Overseeding of fairways

Transcription

1 Overseeding of fairways Anne Mette Dahl Jensen 1, Oliver Bühler 1a, Trygve Aamlid 2, Agnar Kvalbein 2 1 Principal investigator University of Copenhagen, Department of Geosciences and Natural Resource Management Rolighedsvej 23, DK-1958 Frederiksberg C, Denmark amdj@ign.ku.dk 1a University of Copenhagen, Department of Geosciences and Natural Resource Management 2 Norwegian Institute for Bio economy Research, Section for Urban Greening and Environmental Technology NIBIO, Norway Final report December 2015

2 Table of content Table of content... 2 Summary... 3 Introduction... 4 Turfgrass species and varieties for overseeding... 4 Objectives... 6 Hypothesis... 6 Materials and Methods... 7 Landvik trial... 7 Registrations Landvik trial... 8 Trial at 3 Danish golf courses... 9 Registrations Danish trial Statistical analyses Results and Discussion Introduction of new grass species by overseeding Overseeding in Denmark Overseeding in Norway Factors affecting grass introduction Grass occurrence When to overseed Appearance data Broadleaved weed occurrence Denmark Norway Weed and overseeding Cost of overseeding Conclusions and perspectives Dissemination of results Dissemination in Acknowledgement References... 37

3 Summary Research concerning management of golf course fairways has so far been neglected compared to golf greens. Legislations regarding pesticide reduction in the Nordic countries will cause serious problems with annual bluegrass (Poa annua) and broadleaved weeds in years to come. Establishing a strong competitive turfgrass sward might be one solution in order to avoid massive invasion of weeds. This can be accomplished with an appropriate overseeding strategy adapted to the Nordic countries. Today, most knowledge regarding overseeding is related to Bermuda grass (Cynodon dactylon) fairways in the USA. Therefore, more information regarding the optimal overseeding strategy in the Nordic countries is crucial in order to create a strong fairway with a minimum of annual bluegrass and broadleaved weeds. The aim of the project was to test a variety of parameters that might influence the outcome of an overseeding procedure. This includes overseeding date (spring or autumn), fertilisation when overseeding and turfgrass species; perennial ryegrass (Lolium perenne), red fescue (Festuca rubra) and Kentucky bluegrass (Poa pratensis) sown as single species. The experiment was performed on an experimental fairway at Landvik (NIBIO, Norway) and on 3 golf courses in Denmark (Korsør, Skovbo and Hornbæk) for the period and with registrations in Registration was on grass species composition and on weed occurrence (covering degree of the different weed species) and in the Norwegian trial visual appearance was scored. Additionally the economic costs of the overseeding procedure was estimated for a standard golf course (18 hole course). The main findings were that perennial ryegrass was established, both on the three Danish golf courses where no watering and topdressing was applied after overseeding and in the Norwegian experimental fairway where water and topdress was applied after the overseeding. There is no perfect time for overseeding. It can be performed in spring or in fall. However there is a tendency, in the Norwegian experimental fairway that an autumn overseeding is more successful especially with the fertiliser but at all time and with and without fertiliser perennial ryegrass is introduced. In the Norwegian experimental fairway red fescue could also be successful introduce. On the 3 golf course fairways in Denmark this could not be demonstrated. For Kentucky bluegrass there is no indication of an establishment after 3 years of overseeding in Denmark or Norway. The amount of weed is not reduced by the overseeding procedure over the 3 year period. However this might not indicate that there isn t an effect on weed in the long run. A greenkeeper experience from another overseeding trial in Denmark years ago indicates that the effect on weed is more pronounced after approx years. Registrations of appearance, density etc. in the Norwegian trial does not demonstrate a significant effect of any of the treatments. A small increase in appearance and density is seen shortly after the spring overseeding in the perennial ryegrass plots, but the effect is not pronounced over the growing season.

4 Introduction The EU legislation regarding pesticide use, the Danish pesticide agreement, and a more extreme climate will create many challenges on the golf courses. Most focus has been on greens but serious problems will occur also on fairways. More and more broadleaved weeds are seen as herbicide use is decreased. The Danish experience so far is that many courses can cope with the weeds for 3-4 years after an herbicide ban is introduced, but then problems escalate to levels severely reducing not only the aesthetic quality, but also the playing quality (ball lie) on the fairways. Unfortunately, experiences from an on-going STERF project do not indicate that the weed problem can be controlled by mechanical treatments alone. Another, just as serious problem, is the encroachment of annual bluegrass on fairways. This problem seems to increase with increasing pressure from players (longer seasons, more golf rounds), heavier machinery and warmer, wetter and less stable winters due to climate change. Among the consequences of this encroachment is less drought tolerance in summer, more physical winter damage and more diseases, especially Microdochium nivale. Since fairways make up such a high proportion of the total golf course area, a complete strategy for less pesticide use and integrated pest management on golf courses must include fairways, not only the greens and tees. Fairway overseeding is a management technique not often used in the Nordic countries. It might, however, be the solution, or at least part of the solution, to the broadleaved weed and Poa annua problems. This management technique helps to replace unwanted grasses, or it is used as at strategy to improve recovery after intense wear. Overseeding is also important for introduction of new and improved species/varieties into the turf. Improved characters can be improved disease resistance, better drought tolerance, higher shoot density etc. Most scientific data on fairway renovation by overseeding comes from the transition zone in the USA where Bermuda grass fairways are often overseeded with rough bluegrass (Poa trivialis), annual ryegrass (Lolium multiflorum) or perennial ryegrass (Lolium perenne) (Morris, 2004; McCarthy, 2009). A quick search in the American turfgrass information database showed 120 references with the combination of fairway and overseeding in the title, however, more than 100 of these references were related to optimal use of pesticides or chemical growth regulators during the overseeding process. A single experiment from the UK demonstrated a pronounced effect of overseeding (Harris, 2008). Observations on Viborg Golf Course, Denmark, showed a conspicuous effect of overseeding. Even ten years after a large-scale demonstration trial was finished, plots which included overseeding had more perennial grasses and less weeds compared to plots with no overseeding (Nyholt, 2010). Turfgrass species and varieties for overseeding The choice of turfgrass-species for overseeding is mainly determined by the climatic zone and based on the initial species composition. Fairways in the Nordic countries are primarily established

