29 1. Introduction Bermudagrass is still the dominant warm-season turf- grass in warm to temperate climatic regions of the world. It is well adapted to a wide range of soil types, and its drought tolerance, recuperative ability, salt tolerance, wear tolerance, aggressive stoloniferous and rhizomatous growth habit, and overall appearance make bermudagrass an ideal turfgrass in many environments (Taliaferro, 2003; Shearman, 2006). Bermudagrass includes several taxa of the genus Cyno- don (L.) Rich. but the two species that represent the genetic pool from which the present cultivars descend are Cynodon dactylon (L.) Pers. Var. dactylon and Cynodon transvaal- ensis (Burtt-Davy) also known as African bermudagrass (Taliaferro, 2003). C. transvaalensis is morphologically distinct from C. dactylon due to narrow erect pale leaves producing a fine textured turf with a yellowish-green colour (De Wet and Harlan, 1970; Taliaferro, 1992). C. dactylon can be found as far north as 53° N latitude and from sea level to 3000 m altitude (Taliaferro, 2003). Asexual repro- duction has played a role in bermudagrass enhancement as well. Remarkable breeding progress has been obtained from inter-specific hybridization and mutation breeding. The inter-specific hybridization of C. dactylon and C. trans- vaalensis has been extensively used to obtain sterile culti- vars for which clonal propagation is necessary due to a lack of viable seeds. Among hybrid genotypes, a number of cul- tivars have been selected for plant size and morphology in response to the lower cutting height adopted over the years on golf greens. The more recently released “ultradwarf” cul- tivars have become routinely adopted thanks to an improved density, a slower vertical leaf extension and an increased dominance of stoloniferous growth relative to rhizomes at low mowing heights (Beard and Sifers, 1996). Parameters used to evaluate Cynodon turf typically in- clude turfgrass quality, colour, percent spring green-up, establishment rates, leaf texture, and density. As the turf- grass industry moves towards more sustainable manage- ment practices, the types of parameters potentially related to better stress tolerance are increasingly important (Bald- win and Liu, 2013). One of the most important parameters is cold toler- ance and rapidity of recovery from winter dormancy in the spring (Anderson et al., 2007; Patton et al., 2008). Low- temperature tolerance depends on a combination of sev- Bermudagrass adaptation in the Mediterranean climate: phenotypic traits of 44 accessions S. Magni*, M. Gaetani*(1), N. Grossi*, L. Caturegli*, S. La Bella**, C. Leto**, G. Virga**, T. Tut- tolomondo**, F. Lulli***, M. Volterrani* * Dipartimento di Scienze, Agrarie, Alimentari e Agro-ambientali, Università di Pisa, Via del Bor- ghetto, 80, 56127 Pisa, Italy. ** Dipartimento di Scienze Agrarie e Forestali, Università di Palermo, Viale delle Scienze, 208, 90128 Palermo, Italy. *** Turf Europe R&D, Pisa, Italy. Key words: colour, Cynodon dactylon, green-up, node density, quality, shoot density. Abstract: The use of bermudagrass in the Mediterranean area is increasing for its outstanding tolerance to heat and drought, and its aggressive growth and high recuperative potential make it particularly suited to heavily worn areas and appreciated for sports turfs. However, the overall performance of a given genetic type can be affected by the adaptation to a specific environment. The objective of this research was to determine the variability of a number of phenotypic traits that can affect bermudagrass turf performance on a wide range of bermudagrass accessions grown in two locations in Italy. In May 2010, 44 accessions of bermudagrasses, grouped in “wild”, “improved” “hybrid” and “dwarf” types were transplanted in the center of field plots in Pisa and Palermo. In 2011, when the turf was completely established, the following traits were determined: shoot density, horizontal stem density, node