deep root growth (Harper et al. 2007), and for polycultures over monocultures (Sanderson et al. 2005, Deak et al. 2010). However, in my study the response to drought from NWSG and GMEP was similar because a Treatment ´ Year interaction for ADG was not supported. This trend may reflect the recent establishment of NWSG pastures, but may also be due to the fairly dry conditions of both years during my study, preventing me from quantifying the response of forages to the full range of precipitation experienced in this region.
Educational & Outreach Activities
Monroe, A. P. Grassland bird response to NWSG establishment. Native Warm-season Grass Workshop, 31 May 2012, Prairie, MS. 30 attendees.
Monroe, A. P., S. K. Riffell, J. A. Martin, and L. W. Burger, Jr. In review. Testing hypotheses on avian nest site selection and breeding success in native and exotic grazing systems. Ecosphere.
Monroe, A. P., J. A. Martin, R. B. Chandler, and S. K. Riffell. In revision. Estimating avian nest density using capture-recapture models. The Condor: Ornithological Applications.
Monroe, A. P., J. A. Martin, S. K. Riffell, and L. W. Burger, Jr. 2014. Effects of measuring nestling condition on nest success in the Dickcissel (Spiza americana). The Wildlife Society Bulletin. In press [doi: 10.1002/wsb.412]
Monroe, A. P., J. A. Martin, R. B. Chandler, and S. K. Riffell. 2014. Accounting for detectability and survival in avian nest density estimates with capture-recapture models. Southeastern Natural Resources Graduate Student Symposium at Mississippi State University, Mississippi State, MS
Monroe, A. P., S. K. Riffell, J. A. Martin, L. W. Burger, Jr., and H. T. Boland. 2013. Effects of vegetation structure and weather on Dickcissel nest survival in grazing systems. Wilson Ornithological Society 125th Anniversary Meeting at the College of William & Mary, Williamsburg, VA
Monroe, A. P., S. K. Riffell, J. A. Martin, L. W. Burger, Jr., and H. T. Boland. 2012. Response of a breeding grassland bird to native warm-season grass pasture conversion. The 8th Eastern Native Grass Symposium in Charlottesville, VA
Monroe, A. P., S. K. Riffell, J. A. Martin, L. W. Burger, Jr., and H. T. Boland. 2012. Converting bermudagrass to native warm-season grass pasture: effects on Dickcissel nest success. Southeastern Prairie Symposium at Mississippi State University, MS
Monroe, A. P., S. K. Riffell, J. A. Martin, L. W. Burger, Jr., and H. T. Boland. 2011. Dickcissel nest success in bermudagrass and native warm-season grass pastures. Mississippi Chapter of The Wildlife Society annual meeting at Lake Tiak-O’Khata Resort, Louisville, MS. (Received Outstanding Student Poster Award)
The results of this study are an important first step in examining the viability and wildlife response of cattle production on NWSG. I hope other researchers will consider our results when designing future studies to replicate and expand this research across the region and over longer time intervals. The next steps for this research will be to publish our results in peer-reviewed journals, as well as develop prediction models for broader adoption of NWSG across the region. For example, in the Southeastern United States, bermudagrass (Cynodon dactylon) and tall fescue (Schedonorus arudinaceus) are established on millions of ha (Barnes et al. 2013), so I will use estimates from this study to predict the regional change in Dickcissel productivity and nitrogen fertilizer use with 5%, 20%, or 50% conversion of exotic forages to NWSG.
Establishment cost was 10.5% greater for GMNP ($880.12 ha–1) than GINP ($796.13 ha–1), driven by higher cost of pure live seed for the native grass mix (Table 1). When prorated over ten years, establishment cost per annum was $116.76 ha–1 for GMNP and $105.62 ha–1 for GINP. In contrast, initial maintenance cost for GMEP incurred a prorated annual cost of $12.75 ha–1. Both years, all treatments received positive net revenue from operations, though net revenue declined by >75% from 2011 to 2012 (Table 4). However, higher net benefits in native treatments relative to GMEP compensated for higher total marginal costs, with 35.9% and 28.5% MRR for GINP in 2011 and 2012, respectively. Marginal rate of return was also positive for GMNP in 2011 (12.8%), but not in 2012 (–2.0%). The disparity between the two years for GMNP was likely driven by a combination of lower ADG (Table 3), heavier starting weights, and 18% higher spring purchase price for cattle in 2012 but only a 5% increase in price received in the fall. Price of ammonium nitrate was also 5% higher in 2012 than 2011. Sensitivity analyses indicated that ADG and selling price contributed greatly to changes in MRR for GMNP, with >150% change in MRR from 20% change in gain or selling price (Table 5). GINP responded similarly, but to a lesser degree. Decreasing cost of prescribed fire and establishment had a positive effect on MRR, particularly for GMNP with >50% increase in MRR from each parameter. Fertilizer, seed, and interest rates had comparatively smaller effects on net revenue.