5 with red fescue or a mixture of red fescue, perennial ryegrass and Kentucky bluegrass. A few courses have established American-style fairways with creeping bentgrass (Agrostis stolonifera). So far, only a few Danish courses have included overseeding of fairways in their management plan. Most of them do it only sporadically and infrequently. Red fescue, Kentucky bluegrass and perennial ryegrass are used for overseeding in different mixture combinations. Using Kentucky bluegrass for overseeding is questionable because it does not establish easily in a mixture (Larsen et al., 2004). This species has a poor establishment compared to red fescue and perennial ryegrass due to a larger thermal time required for germination (Larsen and Bibby, 2005). But the species is wanted on fairways due to its high recuperative capacity. Many new cultivars that tolerate mowing heights down to 12 mm are now coming to the market (Molteberg and Aamlid, 2007). Some courses favour red fescue as the overseeding grass species; however perennial ryegrass might be an option, too. Ryegrass germinates faster and has better shoot and root competition against weeds. This also includes competition for nitrogen (Haugland and Froud-Williams, 1999). Increasing the amount of ryegrass degreases the dry weight of voluntary seedlings such as Poa annua (Snaydon and Howe, 1986). Root competition is generally very important between established species and seedlings (Jeangros and Nösberger, 1990). Perennial ryegrass may be a good competitor in relation to Poa annua and other weeds. Red fescue is a reasonable choice if a course wants to reduce input such as fertilizer on fairways. However, unlike Kentucky bluegrass, and to a certain extent perennial ryegrass, red fescue has little seed dormancy, and must therefore be seeded every year. The effect of overseeding varies. Generally it is difficult to achieve an efficient establishment of the sown species because the new seedlings must compete with the established sward (Henriksson, 1991). To enhance the effect of overseeding the seed must come in contact with the soil. The seed must be placed in the growing layer and the grass must be sown when annual bluegrass and broadleaved weeds species that dominate on golf fairways have a reduced growth rate. (Vargas and Turgeon, 2004) Some of the factors that might affect the outcome of the overseeding are time of seeding and the amount of fertilizer. The most important fertilizer component in the establishment phase is phosphorous (McVey, 1968). It is immobile and in order for the small seedling with a limited root volume to be able to get in contact with it, there must be a fairly large amount of phosphorous in the soil. In contrast, the nitrogen level must not be too high because it can burn the small seedlings (Christians, 2007). When the seedlings are established the grass need N for further growth. In Denmark the average amount of fertilizer used on fairways is approx. 50 kg N /ha/ year. This might be too low in order to provide the desired grass species with a satisfactory growth and the ability to compete with broadleaved weeds. The dominant weed grass on fairways is annual bluegrass. Populations of this weed on golf courses are morphological and physiological different in relation to origin (green, fairway or rough). This includes differences in germination and flowering (Lush, 1989). Annual bluegrass populations from less managed areas (rough and fairway) have lower germination ability at high

6 temperature (25 C, Wu et al., 1987). Additionally an examination of annual bluegrass ecotypes demonstrated that all ecotypes germinated to some degree in complete darkness, which indicates that maintaining a dense turf canopy to eliminate annual bluegrass germination may not be completely effective. This justifies the focus on overseeding and timing in order to obtain detailed information about annual bluegrass occurrence and fairway density. Overseeding fairways might present a budget problem for many golf clubs, especially if the overseeding program includes topdressing and irrigation. It is a standard procedure on greens and tees, however, these areas usually cover less than 5% of the total course area. Overseeding fairways is a larger operation that includes around 20 % of the course area. Therefore, a complete overseeding program will have an impact on the golf clubs maintenance budget (man hours, equipment, seed, etc.). Evaluation of costs for the overseeding programs is therefore of large interest. Objectives To evaluate the impact of time, turfgrass mixture and fertilization practises when overseeding fairways on Nordic golf courses and in an experimental site To conduct a cost calculation of different overseeding programs Hypothesis There will be less dicot weeds and annual bluegrass by frequently overseeding fairways on Nordic golf courses Increasing the amount of fertilizer used on fairways will increase the covering degree of desirable grasses and reduce the amount of broadleaved weeds Using perennial ryegrass will result in the fastest improvement in fairway quality, however, after two years of repeated overseeding, the effects of red fescue and Kentucky bluegrass will also become significant

