Nitrogen fertilizer and clover inclusion effects on the establishment of fine fescue taxa

Abstract

Little information exists on establishment vigor differences among fine fescue taxa (Festuca L. spp.) and on the effects N fertilizer levels or clover (Trifolium L. spp.)-inclusion during establishment. Five replicated field experiments in Indiana, Minnesota, and Oregon were conducted from 2019 to 2021. Data collection continued for 9 mo after planting to investigate differences among four fine fescue taxa and determine optimal N fertility or clover-inclusion programs for the establishment of fine fescue taxa. Six N fertilizer levels ranged from 0 to 122.5 kg N ha⁻¹ during the 8 wk after planting, and two treatments included clover-inclusion + 0 kg N ha⁻¹ at seeding. The three taxa of the Festuca rubra complex—strong creeping red fescue (F. rubra L. ssp. rubra Gaudin), slender creeping red fescue [F. rubra ssp. littoralis (G. Mey.) Auquier], and Chewings fescue (F. rubra ssp. commutata Gaudin)—were similar to one another and required the shortest amount of time until 90% establishment (i.e., faster establishment) compared with hard fescue (Festuca brevipila Tracey). Providing 24.5 kg N ha⁻¹ at seeding hastened establishment of all fine fescues compared with no N fertilizer (0 kg N ha⁻¹). Applying 49 kg N ha⁻¹ at establishment provided faster establishment than the 24.5 kg N ha⁻¹ treatment; however, treatments receiving 73.5–122.5 kg N ha⁻¹ during the first 8 wk after planting provided a similar rate of establishment as 49 kg N ha⁻¹. Inclusion of clover with fine fescue resulted in slower establishment than fertilizer levels of ≥24.5 kg N ha⁻¹.

Abbreviations

DAP

d after planting

MAP

mo after planting

1 INTRODUCTION

Five fine fescue (Festuca L. spp. and ssp.) taxa used in turfgrass systems are collectively known as the grouping “fine fescues.” The five taxa include strong creeping red fescue (Festuca rubra L. ssp. rubra Gaudin), slender creeping red fescue [F. rubra L. ssp. littoralis (G. Mey.) Auquier], Chewings fescue [F. rubra L. ssp. commutata Gaudin; syn. F. rubra L. ssp. fallax (Thuill.) Nyman], hard fescue (Festuca brevipila Tracey), and sheep fescue [Festuca ovina L.; syn. F. ovina L. ssp. hirtula (Hack. ex Travis) M.J. Wilk.] (Braun et al., 2020b). Overall, fine fescues are generally regarded as “low-input” because once established, they can be successfully managed at low-to-moderate annual nitrogen (N) fertilization rates (i.e., ≤98 kg N ha⁻¹) (Braun et al, 2020a; Braun et al., 2021c; Hugie & Watkins, 2016; Watkins et al., 2011, 2014), which demonstrate one of the low-input characteristics of these turfgrasses. In addition, fine fescues can maintain acceptable turfgrass quality and better resist summer annual grassy and perennial broadleaf weed invasion than other turfgrass species, such as tall fescue [Festuca arundinacea Schreb.; syn. Schedonorus arundinaceus (Schreb.) Dumort., nom. cons.] or Kentucky bluegrass (Poa pratensis L.), when maintained at low-to-moderate annual N fertilization levels (49 to 98 kg N ha⁻¹) (Askew et al., 2013; DeBels et al., 2012; Dernoeden et al., 1994). Several researchers (Bourgoin, 1997; Skogley & Ledeboer, 1968) have reported minimal improvement in fine fescue turf quality when greater than 98 kg N ha⁻¹ yr⁻¹ was applied. Similarly, Braun et al. (2021a, 2021b) reported minimal-to-no differences in turf cover during establishment with increasing N fertilizer rates (from 98 to 294 kg N ha⁻¹ yr⁻¹) for either strong creeping red fescue or Chewings fescue and also increasing N fertilizer rates had no influence on fine fescue turfgrass quality at multiple sod harvest intervals.

Overall, fine fescue taxa have an intermediate establishment vigor (i.e., combination of seed germination, seedling emergence, and initial growth rate) compared with other turfgrass species (Braun et al., 2021a; Braun, unpublished data, 2021) and while there is variation among the fine fescue taxa, it is not well documented (Braun et al., 2020b). A more rapid establishment (i.e., better establishment vigor) can improve weed invasion resistance (Bertin et al., 2009; Braun et al., 2021a, 2021c). Additional research would allow for a better understanding of best establishment practices for fine fescues to overcome an intermediate establishment vigor, which is important because it can influence the plant's competitiveness with weeds, thus reducing the need for herbicide inputs (Harper, 1961; Ross & Harper, 1972).