During 2000–2014, price received in January for steers ranged from $2.11 kg–1 in 2000 to $3.32 kg–1 in 2013. Assuming all other values were constant from the 2011 partial enterprise budget, GINP would yield a positive MRR for 11 of 15 years under consideration (Fig. 5a), with a breakeven selling price of 2.06 kg–1 (or $93.53 cwt–1). Marginal rate of return would be positive for cattle grazing GINP each year that selling price was at or above this value, though in 1999–2000 MRR was near 0. Conversely, MRR from GMNP was positive for 4 of 15 years under consideration (Fig. 5b). The breakeven selling price for this treatment was $2.48 kg–1 (or $112.68 cwt–1).
Regarding MMR from Dickcissel productivity, in 2011 GINP increased productivity by 0.15 fledglings ha–1 relative to GMEP for every $100 invested, whereas productivity in GMNP increased by 0.32 for the same investment. In 2012, decreases in productivity for GINP lead to slightly negative MRR for every $100 investment (–0.03), whereas GMNP yielded a 0.30 increase in productivity. This suggests that a greater investment in GINP (such as by converting a larger area) would be required to achieve increases in productivity equivalent to GMNP.
Price of nitrogen fertilizer may affect rates of return for NWSG conversion (Doxon et al. 2012), and I found a positive effect of nitrogen prices on MRR. GMEP required twice the amount of fertilizer than native pastures, and as fertilizer prices increase, the difference in cost between GMEP and native pastures also increases, resulting in a higher MRR for native pastures. However, sensitivity analysis indicated that the response to fertilizer prices was relatively small, indicating that benefits from increased ADG due to NWSG conversion surpassed potential savings from reduced fertilizer costs. One study of fescue-bermudagrass pastures demonstrated that higher gains (max. 0.66 kg d–1) can be achieved at higher stocking rates with annual fertilizer applications up to 4.6 times greater than in my study (Franzluebbers et al. 2012). Applying fertilizer at such rates would increase fertilizer costs for GMEP from $36–38 ha–1 to $164–173 ha–1, which approaches the $182–194 ha–1 from maintenance and prorated establishment costs of NWSG pastures. However, intensive management of exotic forages is also accompanied with other costs and risks beyond the price of fertilizer, such as greater nonpoint source pollution and financial risk from fluctuations in fertilizer prices and drought.
The comparative advantage of NWSG pastures over GMEP was primarily from higher ADG. Importance of ADG and selling price also suggests that the advantage of NWSG conversion over current production from exotic forages may depend on external factors such as beef markets and weather. If recent trends for higher cattle prices persist in coming years there is a high likelihood that conversion to either GINP and GMNP would yield a positive MRR. Marginal rate of return was also sensitive to establishment costs and price of seeds. Establishment success may vary due to precipitation (Bakker et al. 2003) and control of exotic warm-season grasses (Barnes 2004), and in my study the incomplete eradication of bermudagrass followed by disturbance from grazing likely encouraged the spread of this grass in native pastures (Monroe 2014). Continuation of this study in subsequent years may therefore have incurred additional costs for herbicide treatment and removing NWSG pastures from production for recovery. These results illustrate several important potential sources of risk that producers should consider when investing in NWSG conversion.
Whereas lower establishment costs and slightly higher ADG for GINP led to a higher financial MRR for this treatment, MRR on Dickcissel productivity was consistently higher for GMNP. The difference in productivity between the grazed native grass treatments was a function of higher nest densities in GMNP because daily survival rates and brood size were similar to GINP, suggesting a greater availability of nest sites in the polyculture than in the NWSG monoculture. Sensitivity analysis suggests that subsidies aimed at promoting wildlife should seek to offset higher seed and establishment costs of GMNP, and reduce the comparative advantage of higher ADG with GINP given variable livestock prices.