7 Materials and Methods This experiment included a field trial at NIBIO Landvik in Norway and field trials on three golf courses in Denmark (Korsør, Skovbo and Hornbæk). The experimental treatments were implemented in 2011, 2012, 2013, and the collection of data completed in Landvik trial In the small-scale trial at Landvik, conditions for establishment of the overseeded grasses was optimized in order to study the colonization potential of the various species and also the effect of additional fertilizer when water is not a limiting factor. The trial was conducted on an experimental fairway seeded in 2003 on a silt loam soil (64% sand, 29% silt, and 7% clay) with a seed mixture based on available seed of red fescue, Kentucky bluegrass and colonial bentgrass. Partly because varieties were not selected very carefully, and partly because the fairway has never been overseeded, the turfgrass cover is dominated by annual bluegrass in The following experimental treatments were compared: 1. No overseeding, no additional fertilizer 2. No overseeding, additional application of 50 kg N/ha when overseeding around 1 May 3. No overseeding, additional application of 50 kg N/ha when overseeding around 1 Sep 4. Seeding of red fescue around 1 May, no additional fertilizer 5. Seeding of red fescue around 1 May, additional application of 50 kg N/ha when overseeding 6. Seeding of Kentucky bluegrass around 1 May, no additional fertilizer 7. Seeding of Kentucky bluegrass around 1 May, additional application of 50 kg N/ha when overseeding 8. Seeding of perennial ryegrass around 1 May, no additional fertilizer 9. Seeding of perennial ryegrass around 1 May, additional application of 50 kg N/ha when overseeding 10. Seeding of red fescue around 1 Sep., no additional fertilizer. 11. Seeding of red fescue around 1 Sep., additional application of 50 kg N/ha when overseeding 12. Seeding of Kentucky bluegrass around 1 Sep., no additional fertilizer 13. Seeding of Kentucky bluegrass around 1 Sep., additional application of 50 kg N/ha when overseeding 14. Seeding of perennial ryegrass around 1 Sep., no additional fertilizer 15. Seeding of perennial ryegrass around 1 Sep., additional application of 50 kg N/ha when overseeding The standard fertilizer program on the experimental fairway was 50 kg N/ha, split between three applications of granular fertilizer (Fullgjødsel or OPTI-NK ) per year. The

8 additional fertilizer at seeding in half of the treatments was in the form of Sierraform Preseeder GT containing about 50% slow-release nitrogen and also ensuring P to the developing seedlings. To facilitate detailed studies on botanical composition, plot size in the trial at Landvik was limited to 1 m x 2 m. The experiment had four replicates (blocks). Overseeding was conducted according to the following procedure: Mowing height during the last week before overseeding was reduced from the usual 15 mm to 10 mm to reduce competition from annual bluegrass The entire plot was hollow-tine cored to 4 cm depth using a walk-behind JD Aerocore, 1 m wide and with 12 mm pipes. Cores was removed with an air-blower Exact seeding and application of additional fertilizers was conducted using a Scot s drop spreader, 1 m wide. The sowing rate of all species was 3 kg per 100 m2. This amount is chosen for all three species mixtures and the equal amount compensate for differences in field germination potential for the different species Seed and/or fertilizer was raked or blown into the holes. The entire main plot was topdressed with pure sand, 5 kg/m2 and the sand brushed into the holes Mowing was suspended for one week after seeding The trial was irrigated after seeding and the irrigation amount recorded Registrations Landvik trial Precipitation and temperature automatically on a daily basis throughout the year. Soil temperature was measured at sowing and once per week during the two following weeks. Visual recordings of turfgrass overall impression, tiller density, coverage, diseases every month throughout the project period. Fairway tiller density and botanical composition / the amount of weed prior to seeding in spring 2011, and then on 15 October in

9 Trial at 3 Danish golf courses In the Danish fairway trials the colonization potential of the various species and also the effect of additional fertilizer were investigated under real fairway overseeding conditions with no additional water and no topdressing. The trials were conducted on natural soil on established golf course fairways of different age, with different grass species composition and differences in weed covering degree. All 3 golf courses used approx. 50 kg N per ha per year Table 1. Soil characteristic nutrient content spring 2011 Rt Pt Kt Mg Cut Mnt Nit Korsør 6,4 5,3 29,2 11,8 22,3 42,1 5,3 Hornbæk 5,4 5,2 14,7 4,2 1,6 13,9 193,2 Skovbo 6,2 4,5 26,4 10,5 4,7 3,7 24 Table 2. Soil texture spring 2011 Humus Clay Under 0,002 mm Silt 0,002 0,02 mm Fine sand 0,02 0,2 mm Korsør 3,46 6,6 12,6 47,0 30,4 Hornbæk 2,01 4,7 7,8 35,5 40,0 Skovbo 5,23 4,5 9, ,6 Korsør soil type is a JB 4 fine loamy sand Hornbæk and Skovbo soil type is a JB 1 - coarse sandy soil Coarse sand 0,2 2,0 mm The following experimental treatments were compared: Plot size in the trials at three Danish golf courses was 2 m x 15 m. The experiment had four replicates (blocks). 1. No overseeding, no additional fertilizer 2. No overseeding, additional application of 50 kg N/ha when overseeding around 1 May 3. No overseeding, additional application of 50 kg N/ha when overseeding around 1 Sep 4. Seeding of red fescue around 1 May, no additional fertilizer 5. Seeding of red fescue around 1 May, additional application of 50 kg N/ha when overseeding 6. Seeding of Kentucky bluegrass around 1 May, no additional fertilizer 7. Seeding of Kentucky bluegrass around 1 May, additional application of 50 kg N/ha when overseeding 8. Seeding of perennial ryegrass around 1 May, no additional fertilizer