One well-known way to enhance establishment vigor is providing N and phosphorus (P) fertilization at seeding or during establishment. However, although the effects of N and P fertilization rates at seeding has been investigated for other cool-season species, such as Kentucky bluegrass (Reicher et al., 2000), research is lacking for fine fescues (Braun et al., 2020b). Chang et al. (2014) reported the germination and establishment of strong creeping red fescue showed minimal-to-no response to the application of P or increasing P levels unlike Kentucky bluegrass and creeping bentgrass (Agrostis stolonifera L.). Similarly, Braun et al. (2020a) reported application of N (32 kg N ha⁻¹) and P (22 kg P ha⁻¹) from a starter fertilizer had minimal effect on the rate of establishment for a fine fescue mixture containing 25% (by weight) strong creeping red fescue, 25% Chewings fescue, 25% hard fescue, and 25% slender creeping red fescue; however, greater turf cover compared with no fertilizer was observed at 16 wk after planting. Overall, fine fescues may possibly require less soil P than other turf species to establish (Braun et al., 2020a, 2020b; Chang et al., 2014).

Although fine fescues can be successfully managed as low-input turf in the long-term at low-to-moderate annual N fertilization rates (≤98 N ha⁻¹ yr⁻¹), further research is needed to discover if establishment vigor may potentially be improved with increasing N rates during establishment to promote faster grow-in and greater probability of long-term success at low-input sites. This is an area where research is lacking. Bourgoin (1997) observed minimal responses in turf quality from increasing N fertilizer levels within multiple fine fescue taxa; however, this experiment was conducted on well-established turf. Would a greater response be observed when increasing N fertilizer levels during establishment since the turf sward is young, actively growing, and tillering? Furthermore, would the response to increasing N fertilizer levels be similar among fine fescue taxa or more apparent in slower establishing fine fescue taxon?

Inclusion of a N-fixing legume species at establishment could be a potential alternative to N fertilization; however, few experiments have examined the influence of cover crops on turfgrass establishment. Including a low-growing legume, such as white clover (Trifolium repens L.), would also support the growing interest in “flowering bee lawn” turf mixtures (turfgrass with a flowering forb) (Ramer & Nelson, 2020; Ramer et al., 2019). Cool-season flowering bee lawn mixtures typically contain turfgrasses with low-growing forbs, such as white clover, common selfheal (Prunella vulgaris L.), creeping thyme (Thymus serpyllum auct. non. L.), and groundplum milkvetch (Astragalus crassicarpus Nutt.) to name a few (Wolfin, 2020). In addition, fine fescues, specifically hard fescue, or fine fescues included in cool-season species mixtures have shown to be promising companion species for forb/turfgrass interseeding (Bigelow et al., 2021; Lane et al., 2019).

Overall, there is a need for further investigation to gain a better understanding on (a) establishment vigor differences among fine fescue taxa, (b) influence of increasing N fertilizer levels during establishment, and (c) influence of N-fixing legume inclusion within fine fescue taxon during establishment. Therefore, our objective was to investigate differences among fine fescue taxa and determine optimal N fertility or clover-inclusion programs for fine fescue taxa during establishment for future low-input sites.

Core Ideas

• Festuca brevipila has a slower establishment rate than the three Festuca rubra subspecies.
• Fine fescues should receive 24.5–49 kg N ha⁻¹ during the first 8 wk after planting.
• Applying >49 kg N ha⁻¹ during the first 8 wk provides negligible benefits.
• Clover-inclusion in fine fescue seed mixtures may slow establishment.

2 MATERIALS AND METHODS

2.1 Experimental site design and management

Field experiments were repeated across 2 yr at two locations (Indiana and Oregon) and 1 yr at a third location (Minnesota), providing five replications across time and location (Table 1). These experiments were initiated at all three locations in autumn 2019 and two locations in autumn 2020 in full sun. These locations represent cool-season or northern transitional climatic zones.

TABLE 1. Location and site details

Variable	Indiana	Oregon	Minnesota
Location, city	W.H. Daniel Turfgrass Research and Diagnostic Center, West Lafayette	Lewis-Brown Horticulture Farm, Corvallis	Minnesota Agricultural Experiment Station, St. Paul
Latitude and longitude	40°26′31″ N; 86°55′47″ W	44°32′52″ N; 123°12′57″ W	44°59′22″ N; 93°10′34″ W
Soil type	Silty-clay loam	Silty-clay loam	Silt loam
Soil pH	6.7	6.3	5.7
Organic matter (g kg^−1)	44^a, 66^b	67	68
Soil phosphorus (mg kg^−1)^c	58 ^a, 63^b	64	120
Soil potassium (mg kg^−1)	193^a, 300^b	349	372
Seeding date(s)	5 Sept. 2019 12 Sept. 2020	2 Oct. 2019 25. Sept. 2020	30 Aug. 2019

• ^a Experimental year (2019–2020).
• ^b Experimental year (2020–2021).
• ^c Bray-1 extraction method at Indiana and Minnesota and Mehlich-3 extraction method at Oregon.