Higher MRR in Dickcissel productivity for GMNP also indicates fledglings are relatively more expensive to produce with GINP. Interpreted from this perspective presents a shift from payment for actions towards payment for results. Instead of mandating a specific management regime, result-oriented payments are offered to producers for specific outcomes such as for number of nests found and protected (Musters et al. 2001). Benefits of result-oriented schemes over payment for actions (reviewed by Burton and Schwarz 2013) include greater cost-efficiency of conservation subsidies, increased flexibility for landowners in making management decisions, and stronger ties between landowners and biodiversity by treating environmental outcomes as another source of revenue in their enterprise. For example, if producers managing NWSG grazing are compensated for each Dickcissel nest in their pasture, they may be further motivated to reduce stocking rates during drought to maintain tall cover and nest site availability. However, ability of landowners or monitoring agencies to effectively and efficiently estimate nest density remains a significant challenge in using such an index as a biodiversity indicator for result-oriented payments (Matzdorf et al. 2008, Matzdorf and Lorenz 2010, Burton and Schwarz 2013).
Given an enterprise that purchases steer in May, grazes continuously through summer, and sells the following January, my results suggest NWSG conversion has clear potential to produce positive MRR on investment relative to exotic grass pastures currently used for livestock production, even when including prorated costs from establishment. My study suggests that at equivalent stocking rates NWSG pastures may yield greater livestock gain than exotic forages despite lower intensity (reduced fertilizer inputs), and the additional net benefit for wildlife suggests a win-win scenario for producers and conservationists. Furthermore, the economic benefit of intensive grazing is increasingly questioned given the higher production costs due to heavy fertilizer inputs, which then reduce the return on investment from each animal in production (Coleman et al. 2001, Fuhlendorf and Engle 2001, Gillen and Berg 2001, Burns and Fisher 2013). Currently the majority of Natural Resources Conservation Service (NRCS) assistance and funding promotes intensive livestock management on private lands, such as increased fencing and water source distribution (Toombs and Roberts 2009). Diverting efforts instead towards promoting NWSG conversion and less intensive management may increase the sustainability and wildlife value of private pastureland in the United States.
My analysis shows that reduction in costs from establishment and management of NWSG can reduce risk from NWSG conversion through substantial increases in MRR. This indicates several potential avenues for subsidies and cost-share to encourage participation. For example, working-land programs such as the Environmental Quality Incentives Program (EQIP) offer payments for a variety of environmentally-beneficial structural and management practices, and >60% of funds are appropriated for livestock producers (Claassen et al. 2008). Surveys indicate a general interest in conservation and the environment among cattlemen (Jacobson et al. 2003, Doll and Jackson 2009, Willcox et al. 2012), which suggests that they may be receptive to NWSG. Furthermore, successful programs such as the Conservation Reserve Program may have unwittingly discouraged enrollment by excluding grazing and haying from contracts (Esseks and Kraft 1986), so the ability of cattlemen to keep land in production will likely encourage participation in programs that promote NWSG conversion. Many landowners are unaware of the availability and their eligibility for conservation programs, and this is consistently a major factor limiting landowner participation (Esseks and Kraft 1986, Jacobson et al. 2003, Doll and Jackson 2009, Lubell et al. 2013). Active outreach is therefore needed to promote the different cost-share programs available for NWSG conversion while ensuring that both landowners and funding agencies can make informed decisions.
Areas needing additional study
Additional study is needed on the long-term viability of NWSG conversion for producers, and on the appropriate grazing regime that promotes cattle gain without reducing Dickcissel nest site availability or increasing bermudagrass spread. Long-term studies are also needed to characterize the response of pest species such as IFA to NWSG establishment. Another important caveat is that I did not consider other members of the avian grassland community such as Grasshopper sparrows (Ammodramus savannarum) that may respond positively to low structure in exotic cool-season grass fields (McCoy et al. 1999, McCoy et al. 2001). Incorporating NWSG into current exotic forage operations to foster a mosaic of tall and short structure may be beneficial for the full complement of avian biodiversity in this region (Giuliano and Daves 2002), but this warrants further study.