10 9. Seeding of perennial ryegrass around 1 May, additional application of 50 kg N/ha when overseeding 10. Seeding of red fescue around 1 Sep., no additional fertilizer 11. Seeding of red fescue around 1 Sep., additional application of 50 kg N/ha when overseeding 12. Seeding of Kentucky bluegrass around 1 Sep., no additional fertilizer 13. Seeding of Kentucky bluegrass around 1 Sep., additional application of 50 kg N/ha when overseeding 14. Seeding of perennial ryegrass around 1 Sep., no additional fertilizer 15. Seeding of perennial ryegrass around 1 Sep., additional application of 50 kg N/ha when overseeding Overseeding was conducted according to the following procedure: Seeding of subplots was conducted using a BLEC Disc Seeder 1800 (Working width 1,8m, 5 cm between each slit). The machine was provided by Prodana. Sowing rate of all species was 3 kg per 100 m2. The equal amount compensate for differences in field germination potential for the different species Fertilizer 16:11:8 HiGreen Prodana was spread on subplots using equipment available at the golf course. The fertiliser containing more P to ensure the developing seedlings There was no topdressing and no irrigation on fairways during the growing season and after the overseeding Fairways at the 3 golf courses was mowed according to normal practise on each course - normally the day before overseeding and then 3 days after the overseeding took place. Sowing depth perennial ryegrass 0,75 cm, red fescue 0,7, Kentucky bluegrass 0,5 cm Seed mixtures 2 varieties for each species o Red fescue; Nikky and Valdora o Perennial ryegrass; Dickins and Columbine o Kentucky bluegrass; Julius and Yvette Seed blends was selected in collaboration with Prodana, which also sponsored the seed Registrations Danish trial Soil temperature was measured when overseeding is performed in May and September (one measurement in each plot) to ensure that the temperature was above 7-8 degree Celsius Botanical grass composition was made prior to seeding in spring 2011 and then each year in May. Small cores 2 cm in diameter - from each plot were collected, placed in the

11 greenhouse for approx. 2-3 weeks and then number of plants (on species level) was counted Percent broadleaved weed was estimated prior to seeding in 2011 and again in spring 2012, 2013 and Registration was performed on the course using a counting frame (5 counts in each plot). For Taraxacum spp. and Plantago spp. the number of plants were additionally counted in the frame Picture 1. Sampling for counting of grass plants. Picture 2. Disc seeding. Picture 3. Disc seeding Picture 4. Disc seeding the seeds are placed in the slits. Statistical analyses Grass composition: For each replicate 5 counts were made. Statistical analysis was performed on the average of these 5 counts. For the Danish results data from the 3 locations were combined in order to see whether overseeding guidelines can be produced. Weed composition: For each replicate 5 counts were made. Statistical analysis was performed on the average of these 5 counts. Each location in Denmark were analysed separately due to the fact that the weed composition was very different between the 3 locations.

12 The effects of overseeding, time and fertilizer on grass composition and weed occurrence were analysed. The models were based on a mixed linear model approach (MIXED procedure in SAS 9.3). The response variables were analysed in response to the whole plot factor the grass introduced by overseeding, the subplot factor seeding time and fertilization and interaction effects. Block and locations (for the Danish data) were included as random effect. For all models, uniformity of variance was tested by visual inspection of residual plots produced with the SAS procedure GLM. Picture 5. The fairway at Skovbo golf course where the experiment took place.

13 Results and Discussion Introduction of new grass species by overseeding Overseeding in Denmark Overseeding was performed according to plan in 2011, 2012 and 2013 although the time for overseeding was adjusted in relation to the weather condition. Data were analyses as an average of the 3 locations. In order to be able to produce guidelines the effects must be overall. Perennial ryegrass is established in all four treatments with overseeding of perennial ryegrass (Table 3). There is no statistical effect of the different treatment although there is almost a significant effect of fertilization in fall (Table 4). More plants are established when fertilizer is applied in the fall overseeding with perennial ryegrass. Red fescue and Kentucky bluegrass is not successful established. For both species there was a reduction in number of plants per m 2 from 2011 to 2014 (Table 5 and table 6). Table 3. Number of plants of L. perenne per m 2 average of 3 locations and 4 repeats Time overseeding Fertilizer Diff L. perenne Spring No L. perenne Spring Yes L. perenne Autumn No L. perenne Autumn Yes Control No No Control No Spring Control No Autumn Table 4. Differences of Least Squares Means Fertiliser Time Fertiliser Time Estimate P vale No Yes -358,1 0,5225 e f 70,9549 0,8989 No e No f -1039,09 0,1979 No e Yes e -1468,14 0,0716 No f Yes f 751,95 0,3384

14 Table 5. Number of plants of F. rubra per m 2 average of 3 locations and 4 repeats Time overseeding Fertilizer Diff F, rubra Spring No F, rubra Spring Yes F. rubra Autumn No F. rubra Autumn Yes Control No No Control No Spring Control No Autumn Table 6. Number of plants of P. pratensis per m 2 average of 3 locations and 4 repeats Time overseeding Fertilizer Diff P.pratensis Spring No P. pratensis Spring Yes P. pratensis Autumn No P. pratensis Autumn Yes Control No No Control No Spring Control No Autumn Overseeding in Norway Overseeding with perennial ryegrass in 2011, 2012 and 2013 introduced this species into the experimental fairway (Table 7). There was no perennial ryegrass when the experiment was established in 2011 however overseeding demonstrated an increase in perennial ryegrass plants per m 2 from year to year. Overall perennial ryegrass is clearly established after 3 years in the plots where perennial ryegrass is seeded. It is almost significant (P 0,05 level) (Table 8 and table 9). But there is no effect of the seeding time and the application of fertilizer. The effect of the different treatments varies a lot over the 3 years. Red fescue was a part of the experimental fairway sward before the experiment started. Overseeding with red fescue in 2011, 2012 and 2013 introduced more plants into the turf. Overall red fescue is clearly established after 3 years in the plots where red fescue is seeded (Table 10). It is significant at the 0,05 level (Table 11 and table 12). The effect of the different treatments varies a lot over the 3 years and no clear conclusion in relation to sowing time or maintenance (+/- fertiliser) can be given. Overseeding with Kentucky bluegrass in 2011, 2012 and 2013 did not result in more plants in the turf. In contrary fever plants were seen in all 4 treatments after 3 years of overseeding (Table 13). However the same was seen in the controls where the number of Kentucky bluegrass plants decreased.