Experimental design was a two-way factorial, randomized, complete-block design with four complete blocks totaling 128 plots measuring 1.2 × 1.2 m per plot. The species main effect included four fine fescue taxa: (a) ‘Cardinal II’ strong creeping red fescue; (b) ‘Seamist’ slender creeping red fescue; (c) ‘Radar’ Chewings fescue; and (d) ‘Beacon’ hard fescue. The fertility and cover crop program main effect included eight N fertilization or cover crop treatments (Table 2). Six of the eight fertility and cover crop program treatments investigated the effects of increasing levels of N fertilization rates during establishment, while the cover crop treatments investigated the effects of inclusion of clover without N. The two cover crop treatments consisted of (a) an annual clover, ‘Frosty’ berseem clover (Trifolium alexandrinum L.), and (b) a perennial clover, ‘Aberlasting’ white clover × kura clover (also known as caucasian clover) (Trifolium ambiguum M. Bieb.) hybrid.

TABLE 2. Nitrogen (N) fertilization and cover crop program treatment levels in this experiment

Fertility and cover crop main effect	Application notes
Nontreated (0 kg N ha^−1)	–
24.5 kg N ha^−1	24.5 kg N ha^−1 at seeding^a
49.0 kg N ha^−1	49.0 kg N ha^−1 at seeding^a
73.5 kg N ha^−1	24.5 kg N ha^−1 at seeding^a + 49.0 kg N ha^−1 at 4 WAP^b
98 kg N ha^−1	49.0 kg N ha^−1 at seeding^a + 49.0 kg N ha^−1 at 4 WAP^b
122.5 kg N ha^−1	49.0 kg N ha^−1 at seeding^a + 49.0 kg N ha^−1 at 4 WAP^b + 24.5 kg N ha^−1 at 8 WAP^b
Annual clover cover crop + 0 kg N ha^−1	‘Frosty’ berseem clover^c seeded at a rate of 1.1 g m^−2 at fine fescue seeding
Perennial clover companion cover crop + 0 kg N ha^−1	‘Aberlasting’ White × kura clover^c seeded at a rate of 1.1 g m^−2 at fine fescue seeding

• ^a A quick-release N fertilizer (Urea; 46-0-0) was utilized for all N fertilizer treatments at seeding.
• ^b A slow-release N fertilizer (XRT N-3 Polymer-Encapsulated Urea; 44-0-0; SGN:150, Knox Fertilizer Co., Inc.) was utilized for all N fertilization treatments applied at 4 and 8 wk after planting (WAP).
• ^c Inoculated with Sinorhizobium meliloti and Rhizobium leguminosarum biovar trifolii (N-DURE, Verdesian Life Science U.S.) prior to seeding.

Pure live seeding rate was calculated ≤4 wk prior to each field experiment following germination tests according to the Association of Official Seed Analysts principles and procedures (Association of Official Seed Analysts, 2016). Prior to seeding, the soil was tilled at each site and a starter fertilizer of triple superphosphate (0-46-0) was applied at a rate of 22 kg P ha⁻¹ across the entire study. At each respective seeding date for each experimental run listed in Table 1, fine fescue taxa treatments were seeded at a rate of 2 pure live seed cm⁻², which corresponded to seeding rates of 30–34 g m⁻² for strong creeping red fescue, 25–27 g m⁻² for slender creeping red fescue, 25–28 g m⁻² for Chewings fescue, and 24 g m⁻² for hard fescue. Both clover treatments were inoculated with Sinorhizobium meliloti and Rhizobium leguminosarum biovar trifolii (N-DURE, Verdesian Life Science U.S., LLC) and then mixed with each respective fine fescue treatment above immediately prior to seeding and seeded at a clover rate of 1.1 g m⁻². All seeds were planted with hand-shakers in multiple directions across the individual plots. The N fertility treatments were also applied with hand-shakers at the respective timing listed in Table 2. After seeding, the soil surface was lightly raked to promote seed-to-soil contact. A germination blanket (PR1724, A.M. Leonard) covered the experiment area for 7 to 12 d to control erosion, prevent seed movement, and promote seedling germination and emergence. Irrigation was applied daily or four to five times per week for the first 2 wk after seeding to promote turf establishment. Plots then received an average of 2.5 cm wk⁻¹ during the growing season from either irrigation or precipitation for the remainder of the study. No pesticides or additional fertilizer were applied during the experiment. Mowing was performed when needed at a height of 7.6 cm with a rotary mower with clippings mulched and returned to the surface.

2.2 Data collected

Daily air and soil temperature (5-cm depth) and precipitation were collected from an on-site weather station at each location.

At each site, digital images (size 4,000 × 3,000 pixels) were collected with a lighted camera box, described by Heineck et al. (2020), every 7 or 14 d until 70 to 90 d after planting (DAP), and after winter on 1 May (∼8 mo after planting [MAP]), and 1 June (∼9 MAP) and analyzed with ImageJ (hue: 47–150; saturation: 0–255; brightness: 10–255) (v.1.52a; Schneider et al., 2012) to assess green vegetation cover (0-100%). Using the methods of Braun et al. (2020a), fine fescue establishment rate was calculated from digital image analysis regressed against DAP by GraphPad Prism (v.9.0, GraphPad Software Inc.) and a sigmoidal model. Days until 90% establishment were calculated from the regression curves.