density, leaf width, colour, quality, spring green-up, and fall colour retention. Dwarf and hybrid types yielded the best aesthetic characteristics. With respect to colour retention and spring green-up, great variability was recorded within the groups. Dwarf types pre- sented the earliest dormancy, while the hybrid types were in general the ones to green-up first in spring. Adv. Hort. Sci., 2014 28(1): 29-34 1 Corresponding author: monica.gaetani@unipi.it Received for publication 14 April 2014 Accepted for publication 19 May 2014 30 eral factors, including environmental conditions, cultural practices, and especially genetic factors (Blum, 1988; An- derson and Taliaferro, 2002). Bermudagrass survives the dormancy period using its reserves of nonstructural car- bohydrates and nitrogen compounds accumulated during the previous growing season in storage organs such as sto- lons and rhizomes (Macolino et al., 2010; Volterrani et al., 2012; Giolo et al., 2013; Pompeiano et al., 2013). In the last two decades several southern European univer- sities have developed research programs to study warm-sea- son turfgrass species, including bermudagrass (Volterrani et al., 2008; Lulli et al., 2011; Lulli et al., 2012; Nikolopoulou et al., 2012; Agati et al., 2013; Gómez de Barreda et al., 2013), and in particular their adaptability to the Mediterra- nean environment (Volterrani and Magni, 2004). The aim of our research was to determine the variability of a number of phenotypic traits and aesthetical characteris- tics that can affect bermudagrass turf performance in a wide range of bermudagrass accessions grown in two locations in Italy. This information can provide further insight into ber- mudagrass adaptability in the Mediterranean climate. 2. Materials and Methods Plant material With the objective of expanding morphological diver- sity of the plant material, 44 accessions of bermudagrass [Cynodon (L.) Rich.], representative of both wild popula- tions and cultivars, were included in the present study. Group one included 13 entries that were called “wild types”, naturally occurring populations of C. dactylon (L.) Pers. collected from contrasting environments supposed to generate a selective pressure. Collection sites were located in Italy (CeRTES-1= warm temperate, salt affected soil; CeRTES-2= warm temperate, fertile soil; CeRTES-3= warm temperate, polluted soil; CeRTES-13= warm tem- perate, fertile soil), France (CeRTES-4= cool humid, fer- tile soil), Greece (CeRTES-5, -6 and -7= warm temperate, salt affected soils), Croatia (CeRTES-8= warm temper- ate, salt affected soil), Argentina (CeRTES-9= warm tem- perate, salt affected pastureland), United Arab Emirates (CeRTES-10 and -11= warm arid, desert sand), and Mal- dives (CeRTES-12= tropical humid, salt affected soil). Group two included 13 entries that were called “im- proved types”. These were experimental or commercial vegetative and seeded improved C. dactylon cultivars. Group 3 included 11 entries called “hybrid types”, commercial or experimental inter-specific hybrids (Cyn- odon dactylon x transvaalensis Burtt.-Davy) of which those labelled Tif- were kindly provided by Dr. W. Hanna (University of Georgia, USA). Group four included seven entries called “dwarf types”, commercial interspecific dwarf and ultradwarf hybrid cul- tivars and two Cynodon transvaalensis Burtt.-Davy acces- sions, one a commercial cultivar (Uganda) and the other (Roma) a line of African bermudagrass that was collected in a turf nursery in Rome (Italy) where the species was first introduced presumably as a weed. African bermu- dagrasses were included in the “dwarf types” due to their similarity in leaf texture, density and growth habit with the well-known hybrid dwarf bermudagrasses. 