15 Table 7. Number of plants of L. perenne per m 2 average of 4 repeats Time overseeding Fertilizer Diff L. perenne Spring No L. perenne Spring Yes L. perenne Autumn No L. perenne Autumn Yes Control No No Control No Spring Control No Autumn Table 8. Least squares means Treatment Estimate diff P value L. perenne ,0069 Control 0 1 Control spring 120,34 0,9839 Control autumn 3008,62 0,6214 Table 9. Difference of Least Squares means Treatment Treatment Estimate P value L. perenne Control ,0589 L. perenne Control spring ,0598 L. perenne Control autumn ,089 Table 10. Number of plants of F. rubra per m 2 average of 4 repeats Time overseeding Fertilizer Diff F, rubra Spring No F, rubra Spring Yes F. rubra Autumn No F. rubra Autumn Yes Control No No Control No Spring Control No Autumn Table 11. Least squares means Treatment Estimate diff P value F. rubra ,0013 Control 4212,06 0,1942 Control spring 722,06 0,7937 Control autumn 4031,55 0,209

16 Table 12. Difference of Least Squares means Treatment Treatment Estimate P value F. rubra Control ,0315 F. rubra Control spring ,0149 F. rubra Control autumn ,0302 Table 13. Number of plants of P. perenne per m 2 average of 4 repeats Time overseeding Fertilizer Diff P.pratensis Spring No P. pratensis Spring Yes P. pratensis Autumn No P. pratensis Autumn Yes Control No No Control No Spring Control No Autumn Factors affecting grass introduction The Danish trials on 3 golf courses and the Norwegian trial at the experimental fairway at Landvik demonstrated that the success of overseeding varied. Perennial ryegrass was introduced on all locations but with different success whereas red fescue was only introduced with success in the Norwegian trial. These differences among countries might be ascribed to different factors. A successful introduction and establishment of new grass plants after an overseeding depends on several factors. The first key to success is to create a good seed-soil contact. In Norway core aeration created holes within the soil. Seed applied afterwards was blown into the holes and topdress was applied. That procedure produced at good seed to soil contact and might explain the better success with the overseeding. In Denmark disc seeding was used but the only preparation before seeding was that the fairways were mowed the day before overseeding. A disc seeding deposits the seed within the slit but in Denmark there was difficulties with the red fescue seeding. It was difficult to place all the seeds in the slits. Many seeds were deposited on top of the turf and not in the slit. In retrospect methods that helps work the seed down through the grass blades could have been beneficial. But the experiment on the golf course was to try a method that was reasonable in cost and labour work without too much preparation. Producing slits on a fairway is also difficult when it is uneven. Sometimes no slits were made and the seed was place on the turf and sometimes the slits were too deep and the seed was placed at a depth from where it cannot germinate. Especially for Kentucky bluegrass it is critical if

17 the seed is placed below 0,5 cm and red fescue and perennial ryegrass should not be placed below 1 cm. Picture 6. Seeding at Korsør May Red fescue seed lays on top of the turf. Picture 7. Ryegrass seed is deposited in the slits. Another key to success is to have a soil with good conditions for seed germination and seedling growth. The ph must be appropriate and no fertilizer deficit for the nutrients necessary to support seedling establishment and growth. Soil analysis of the Danish golf courses did not give any alarming results. P value was within the normal range (3,5 5 on clay soil and 2,5-3,5 on sandy soil). In literature and web advice from consultants regarding overseeding of athletic field s state that overseeding generally fail due to a lack of moisture (Gaussoin and Branham, 1989 and Eggers, 1979). If the seed is not exposed to moisture it will not germinate. Moisture is needed from the point the seed is applied, through germination and seedling establishment. Frequent, light irrigation cycles are therefore a key to success. Therefore the success of overseeding on a fairway with no irrigation can be low. The Norwegian way of seeding might have been a better solution in order to ensure moisture for germination. The seed was placed in a hole, top-dressed and irrigated. Irrigation after overseeding was not a possibility on the Danish golf courses and the success was dependent on the moisture in the soil and precipitation. The aim of the Danish experiment was to test the effect of an overseeding with no irrigation because this is the reality that most golf courses face. The Danish experiment tried to compensate by creating a good soil/seed contact by disc seeding. Climate data from Denmark demonstrates that the conditions for obtaining success with the overseeding might have been challenged in some years was a warm year especially in the spring it was record warm in April and a record max temperature was reached in October. There