Grid counts for percentages of fine fescue cover were collected after winter on 1 May and 1 June in 2020 or 2021 for respective experiments. On these dates, estimates of fine fescue turf were taken using a modification of the vertical point quadrat method (Braun et al., 2020a; Gaussoin & Branham, 1989), where a 1.0 by 1.0 m frame was laid over the plots with an interval filament grid of at least 100 intersections. The total number of times fine fescue was present under each intersection was recorded for each plot, and then percent fine fescue cover was calculated.

Additional data included visual fine fescue cover (0-100%), visual clover cover (0-100%), visual winter annual weed or annual bluegrass (Poa annua L.) cover (0-100%), and visual turfgrass quality on a 1-to-9 scale, where 1 = poorest quality, 6 = minimally acceptable, and 9 = highest quality) according to color, texture, density, and uniformity was collected monthly in the autumn during establishment and then on 1 May and 1 June the following year of each experiment.

2.3 Data analysis

All data were analyzed using the GLIMMIX procedure of SAS version 9.4 (SAS Institute Inc., Cary, NC). Residual normality was tested using the UNIVARIATE procedure of SAS. Raw data, such as days until 90% establishment and other output from the regression curves were log transformed prior to analysis to normalize the variance. Analysis was combined across all five environments (experiments) because our central objective was to estimate treatment effects (Blouin et al., 2011). Therefore, location and block (nested within location) were considered random effects and treatment by location effects were excluded from the model to increase statistical power (Blouin et al., 2011). Means were separated with Fisher's LSD test (α = .05) when the F-tests were significant (P ≤ .05), and transformed means were back-transformed for presentation. Correlation analyses between grid count data and visual turf cover in May and June were conducted using GraphPad Prism.

The relationship between days until 90% establishment and treatments were calculated using exponential decay model using GraphPad Prism via Equation [1]:

$$\begin{equation}y\; = \left( {{y_0} - Plateau} \right)\;{^{*}}\exp \left( { - K{^{*}}x} \right) + Plateau\end{equation}$$(1)

where y₀ (days) is the y value when x [mean fertility (kg N ha⁻¹)] is zero; Plateau is the y value at infinite fertility; expressed in the same units (d) as y; and K is the rate constant.

3 RESULTS AND DISCUSSION

3.1 Establishment rate

Fine fescue species had a significant effect on establishment rate (Table 3). Sigmoidal regression results indicate that the three F. rubra complex taxa (strong creeping red fescue, slender creeping red fescue, and Chewings fescue) were similar to one another and required the shortest amount of time until 90% establishment compared with hard fescue (Table 3). Research conducted prior to 1970 reported minimal differences in seedling or establishment vigor among strong creeping red fescue, Chewings fescue, or slender creeping red fescue (DeFrance & Simmons, 1951; Parks & Henderlong, 1967). Prior to this project, hard fescue has been anecdotally reported to be slower establishing compared with other fine fescue taxa, this may be due to smaller seed size, less seedling vigor, or other unknown factors (Braun et al., 2020b; Bertin et al., 2009). With newer cultivars in National Turfgrass Evaluation Program (NTEP) data from 1999–2002, establishment vigor of strong creeping red fescue and Chewings fescue were similar, but visually faster than hard fescue (Bertin et al., 2009). Results from this study clearly indicate establishment rate differences between fine fescue taxa, especially hard fescue compared with three subspecies of the F. rubra complex. The differences in rate of establishment between hard fescue and the other three fine fescues is also demonstrated by the hillslope values, which a higher hillslope value indicate more rapid establishment (Table 3). Our results are in agreement with Bertin et al. (2009) and Meyer and Funk (1989), who reported differences in establishment vigor among fine fescue taxa; however, there also has been reported variability among cultivars within fine fescue taxon (Bertin et al., 2009) and other turfgrass species (Bonos & Huff, 2013). Therefore, new release of fine fescue cultivars, as well as variable cultivar performance in differing environments (NTEP, 2014, 2018), necessitates further establishment research to continue to provide more information and understand inter- and intraspecies establishment differences among fine fescue taxa.

TABLE 3. Sigmoidal regression model^a results for the estimated number of days until 90% establishment of four turfgrass species and fertility and cover crop program treatments. Data combined across five experimental runs

Source of variation	Days until 90% establishment^a	Top (%)^a	Hillslope^a	R^2
Turfgrass species
Strong creeping red fescue	44.1 b	89.4 c	0.067 a	0.97
Slender creeping red fescue	42.0 b	91.9 a	0.067 a	0.98
Chewings fescue	43.5 b	91.0 ab	0.064 a	0.97
Hard fescue	50.7 a	90.3 b	0.058 b	0.97
Fertility and cover crop program
0 kg N ha^−1	55.8 a	87.6 d	0.052 d	0.97
24.5 kg N ha^−1	43.1 c	88.4 d	0.062 b	0.97
49.0 kg N ha^−1	39.7 d	90.0 c	0.069 a	0.98
73.5 kg N ha^−1	40.6 cd	90.8 bc	0.070 a	0.97
98 kg N ha^−1	40.7 cd	91.0 bc	0.069 a	0.97
122.5 kg N ha^−1	39.1 d	90.5 bc	0.070 a	0.97
Annual clover cover crop + 0 kg N ha^−1	52.1 ab	91.6 ab	0.057 cd	0.98
Perennial clover companion cover crop + 0 kg N ha^−1	49.4 b	93.1 a	0.061 c	0.97
ANOVA
Source	p-value
Species	<.0001	<.0001	<.0001	–
Fertility and cover crop program	<.0001	<.0001	<.0001	–
Species × fertility and cover crop program	.8710	.7702	.8446	–