2010 Turf establishment On 15 April 2010 at the University of Pisa, Italy, all the accessions were propagated in the greenhouse (24±5°C) in peat-filled honeycomb seed trays (7 cm2 area and 25 cm3 volume each cell). Vegetatively propagated genotypes were planted as single stolon and seeded cultivars were seeded as single seed. On 13 May 2010 plants in the greenhouse were fertil- ized (30 kg ha-1 N, 10 kg ha-1 P, and 10 kg ha-1 K) using a soluble fertilizer (Grow More Inc., Gardena, CA, USA). On 31 May 2010, plants were mown to 5 cm and trans- planted into field plots in two locations in Italy: the re- search station of the University of Pisa (43°40’N, 10°19’E, 6 m a.s.l.) and the research station of the University of Palermo (38°06’N, 13°20’E, 50 m a.s.l.). Experimental plots were 1.5 by 1.5 m with 0.5 m bare soil pathways arranged in a randomized complete-block design with four replications. One plant of bermudagrass was transplanted in the centre of each plot. Soil type at Pisa was silt-loam (28% sand, 55% silt and 17% clay) with a pH of 7.8 and 18 g kg-1 organic matter while at Palermo soil type was sandy clay loam (54% sand, 23% silt and 23% clay) with a pH of 7.6 and 14 g kg-1 organic matter. Irrigation was applied as needed to encourage establish- ment. Plots received 50 kg ha-1 N, 10 kg ha-1 P, and 40 kg ha-1 K per month from June to September 2010. In or- der to minimize weed competition, from two years before establishment, the experimental areas were treated twice a year with glyphosate [N-(phosphonomethyl) glycine] at 2.88 kg ha-1 a.i. The day before planting, oxadiazon [5-tert-butyl-3-(2,4-dichloro-5-isopropoxyphenyl)-1,3,4- oxadiazol-2(3H)-one] was applied at 3.36 kg ha-1 a.i. Plots were not mowed during the year of establishment to avoid genotype x mowing interaction. Weeds occurring during the trial period were manually removed, even if it is not a standard cultivation technique, as the accessions could be differently injured by chemical removal. In the trial, no pesticides were applied. Encroachment of stolons into adjacent plots was avoided by using toothpicks that redi- rected the growing tip back toward the plot centre. 2011 In 2011, in the first and second weeks of March at Palermo and Pisa, respectively, scalping was carried out. From the end of April (the end of green-up) to October 2011, the turf was mowed weekly with a reel mower (John Deere 20SR7) at a mowing height of 2.5 cm. The irriga- tion program was adjusted according to soil temperature and evapotranspiration rate, with supplement irrigations applied as needed to prevent visual wilt of the turf (Croce et al., 2004). Plots received 50 kg ha-1 N, 20 kg ha-1 P, and 40 kg ha-1 K per month from May to August 2011. 31 Weeds were manually removed inside the plots dur- ing the trial period as the accessions could be differently injured by chemical treatments. In the trial, no pesticides were applied. To avoid the encroachment of stolons into adjacent plots, the corridors were treated with glyphosate at 2.88 kg ha-1 a.i. every other week. No turf cultivation, or verticutting or phytosanitary treatment was practiced on the plots. Monthly mean maximum and minimum temperatures recorded at the two trial sites are reported in Table 1. There were 13 days in Pisa with air temperatures below 0°C from November 2010 to March 2011; zero days were recorded in Palermo for the same period. Assessments Spring green-up (15 March-15 May 2011) and fall co- lour retention (15 November 2011-15 January 2012) were estimated and expressed as percentage of green ground cover. In the second week of October at both testing sites a 50 cm2 core sample per plot was collected and the follow- ing parameters determined: leaf width (20 fully expanded leaves per plot measured with a precision Vernier caliper and data reported in millimeters), shoot density (direct counting with data reported as shoot no cm-2), horizontal stem density (stolons and rhizomes collected after soil washing measured with a ruler and data reported as cm cm-2) and node density (nodes of stolons and rhizomes col- lected from the core samples counted and reported as no cm-2) (Roche