18 was a rain deficit in spring and late fall whereas the summer was very wet. This was not optimal for grass establishment after overseeding was a bit colder than The late spring/early summer and the beginning of fall were very wet. Maybe a bit too wet was characterised with a rain deficit in the early spring and in the summer (Figure 1 and figure 2). This might have affected the spring overseeding. But these are the conditions that a golf course face if watering after overseeding is not a possibility. Therefore it is important to overseed more than once in order to increase the chance of success. Figure 1. Temperature average in each month over the experimental period Temp - gennemsnit Temp - normal ( ) Temp - normal ( ) Figure 2. Precipitation in each month over the experimental period Nedbør Nedbør - normal ( ) nedbør - normal ( )

19 Another important factor is the soil temperature. It needs to be in the range for germination of the three grass species chosen. Normally a fall overseeding is not a problem for cool-season grasses. The soil is warm enough and in fall a lot of dew is produced which can help in the germination when there is no watering. In the spring it is different. A soil temperature above 7-8 degree Celsius is needed. In this experiment the soil temperature was checked before each overseeding in order to assure that the temperature was above this limit. Grass occurrence Introduction of perennial ryegrass and red fescue might give a denser turf with more plants per m 2 and additionally the assumption is that it will decrease the occurrence of annual bluegrass, a weed grass. The results from Norway demonstrate that the amount of annual bluegrass is reduced in most treatment but unfortunately also in the control and also where Kentucky bluegrass is overseeded (Table 14). So the reduction in annual bluegrass is not a result of new plants introduced in an overseeding. Other factors might determine the number of annual bluegrass plants such as climate and water availability. Table 14. Difference in number of P. annua plants and total plant per m 2 between 2011 and average of 4 repeats Time overseeding Fertilizer Diff P. annua Diff Total plants P. pratensis Spring No P. pratensis Spring Yes P. pratensis Autumn No P. pratensis Autumn Yes Control No No Control No Spring Control No Autumn L. perenne Spring No L. perenne Spring Yes L. perenne Autumn No L. perenne Autumn Yes F, rubra Spring No F, rubra Spring Yes F. rubra Autumn No F. rubra Autumn Yes Overseeding should ideally give a denser turf due to the introduction of new plants. In the Norwegian experiment more plants per m2 was observed in all treatments in 2014 compared to 2011 (Table 14). Unfortunately the number was also increased in the control where no

20 overseeding was performed so the increase in plants per m 2 cannot be ascribed to the overseeding with new species or the aeration performed in relation with the seeding procedure. Examination of the grass data reveal that the increase in number of plants is caused by colonial bentgrass. There is much more bentgrass in 2014 than in The data might reflect the dynamic that occur in the grass population in relation to different climate and soil conditions. A difference between the Norwegian experiment and the one on the golf course in Denmark is the amount of wear. And it might have an influence on the success of the overseeding. Traffic might have a negative influence on the young seedlings. Additionally the traffic from players on the golf course and the maintenance machines, combined with the fact that some of the seeds were place on top of the turf, might explain why an increase in perennial ryegrass was seen in the control plots in Denmark. Seeds might have been dragged from one plot to another. When to overseed The most obvious approach is to overseed when bare soil or thin turf appears. In the Nordic countries, especially Denmark, most golf courses do not have the possibility to irrigate fairways, there are restriction on pesticide use (amount and type of products), growth regulators cannot be used, the cost of topdressing is too high and a minimum of fertilizer is used, restrictions that do not apply to other part of the world. In the US different kind of herbicide treatments are often a part of the overseeding procedure and necessary to obtain success with the overseeding (Patton et al. 2004). Under our conditions an overseeding procedure must be carried out when the conditions for seed germination is at its highest. The question is therefore when to overseed? The data from this experiment cannot be used to give any recommendations on when to overseed. The success with introducing perennial ryegrass does not depend on overseeding time. Only on the Norwegian experimental fairway there is a small indication of the fall overseeding with fertilizer being more successful. One of the purposes of overseeding is also to build a seed bank where the seed will germinate when conditions are right. For a viable seed bank to build up, the seed must be in the soil and not on the surface where it is subjected to water and desiccation. In order to ensure a constant grass cover on a fairway the introduction of different grass species are important. The different species have different dormancy periods and the germination takes place with or without a cold period. Whether we have succeeded in creating a seed bank will be visible in 5-7 years.

21 Appearance data Introducing new grass species or more grass plants to a turf sward might change the appearance of the turf. In the Norwegian trial appearance (overall performance, colour, density and disease occurrence) were scored frequently during the 3 year period. Figure 3. Overall appearance ,5 6 5,5 5 4,5 4 3, P.p fall P.p fall fert P.p spring P.p spring fert Control Control f-fert Control s-fert L.p fall L.p fall fert L.p spring L.p spring fert F.r falll Figure 4. Overall appearance ,5 6 5,5 5 4,5 4 3,5 3 2, P.p fall P.p fall fert P.p spring P.p spring fert Control Control f-fert Control s-fert L.p fall L.p fall fert L.p spring L.p spring fert F.r falll F.r fall fert F.r spring

22 Figure 5. Overall appearance P.p fall P.p fall fert P.p spring P.p spring fert Control Control f-fert Control s-fert L.p fall L.p fall fert L.p spring L.p spring fert F.r falll F.r fall fert F.r spring F.r spring fert Figure 6. Overall appearance ,5 6 5,5 5 4,5 4 3, P.p fall P.p fall fert P.p spring P.p spring fert Control Control f-fert Control s-fert L.p fall L.p fall fert L.p spring L.p spring fert F.r falll F.r fall fert F.r spring