• Note. Within a source of variation, means within a column followed by a common letter are not significantly different according to Fisher's LSD (α ≤ .05).
• ^a Sigmoidal regression model (four-parameter logistic equation) defined by Eq. [1]: $$y{\rm{\; = \;Bottom\; + \;}}\frac{{( {{\rm{Top}} - {\rm{Bottom}}} )}}{{1 + {{10}^{( {{\rm{LogECz}} - x} ) - {\rm{Hillslope}}}}}}$$ where y is percent green vegetation cover, bottom is the y value at the estimated bottom plateau, top is the y value at the estimated top plateau, LogECz is the x value when the response of either D₉₀, which is 90% of the top value (days until 90% establishment), and Hillslope (no units) describes the slope of the curve. Top (%) is the maximum green vegetation cover estimate and higher hillslope values indicate more rapid establishment. Raw regression model output data were logarithmic transformed prior to analysis to normalize. Means were separated using Fisher's LSD test, and transformed means were back-transformed for presentation.

The sigmoidal regression results also reveal that fertility and cover crop program had a significant effect on establishment rate (Table 3). Providing 24.5 kg N ha⁻¹ at seeding hastened establishment of fine fescues by almost 13 d compared with no N fertilizer (0 kg N ha⁻¹) (Table 3 and Figure 1). Moreover, applying 49 kg N ha⁻¹ at establishment, which received 24.5 kg N ha⁻¹ more at seeding compared with the 24.5 kg N ha⁻¹ treatment, had slightly faster establishment (3.4 d) than the 24.5 kg N ha⁻¹ treatment. While treatments receiving ≥ 73.5 kg N ha⁻¹ also provided faster establishment than non-fertilized turf and turf receiving 24.5 kg N ha⁻¹, none of these higher fertility treatments increased establishment rate compared with the 49 kg N ha⁻¹ treatment. Regression analysis indicated an inverse relationship between days until 90% establishment and the six fertility treatments (not including clover treatments) (Figure 1a), which also confirms that as N fertilizer amounts increased, especially from 0 to 49 kg N ha⁻¹, the number of days until 90% establishment decreased. Based on the relationship between fertility treatments and days until 90% establishment (Figure 1a), the days until 90% establishment when N is applied during establishment at 0, 24.5, 49, 73.5, 98, and 122.5 kg N ha⁻¹ are predicted to be 55.9, 42.9, 40.5, 40.0, 39.9, and 39.9 days after planting, respectively. Overall, increasing N fertility levels to 24.5 or 49 kg N ha⁻¹ during the first 8 wk after planting increased establishment of fine fescues, and N levels higher than 49 kg N ha⁻¹ provided negligible benefits. Furthermore, Figure 1b displays the effects of six fertility treatments (not including clover treatments) on percentage of green vegetation cover as time advances from 28 to 42 or 70 DAP. Based on scatter plots (Figure 1a) and standard error of the mean (Figure 1b), there is less variability in treatment means when increasing N fertility above 24.5 kg N ha⁻¹, which may provide greater probability of a successful establishment of fine fescues when planted at low-input sites (Figure 1).

FIGURE 1

The relationship between days until 90% establishment and six fertility program treatments (not including clover treatments) combined across all five experiments calculated using an exponential decay model using Graphpad Prism (a). The independent variable was fertility program treatments with days until 90% establishment as the dependent variable. All observations are shown. Effects of six fertility program treatments (not including clover treatments) on percentage of green vegetation cover at 28, 42, and 70 d after planting across all experiments (b). Treatments include (1) 0 kg N ha⁻¹ (nontreated, no fertilizer applied), (2) 24.5 kg N ha⁻¹ (applied at seeding), (3) 49 kg N ha⁻¹ (applied all at seeding), (4) 73.5 kg N ha⁻¹ (24.5 kg N ha⁻¹ at seeding + 49 kg N ha⁻¹ at 30 d after planting [DAP]), (5) 98 kg N ha⁻¹ (49 kg N ha⁻¹ at seeding + 49 kg N ha⁻¹ at 30 DAP), 122.5 kg N ha⁻¹ (49 kg N ha⁻¹ at seeding + 49 kg N ha⁻¹ at 30 DAP + 24.5 kg N ha⁻¹ at 60 DAP). Error bars represent the standard error of the mean. Dashed red line signifies 90% green vegetation cover

When a quick-release N fertilizer (i.e., urea) was utilized at seeding, followed by slow-release N fertilizer (i.e., polymer-coated urea) for all additional N fertilizer applications at 4 and 8 wk after planting, then results demonstrate that applying greater than 49 kg N ha⁻¹ during the first 8 wk provided negligible establishment benefits with fine fescue taxa. An enhanced N fertility effect may have been observed if a quick-release N fertilizer was utilized at all application timings instead of a slow-release N fertilizer; however, further research is required on this topic. Considering all the sites had high soil organic matter (≥44 g organic matter kg⁻¹), the results from this study may not necessarily apply/translate to sites with poor soils, lacking organic matter. At sites with poor soils, additional N (>49 kg N ha⁻¹) during the first 8 wk after planting could prove beneficial.