and Loch, 2005; Volterrani et al., 2008; Volt- errani et al., 2010; Pompeiano et al., 2012). At the Pisa location two additional parameters were de- termined: at 30-day intervals throughout the growing sea- son (May-October) colour, with a rating scale of 1=light green and 9=dark green, and quality with a rating scale of 9 = best and 1 = poorest (Morris and Shearman, 2007; Pat- ton et al., 2009 ) were estimated. Data were subjected to analysis of variance using Co- Stat software (Monterey, CA, USA). To test the effects of location, accession and their interaction, a factorial combi- nation was used. Significantly different means were sepa- rated using Fisher’s Least Significant Difference (LSD) at the t-probability level of 0.05. 3. Results The interaction of treatments was not statistically sig- nificant for any of the parameters recorded in Pisa or Pa- lermo. For all the parameters, the location and accession mean effects were statistically significant. Location effect is reported as average across accessions and accession ef- fect is reported as average across locations. Location mean effect For the location effect, shoot density and node den- sity were on average higher in Pisa (3.1 shoots cm-2 and 2.4 nodes cm-2 respectively) compared to Palermo (0.8 and 1.0) (Table 2). Also horizontal stem density was higher in Pisa (4.2 cm cm-2) while leaf width recorded in Pisa was on average less (1.4 mm) compared to Pal- ermo (2.4 mm). Spring green-up evaluations as average across loca- tions showed green cover percentages (April 14) higher in Palermo (81%) with respect to Pisa (60%), while fall co- lour retention (December 17) showed higher green cover percentages in Palermo (77%). Accession mean effect Shoot density All wild type entries, with the exception of CeRTES 12 (2.4 shoot cm-2) which produced a density comparable to hybrid and improved types, had a similar shoot den- Table 1 - Monthly mean air temperatures (°C) during the trial period (2011) at Pisa and Palermo Month Air temperature (°C) Pisa Palermo Mean maximum Mean minimum Mean maximum Mean minimum January 10.6 3.5 14.7 9.3 February 11.8 3.3 13.9 9.3 March 13.9 5.2 16.2 10.4 April 19.2 9.0 19.6 13.1 May 23.8 11.8 21.4 15.9 June 26.2 17.0 25.2 20.2 July 27.6 18.3 28.6 22.9 August 30.2 18.5 29.0 23.0 September 26.7 16.0 26.9 21.0 October 21.8 10.5 22.6 16.5 November 17.5 6.6 20.2 14.5 December 13.3 4.7 16.9 12.6 Table 2 - Bermudagrass [Cynodon (L.) Rich.] shoot density, horizontal stem density, node density, leaf width. Location effect averaged across ac- cessions Shoot density (n° cm-2) Horizontal stem density (cm cm-2) Node density (n° cm-2) Leaf width (mm) Spring green-up (%) Fall colour retention (%) Pisa 3.1 4.2 2.4 1.4 60 33 Palermo 0.8 1.9 1.0 2.4 81 77 Means are significantly different at the 0.05 level of probability as determined by Fisher’s protected LSD. 32 sity with values ranging from 0.5 to 1.0 shoots cm-2 (Table 3). The most dense improved type was Wintergreen (2.3 shoots cm-2), hybrid type values ranged from 1.7 shoots cm-2 (Patriot) to 3.8 shoots cm-2 (Tif 00-1), while the most dense dwarf type was Miniverde with 5.1 shoots cm-2. Horizontal stem density The highest value was recorded for Miniverde with 5.5 cm cm-2 while CeRTES 12 had a slightly lower value (5.1 cm cm-2) (Table 3). The variability within the differ- ent groups was high for this parameter with values rang- ing from 1.5 to 5.1 cm cm-2 (respectively for Certes 1 and Certes 3 versus CeRTES 12) for the wild types, from 2.1 to 4.9 cm cm-2 (respectively for Scotts R6LA and Yukon ver- sus Bull’s Eye) for the improved types, from 2.7 to 4.3 cm cm-2 (for Tif 00-18 compared to Santa Ana and Tif 00-1) for the hybrid types, and from 1.6 to 5.1 cm cm-2 (respec- tively for Tifdwarf and Miniverde) for the dwarf types. Node density CeRTES 3 and Miniverde were the entries with the low- est and the highest node density with 0.6 and 5.5 nodes cm-2, respectively (Table 3). The