23 Examination of overall appearance data reveals no clear distinction of a specific overseeding/timing/fertilizer combination. A closer look at figure 3 to 6 illustrates that the spring overseeding of perennial ryegrass especially with fertilizer seems to have the best overall appearance every year in spring and summer in most years. It might be that the new seeds that germinate quickly after seeding in spring has a positive influence on the overall appearance creating a denser and greener turf. Figure 7. Shoot density as an average over the 4 treatments for each grass species ,00 6,00 5,00 4,00 3,00 2,00 P. pratensis Control L. perenne F. rubra 1,00 0, Figure 8. Shoot density as an average over the 4 treatments for each grass species ,00 6,00 5,00 4,00 3,00 2,00 P. pratensis Control L. perenne F. rubra 1,00 0,

24 Figure 9. Shoot density as an average over the 4 treatments for each grass species ,00 6,00 5,00 4,00 3,00 2,00 P. pratensis Control L. perenne F. rubra 1,00 0, Figure 10. Shoot density as an average over the 4 treatments for each grass species ,00 6,00 5,00 4,00 3,00 2,00 P. pratensis Control L. perenne F. rubra 1,00 0, A visual examination of shoot density from demonstrates that plots overseeded with red fescue and perennial ryegrass have a higher density during most of the year (Figure 7 to 10). The magnitude however varies a lot and statistically no overall significant difference could be demonstrated. Data from the overseeding demonstrated that perennial ryegrass and red fescue was successful introduced in the turf in the Norwegian trial and this might explain why these plots

25 have a higher shoot density. Number of plants calculated from the grass species data also supports this (Table 15) Table 15. Number of plants as an average over the 4 treatments for each species Overseeding species F. rubra L. perenne P. pratensis Control Picture 8. The Norwegian trail in November The dark green plots are overseeded with perennial ryegrass A visual scoring of colour was additionally performed in 2013 and Again in late spring early summer the best colour was registered in plots overseeded with perennial ryegrass. The visual appearance of the fairway can shortly be improved by introducing perennial ryegrass as

26 demonstrated in the Norwegian trial. However this effect is not clear throughout the growing season Figure 11. Colour scoring in 2013 and P.p fall 6,5 P.p fall fert P.p spring 6 P.p spring fert Control 5,5 Control f-fert Control s-fert 5 L.p fall 4,5 L.p fall fert L.p spring 4 L.p spring fert F.r falll 3, F.r fall fert F.r spring F.r spring fert

27 Broadleaved weed occurrence Denmark On the 3 Danish golf courses the broadleaved weed occurrence on species level was registered each year from 2011 to 2014 in spring before overseeding. The aim was to test whether 3 year of overseeding had an effect on weed occurrence. The three golf courses were different in relation to weed occurrence. Korsør golf course has for many years, not used pesticides to remove weed on fairways because the course is placed on a conservation area. At Skovbo and Hornbæk golf course herbicides have prior to the experiment been used to remove some weed species on fairways. Therefore the number of weed species and the type of species varied from course to course. Some species occur on all 3 courses and a few species are specific for a single course (Table 16). Table 16. Weed species found on the different locations. Species Hornbæk Korsør Skovbo Taraxacum spp. X X x Trifolium repens X X X Bellis perennis X X Achillea millefolium X X X Plantago major X X Veronica chamaedrys x X X Capsella bursa-pastoris X X Senecio spp. X Geranium spp. X Ranunculus sp. x Cirsium arvense X Cerastium sp x On the three courses all weed species were registered as occurrence frequency using a 100 field frame. Only Taraxacum spp. and Plantago major were also registered with number of plants. On all 3 courses a number of species were only found in a few blocks and the occurrence in relation to overseeding could not be analysed on the species level. These species include Plantago major, Veronica chamaedrys, Capsella bursa-pastoris, Senecio spp. Geranium spp. Ranunculus spp. and Cirsium arvense. A visual examination of all the data did not indicate that some of these weed species was removed after overseeding. Additionally there was no indication of overseeding causing the invasion of specific weed species. Taraxacum spp. Trifolium repens, Bellis perennis and Achillea millefolium were more even distributed and occurred on a larger scale on the experimental sites. All 4 species occurred in most

28 plots on Korsør golf course. At Hornbæk and Skovbo golf course Taraxacum spp. was in all plots whereas T. repens and A. millefolium occurred in some plots and B. perennis in only a few plots. Table 17. Weed occurrence, Hornbæk golf course. Change from the beginning of the project to the end of the project. Treatment Taraxacum spp % Taraxacum spp no Trifolium repens % Achillea millefolium % P.pratensis fall 0,20 0,55-0,00 2,20 P.pratensis fall fert 0,38 0,10 0,00 0,00 P.pratensis spring -0,08 0,25 3,00 1,00 P.pratensis spring fert -0,24-0,15 0,00-4,30 Control 0,14 0,06 0,30 0,90 Control f-fert 0,01 0,03 00,00 1,30 Control s-fert -0,27-0,30 0,10 3,00 L.perenne fall 0,11 0,20 1,58 0,25 L.perenne fall fert -0,05 0,00 0,75 0,60 L.perenne spring -0,11 0,15-0,18 0,50 L.perenne spring fert -0,36-0,35 0,00-0,35 F.rubra fall 0,28 0,40 0,18 1,80 F.rubra fall fert -0,06-0,10 0,00-0,60 F.rubra spring 0,19 0,25 0,00-3,50 F.rubra spring fert -0,09-0,15 0,00 0,00 Table 18. Weed occurrence, Skovbo golf course. Change from the beginning of the project to the end of the project. Treatment Taraxacum spp % Taraxacum spp no Trifolium repens % Achillea millefolium % P.p fall 0,04-0,25 3,41 0,00 P.p fall fert 0,20 0,20 4,20 0,00 P.p spring 0,30-0,15 1,15 6,55 P.p spring fert 0,40 0,35 1,45 0,00 Control 0,07-0,20 0,33 5,10 Control f-fert 0,45 0,55 1,18 0,00 Control s-fert -0,01-0,15 0,10 0,00 L.p fall 0,13-0,05 3,83 10,05 L.p fall fert -0,02 0,00-1,75 2,60 L.p spring 0,06-0,02 2,40 5,65 L.p spring fert 0,07 0,15 0,30 0,00 F.r fall 0,29 0,50 3,98 3,75 F.r fall fert 0,19 0,20-1,13 0,00 F.r spring 0,34 0,40 0,95 3,10 F.r spring fert 0,003-0,10 2,30 1,55