As far as N nutrition, there has been little research conducted on the effect of N fertilization on establishment of fine fescues (Braun et al., 2020b). Research on fully-established fine fescues showed a minimal response in turf quality from increasing N fertilizer levels within each fine fescue taxa (Bourgoin, 1997). Bourgoin (1997) reported differences between and within fine fescues for N use efficiency, especially at lower N levels, which were ranked from most to least efficient as follows: hard fescue > strong creeping red fescue > Chewings fescue ≥ slender creeping red fescue. However, no species × fertility and cover crop program interaction was observed in the current study, indicating all four species of fine fescue taxa responded similarly to the N fertility treatments during establishment.

Inclusion of a clover as a cover or companion crop with fine fescues decreased rate of establishment compared with applying ≥ 24.5 kg N ha⁻¹, which is indicated by greater number of estimated days until 90% establishment and smaller hillslope values (Table 3). While a similar establishment rate was observed between the annual clover cover crop and nontreated (0 kg N ha⁻¹) fine fescue plots, perennial clover included with fine fescues provided slightly faster time required until days until 90% vegetation establishment and greater hillslope values than the nontreated (0 kg N ha⁻¹) fine fescue plots.

Incorporating legumes (e.g., white clover or other clovers) in turf and other crops provides the ability to utilize atmospheric N and improve soil N levels and reduce the dependency on supplemental N fertilization (Evers, 2011; Ledgard & Steele, 1992; Russelle, 2008). The amount of dinitrogen (N₂) fixed by legumes depends on many factors, such as soil type, climate, Rhizobia strain, and soil nutrient status to name a few (Russelle, 2008). While a pure legume stand fixes more N than a grass-legume forage mixture because of less plant, water, and nutrient competition (Evers, 2011), grass–legume mixtures have been reported to fix between 25 and 250 kg N ha⁻¹ yr⁻¹ (Ledgard & Steele, 1992; Giller, 2001; Russelle, 2008; McCurdy et al., 2014; McNeil & Wood, 1990). In addition, Sincik and Acikgoz (2007) estimated white clover fixed greater than 250 kg N ha⁻¹ yr⁻¹ when growing with Kentucky bluegrass, perennial ryegrass, or creeping bentgrass (Agrostis stolonifera L.). Furthermore, Jørgensen et al. (1999) reported white clover mixed with perennial ryegrass (Lolium perenne L.) fixed 23, 187, and 177 kg N ha⁻¹ during the seedling (Year 1 or “grow-in”), first, and second production years, respectively, whereas a pure stand of white clover fixed 28, 262, and 211 kg N ha⁻¹ during seedling, first, and second production years, respectively, in a sandy loam soil in Denmark. McCurdy et al. (2014) also reported an increase in N fixation as years progressed. After fixation, N is transferred from the legume to associated grass in legume/grass mixture by (a) death or decay of legume herbage, roots, and nodules; (b) N excretion from legume roots and nodules; and (c) hyphal links to non-legume roots via arbuscular mycorrhizal fungi (Butler et al., 1959; Dubach & Russelle, 1994; Haystead et al., 1988; McCurdy et al., 2013; Sincik & Acikgoz, 2007; Ta et al., 1986). This amount of transfer of N from legumes to associated grasses in mixtures varies with reported ranges of 4.2 to 42%, N transferred (i.e., 16 to 43 kg N ha⁻¹ yr⁻¹) from white clover to various turfgrass species (McCurdy et al., 2014; McNeil & Wood, 1990; Sincik & Acikgoz, 2007). Transfer of N has shown to increase with stand age and estimations of the amount of N transfer during the first year can be small likely due to less N fixation during establishment and less decomposition of above- and belowground tissue in the first year (Dubach & Russelle, 1994; Jørgensen et al., 1999; Russelle, 2008).

Our hypothesis was that the two treatments containing clover would benefit the fine fescue turf by the fixation and transfer of N from the legume to the turf during establishment. However, days until 90% establishment and hillslope values indicate the two clover-inclusion treatments had similar or only slightly better establishment rates than the nonfertilized (0 kg N ha⁻¹) fine fescue turf (Table 3). Furthermore, the two clover-inclusion treatments did not hasten establishment when compared with fertility treatments of ≥24.5 kg N ha⁻¹ indicated by both days until 90% establishment and hillslope values. This was likely due to the small fixation and transfer amounts of N from legumes to grasses in establishing swards, low legume populations (Table 4), or minimal decomposition of plant tissue (Jørgensen et al., 1999; McCurdy et al., 2014; Russelle, 2008).