variability within the groups was high for this parameter, with the exception of the hybrid types that had values ranging from 1.6 to 2.5 nodes cm-2 (respectively for Tif 00-18 and Santa Ana). Leaf width Coarser leaves were found in the wild types with values ranging from 2.1 mm (CeRTES 10) to 3.0 mm (CeRTES 5, 6, 7) with CeRTES 12 (1.1 mm) the exception (Table 3). Improved Cd types had a leaf width ranging from 1.6 mm (Wintergreen and Yukon) to 2.2 mm (Scotts R6LA and Sovereign). For hybrid types values ranged from 1.3 mm (Tif 00-10) to 1.9 mm (Tif 00-27), while for dwarf types values ranged from 1.0 mm (Roma) to 1.5 mm (Uganda). Spring green-up Green cover percentage evaluated in mid-April showed accessions scoring values above 80% and not statistically different from each other in each group (Table 4). In more detail, the accession with the best score was CeRTES 5 (93%); the lowest score was found in Sovereign (24%). Fall colour retention Green colour retention evaluated in mid-December showed a great variability within the groups (Table 4). The highest value was recorded for Tif 00-2, with the lowest for three dwarf types, Miniverde, Tifdwarf and Tifeagle (11%). Colour In Pisa, the highest and lowest values, Barazur (score 8.4) and Riviera (score 6.0) respectively, were recorded within the improved type group. In the wild type group, the values ranged from 6.1 (CeRTES 1) to 7.4 (CeRTES 4 and 9) (Table 4). With the exception of Barazur, the high- est values were recorded within the hybrid type group with values ranging from 7.1 (Tifsport) to 8.2 (Patriot), while Table 3 - Bermudagrass [Cynodon (L.) Rich.] accessions. Shoot density, horizontal stem density, node density and leaf width. Acces- sion effect averaged across locations Accessions Shoot density (n° cm-2) Horizontal stem density (cm cm-2) Node density (n° cm-2) Leaf width (mm) Wild types (Cd) CeRTES-1 0.8 1.5 0.7 2.8 CeRTES-2 0.7 1.7 0.8 2.8 CeRTES-3 0.9 1.5 0.6 2.8 CeRTES-4 0.8 2.9 1.8 2.5 CeRTES-5 0.9 1.9 0.9 3.0 CeRTES-6 1.0 3.1 1.7 3.0 CeRTES-7 0.6 2.1 0.9 3.0 CeRTES-8 0.5 1.8 0.8 2.9 CeRTES-9 0.9 2.3 0.9 2.8 CeRTES-10 0.9 2.1 0.9 2.1 CeRTES-11 1.0 2.5 1.1 2.7 CeRTES-12 2.4 5.1 3.0 1.1 CeRTES-13 0.7 1.9 1.1 2.8 Improved types (Cd) Argentina 1.0 4.0 1.8 2.1 Barazur 2.0 3.1 2.1 1.7 Bull’s Eye 1.9 4.9 2.4 1.9 Celebration 1.3 2.5 1.3 1.9 Grand Prix 2.2 2.9 1.6 1.8 Princess 77 2.1 3.6 1.9 1.8 Riviera 1.1 2.3 0.9 1.9 Scotts R6LA 1.3 2.1 1.0 2.2 Sovereign 0.9 3.3 1.4 2.2 SR 9554 1.3 2.2 0.7 2.0 Veracruz 1.3 3.4 1.6 1.8 Wintergreen 2.3 2.9 1.5 1.6 Yukon 1.2 2.1 1.1 1.6 Hybrid types (Cdxt) Patriot 1.7 3.9 2.0 1.8 Santa Ana 3.1 4.3 2.5 1.4 Tifsport 2.3 3.4 1.7 1.8 Tifway 2.6 3.8 2.0 1.4 Tif 00-1 3.8 4.3 2.3 1.5 Tif 00-2 3.2 3.3 2.1 1.5 Tif 00-7 3.2 3.2 2.2 1.5 Tif 00-10 3.5 3.7 2.0 1.3 Tif 00-18 2.7 2.7 1.6 1.4 Tif 00-24 2.0 3.7 1.9 1.4 Tif 00-27 2.3 3.6 2.1 1.9 Dwarf types (Cdxt/Ct) Champion 2.2 4.0 2.0 1.2 Miniverde 5.1 5.5 5.5 1.3 Tifdwarf 2.6 1.6 1.0 1.3 Tifeagle 3.1 4.4 4.2 1.1 Tifgreen 2.7 1.9 1.5 1.2 Roma 4.1 3.7 2.2 1.0 Uganda 3.0 4.7 2.6 1.5 LSD 0.05 0.7 1.6 1.2 0.3 33 for the dwarf types the values ranged from 7.0 (Tifgreen) to 7.9 (Miniverde). Quality In Pisa, the highest quality was recorded for the dwarf type Miniverde with a score of 8.4, however no signifi- cant differences were recorded within this group (Table 4). Wild types ranged in quality from 4.3 (CeRTES 1) to 7.0 (CeRTES 12), while improved types ranged from 5.8 (Riviera) to 7.7 (Barazur and Bull’s Eye); the variability of ratings within both these groups is worthy of note. The hybrid types scored from 6.8 (Tif 00-7) to 8.1 (Pa- triot and Tifway). 