29 Table 19. Weed occurrence, Korsør golf course. Change from the beginning of the project to the end of the project. Treatment Taraxacum spp % Taraxacum spp no Trifolium repens % Bellis spp % Achillea millefolium % P.p fall 1,68 1,30-10,17-1,68-0,00 P.p fall fert 1,83 1,10-5,35-5,11 2,60 P.p spring 0,91 1,40 0,15-1,55-2,45 P.p spring fert 0,89 1,25 1,64-0,90 0,60 Control 1,00 1,50 1,16-0,14-0,00 Control f-fert 0,98 1,05-4,93-5,22 1,30 Control s-fert -0,72-0,75-0,94-2,00 1,00 L.p fall 0,72 1,25 3,96-0,38-2,25 L.p fall fert -0,75-0,40-1,57-4,81-5,00 L.p spring 0,74 1,25 1,26-2,31 3,15 L.p spring fert -0,26 0,45-5,70-5,60 0,00 F.r fall 1,67 2,05 2,55-6,84 0,00 F.r fall fert 1,36 1,35-1,74-3,19 4,00 F.r spring -0,31 2,00 4,18 1,40 13,50 F.r spring fert 1,24 1,60-10,83-4,32 0,00 The change in occurrence frequency from the beginning of the experiment to the end in 2014 demonstrates a huge variation in weed occurrence in the different plots and the different treatments (Table 17, 18 and 19). Analysing on the difference between 2011 and 2014, no statistic difference could be observed (P 0,05 level). There was no effect of the different treatments on the broadleaved weed occurrence not overall or on location level. There is no indication of overseeding being a solution in order to reduce the weed occurrence 3 years after the first overseeding. Norway In the Norwegian trial weed occurrence was estimated frequently over the different growing seasons from 2011 to Overall the amount of weed increase from 2011 to Analysis of data does not demonstrate any significant effect of treatment on the weed occurrence (P 0,05 level). Not even the introduction of perennial ryegrass and red fescue did affect the weed occurrence in a negative way.

30 Figure 12. Weed occurrence in perennial ryegrass overseeded plots L.p fall L.p fall fert L.p spring L.p spring fert Control Control f-fert Control s-fert Figure 13. Weed occurrence in red fescue overseeded plots F.r falll F.r fall fert F.r spring F.r spring fert Control Control f-fert Control s-fert

31 Examination of the plots seeded with perennial ryegrass indicate that a spring overseeding with no fertilizer gives the lowest occurrence of weed and it became more and more pronounced over the 3 years (Figure 12). The weed occurrence is even lower than in all 3 controls. The effect is not statistically significant (P 0,05 level) and an explanation is not clear. It might be a coincidence. This effect is not seen in plots where red fescue is introduced (Figure 13). A closer examination of the data visualise that in 2013 and 2014 clover occurred in a few plots with a high covering frequency. This might explain the high variation among the four perennial ryegrass overseeding procedures. Weed and overseeding The effect of overseeding on weed species has not been examined before in detail. In theory introduction of new improved cultural turf grasses with a higher shoot density should compete better against the weeds. The assumption was that an introduction of new grass species into a fairway would have a negative effect on the weed occurrence. As demonstrated in the Danish experiment only perennial ryegrass was successfully introduced into the fairway turf. But there is no effect of perennial ryegrass overseeding, timing and extra fertiliser on weed occurrence. Overall the amount of weed increase from 2011 to For Taraxacum there was an increase in covering degree and number of plants in most treatments on all three locations also in the control. This increase in many of the plots may therefore be ascribed to other factors and the fact that weed populations are dynamic and in some years more weed occurred than in others. It would be interesting to examine the weed occurrence in 2017 or Experience from a Danish greenkeeper (Viborg golf course) is that an area that was overseeded in in the project period did not show any effect on weed occurrence (overall weed occurrence) but after 7-10 years he could visually see that the area that was overseeded years ago had less weed than the rest of the fairway. This indicates the dilemma of 3-4 years research projects. The question is how fast do the grass plants produce shoots and create a more dense grass cover that is able to compete with the weeds. One general problem especially in the Norwegian trial was that the initial amount of weed was very low. In many plots the weed occurrence was below 1 %. General effects on the weed occurrence are difficult to register and the data analysis is more sensible to plots harbouring an aggressive weed such as clover. One of the general problems on fairways in Scandinavia is the change that has taken place over the last years, in the amount of fertiliser used. On Danish golf courses approx. 50 kg N per ha per year is used on fairways (DGU 2011). This might be insufficient to give a dense turf that can compete with the weed. In some cases it might even favour some weed species such as white clover.