TABLE 4. Effects turfgrass species and fertility and cover crop programs during establishment in autumn on fine fescue turfgrass cover by grid counts, visual clover cover, and visual annual bluegrass cover in May and June (∼8 and 9 mo after planting, respectively). Data combined across four experimental runs^a

	May			June
Source of variation	Grid counts of fine fescue turf^b	Clover	Annual bluegrass	Grid counts of fine fescue turf	Clover	Annual bluegrass
	% cover^b
Turfgrass species
Strong creeping red fescue	75 b	5	4.7 b	81 b	7 ab	4.2 b
Slender creeping red fescue	80 a	4	4.2 b	86 a	4 c	2.7 b
Chewings fescue	80 a	5	4.1 b	87 a	5 bc	3.0 b
Hard fescue	66 c	6	12.4 a	74 c	8 a	11.8 a
Fertility and cover crop program
0 kg N ha^−1	76 a	2 c	6.1 abc	83 ab	5 b	3.3 de
24.5 kg N ha^−1	77 a	1 c	6.6 abc	84 ab	3 bcd	4.6 cde
49.0 kg N ha^−1	76 a	1 c	6.8 ab	82 b	3 bcd	4.7 cde
73.5 kg N ha^−1	78 a	0 c	7.2 a	84 ab	2 cd	6.8 abc
98 kg N ha^−1	79 a	0 c	7.6 a	84 ab	2 cd	7.4 ab
122.5 kg N ha^−1	78 a	0 c	7.7 a	83 ab	1 d	8.0 a
Annual clover cover crop + 0 kg N ha^−1	75 a	12 b	4.4 c	85 a	4 bc	5.6 bcd
Perennial clover companion cover crop + 0 kg N ha^−1	61 b	24 a	4.5 bc	73 c	25 a	3.0 e
ANOVA
	p-value
Source
Species	<.0001	.1212	<.0001	<.0001	.0015	<.0001
Fertility and cover crop program	<.0001	<.0001	.0349	<.0001	<.0001	<.0001
Species × Fertility and cover crop program	.9637	.0120	.9959	.4830	.2851	.3233

• Note. Within a source of variation, means within a column followed by a common letter or no letters are not significantly different according to Fisher's LSD (α ≤ .05).
• ^a Although grid count and visual cover data in May and June 2020 was collected in Oregon for the 2019 planting experiment, this site experienced uncharacteristic annual bluegrass infestation of plots during the 2019–2020 winter; therefore, this site was not included in analysis of May and June data. Both Indiana experiments (2019–2020, 2020–2021), Minnesota (2019–2020), and Oregon (2020–2021) are included in analysis.
• ^b 1.5 × 1.5 m frame was laid over the plots with an interval filament grid of at least 100 intersections within a 1.2 × 1.2 m area in May and June and the total number of times fine fescue was present under each intersection was recorded for each plot, and then percent fine fescue cover was calculated. Visual fine fescue cover (0–100%; data not shown), visual clover cover (0–100%), and visual annual bluegrass cover (0–100%) were also collected.

A companion study to investigate the influence of these fine fescue taxa and fertility treatments on sod production immediately following the conclusion of each experimental run in Indiana was conducted in this same experiment (Braun & Patton, 2022). Braun and Patton (2022) reported differences in sod strength due to species (i.e., fine fescue taxon), and increasing levels of N fertility during the first 8 wk of establishment of four fine fescue taxa or inclusion of either an annual or perennial clover will have minimal effect on sod strength and handling when harvesting sod from 9 to 12 mo after planting.

3.2 Final cover

Fine fescue, clover, and annual bluegrass cover based on grid counts or visual estimations indicated treatment differences for May and June ratings (Table 4). The percentage of fine fescue within each respective species plots ranked as follows: Chewings fescue = slender creeping red fescue > strong creeping red fescue > hard fescue at both May and June rating dates, 8 and 9 MAP, respectively (Table 4). When pooled over four experiments, there was a strong correlation between and grid count data of fine fescue turf and visual fine fescue turf cover ratings for both May (r = 0.92, P < .0001, n = 512) and June (r = 0.94, P < .0001, n = 512) rating dates, which suggests both the visual estimations were a good indicator for the percentage of fine fescue.

There were no differences in the percentage of clover by May within the turfgrass species main effect; however, a significant species × fertility and cover crop program interaction occurred (Table 4) with the percentage of clover greatest in the hard fescue × perennial clover treatment. By June, hard fescue had a greater percentage of clover within plots compared with Chewings fescue and slender creeping red fescue and similar percentage of clover within plots as strong creeping red fescue. By June, there was minimal-to-no differences in clover cover among fertility treatments (0–122.5 kg N ha⁻¹) and the annual clover cover crop treatment, and plots that contained inclusion of perennial clover as a companion crop at seeding resulted in higher clover cover (25% clover cover) than all other fertility and cover crop program treatments. Lane et al. (2019) also reported that hard fescue is a good companion grass with clover, but they did not test F. rubra taxa. Our results suggest that strong creeping red fescue, slender creeping red fescue, and Chewings fescue can also serve as good companion grasses for floral or flowering bee lawns.

Annual bluegrass invasion is one of the problematic weed pests of fine fescues (Braun et al., 2020b). The faster establishment of the three F. rubra taxa compared with hard fescue resulted in similar annual bluegrass cover (≤ 5%) among three subspecies and less annual bluegrass invasion than hard fescue (12%) at both rating dates (Table 4). There were differences, albeit minor, in annual bluegrass cover among fertility and cover crop program treatments at both rating dates. Specifically, plots receiving ≥73.5 kg N ha⁻¹ had higher annual bluegrass than either clover treatment in May. By June, plots that had received ≥73.5 kg N ha⁻¹ during establishment had greater annual bluegrass cover than non-fertilized (0 kg N ha⁻¹) or perennial clover inclusion plots. As mentioned, there was only a slight increase in rapidity of establishment once fertility levels increased above 24.5 kg N ha⁻¹ and there were no differences in speed of establishment between treatments receiving 49–122.5 kg N ha⁻¹ (Table 4). Therefore, increasing N fertility above 49 kg N ha⁻¹ during establishment in fine fescues during autumn may likely start to become more detrimental than beneficial, especially with higher N fertility (73.5 to 122.5 kg N ha⁻¹), at which point we observed no beneficial increase in establishment rate during the autumn but higher annual bluegrass invasion the following summer. Further, final turf coverage 8 and 9 MAP was similar among N fertility treatments (not considering clover treatments) (Table 4), which demonstrated that gains in initial establishment within the first 56 days after planting (Table 3) were short-lived.

4 CONCLUSIONS

Results from this study provide information on how to increase the likelihood of success when seeding fine fescues, specifically strong creeping red fescue, slender creeping red fescue, Chewings fescue, and hard fescue. In agreement with Bertin et al. (2009) and Braun et al. (2020b), results indicate there are differences among fine fescue taxa, especially in relation to establishment rate differences. Hard fescue has a slower establishment rate than strong creeping red fescue, slender creeping red fescue, and Chewings fescue, which the latter three are similar to one another. Further research on inter- and intraspecies establishment differences among fine fescue taxa is warranted.

Establishment rate of all four fine fescue taxa improves with the application of 24.5 to 49 kg N ha⁻¹ during the first 8 wk after planting, and N levels higher than 49 kg N ha⁻¹ during the first two MAP provides negligible benefits in establishment rate and final turf cover and may likely increase invasion of annual bluegrass when seeded during autumn. Inclusion of either annual or perennial clover with fine fescue provides similar or slightly faster establishment than the nonfertilized (0 kg N ha⁻¹) fine fescue turf, but clover-inclusion turf did not hasten establishment compared with fine fescue turf receiving ≥ 24.5 kg N ha⁻¹.

In summary, these results combined with recent literature by Braun et al., 2020a, 2021a, 2021c) can provide best management practices for fine fescue establishment as follows. When seeded in autumn, the rate of establishment of strong creeping red fescue and Chewings fescue is similar to tall fescue, and faster than Kentucky bluegrass (Braun et al., 2021a). The optimum seeding timing for fine fescue mixtures is in the late-summer or early-autumn months of August and September in cool-season climate zones in the United States, with a wider seeding window in Mediterranean climates (Braun et al., 2021c). Fine fescues may show a minimal-to-no response to the application of P at seeding, especially on high-quality topsoil (Braun et al., 2020a). Application of N during the first 8 wk after planting fine fescue seed should range from 24.5 to 49 kg N ha⁻¹ in soils with high organic matter and N fertilizer levels above 49 kg N ha⁻¹ during the first 8 wk after planting may become more detrimental than beneficial in the long-term. Sites with poor soils, lacking organic matter, may see a greater effect from additional N (>49 kg N ha⁻¹) during the establishment of fine fescues; however, further researched is needed. Lastly, inclusion of clover with fine fescue may slow the rate of establishment but does not reduce final establishment of strong creeping red fescue, slender creeping red fescue, Chewings fescue, or hard fescue. Overall, results indicate it may be beneficial to provide more N fertilization (24.5 to 49 kg N ha⁻¹) during the first 8 wk of establishment to promote a faster grow-in of fine fescue taxa at low-input sites and then N fertilizer levels can be reduced to a low fertility program (24.5 to 98 kg N ha⁻¹ yr⁻¹) in the long term.

ACKNOWLEDGMENTS

The authors wish to acknowledge the funding support by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Specialty Crop Research Initiative under award number 2017-51181-27222. We also thank the Knox Fertilizer Company (Knox, IN) for supplying fertilizer, and Mountain View Seeds (Salem, OR) and Grassland Oregon Seed (Salem, OR) for supplying seed for this research.

AUTHOR CONTRIBUTIONS

Ross C. Braun: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualization; Writing-original draft; Writing-review & editing. Emily T. Braithwaite: Investigation; Validation; Writing-review & editing. Alexander Robert Kowalewski: Funding acquisition; Project administration; Resources; Writing-review & editing. Eric Watkins: Funding acquisition; Project administration; Resources; Writing-review & editing. Andrew B. Hollman: Investigation; Validation; Writing-review & editing. Aaron J. Patton: Conceptualization; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Visualization; Writing-review & editing.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.