4. Discussion and Conclusions The study carried out on a pool of genetically and mor- phologically different entries belonging to the genus Cyn- odon has highlighted a wide variability of aesthetic and morphological traits. Morphological characteristics such as shoot density, node density, and horizontal stem density highlighted the better quality of the majority of dwarf and hybrid type cul- tivars, with improved types showing performances similar to those of wild types; CeRTES 12 was the exception. The genetic differences among groups are reflected more clearly with regard to leaf width, with values getting lower going from wild to dwarf types, with the exception of CeRTES 12. Recovery from winter dormancy in the spring, ex- pressed as spring green up, showed a great variability within and among the groups. This parameter is associated with carbohydrate reserves accumulated in storage organs as observed by Macolino et al. (2010). Other studies (Volt- errani et al., 2012) focused on carbohydrates in stolons in the first year of establishment and the relationship with growth and establishment rate. Cold tolerance, expressed as fall colour retention, high- lighted the better performances of the hybrid types with the dwarf types being the first cultivars in which dormancy begins. The parameters representing turf aesthetic quality (co- lour and turf quality), although they indicate a great vari- ability within groups, showed improving mean values from wild to dwarf types and confirmed what Patton et al. (2009) observed concerning the differences between im- proved and hybrid types. Acknowledgements The authors wish to acknowledge Dr. Wayne Hanna, the University of Georgia, for providing experimental clones of hybrid bermudagrass. This trial was carried out within the project “Sistemi avanzati per la produzione vivaistica di tappeti erbosi di specie macroterme ad uso multifunzionale a basso consu- Table 4 - Bermudagrass [Cynodon (L.) Rich.] accessions. Spring green- up (April 14 2011) and fall colour retention (percentage of green colour) (17 December 2011). Accession effect averaged across locations. Colour (visual estimation based on a 1-9 scale) and quality (visual estimation based on a 1-9 scale) refer only to the Pisa location (mean values May-October 2011) Accessions Spring green-up (%) Fall colour retention (%) Colour (1-9) Quality (1-9) Wild types (Cd) CeRTES-1 55 16 6.1 4.3 CeRTES-2 75 32 7.0 5.7 CeRTES-3 66 38 6.3 5.8 CeRTES-4 65 38 7.4 6.8 CeRTES-5 93 76 6.4 6.0 CeRTES-6 73 49 7.2 6.4 CeRTES-7 74 37 7.1 6.2 CeRTES-8 59 33 6.6 6.2 CeRTES-9 85 63 7.4 6.7 CeRTES-10 51 53 6.7 5.9 CeRTES-11 72 68 6.2 5.6 CeRTES-12 79 54 6.8 7.0 CeRTES-13 60 41 6.7 5.9 Improved types (Cd) Argentina 66 45 6.4 6.1 Barazur 55 43 8.4 7.7 Bull’s Eye 76 59 7.9 7.7 Celebration 49 59 7.5 6.4 Grand Prix 77 78 6.8 6.9 Princess 77 72 73 6.9 6.9 Riviera 57 50 6.0 5.8 Scotts R6LA 73 57 6.9 6.4 Sovereign 24 56 6.4 6.3 SR 9554 67 56 6.8 6.5 Veracruz 71 61 6.8 6.9 Wintergreen 83 69 7.1 6.5 Yukon 70 35 7.3 7.1 Hybrid types (Cdxt) Patriot 75 33 8.2 8.1 Santa Ana 77 73 7.4 7.9 Tifsport 79 68 7.1 7.2 Tifway 87 83 7.9 8.1 Tif 00-1 86 77 7.9 7.5 Tif 00-2 89 86 7.2 7.6 Tif 00-7 62 75 7.9 6.8 Tif 00-10 82 81 8.1 7.8 Tif 00-18 84 80 8.0 8.0 Tif 00-24 74 79 7.7 7.0 Tif 00-27 87 81 7.4 7.6 Dwarf types (Cdxt/Ct) Champion 86 76 7.7 8.0 Miniverde 53 11 7.9 8.4 Tifdwarf 63 11 7.8 7.9 Tifeagle 52 11 7.4 8.1 Tifgreen 72 31 7.0 7.9 Roma 75 56 7.4 8.0 Uganda 84 78 7.8 7.9 LSD 0.05 17 7 0.5 0.7 34 mo idrico ed energetico” funded by the Italian Ministry of Food, Forest and Agricultural Policies. References AGATI G., FOSCHI L., GROSSI N., GUGLIELMINETTI L., CEROVIC Z. G., VOLTERRANI M., 2013 - Fluorescence- based versus reflectance proximal sensing of nitrogen con- tent in Paspalum vaginatum and Zoysia matrella turfgrasses. - Europ. J. Agronomy, 45: 39-51. ANDERSON J.A., TALIAFERRO C.M., 2002 - Freeze toler- ance of seed producing turf bermudagrasses. - Crop Sci., 42: 190-192. ANDERSON J.A., TALIAFERRO C.M., WU Y.Q., 2007 - Freeze tolerance of seed and vegetatively propagated ber- mudagrasses compared with standard cultivars. - Online Applied Turfgrass Science Doi, 10.1094/ATS-2007-0508- 01RS. BALDWIN C.M., LIU H., 2013 - Aesthetic and physiological characteristics of 42 bermudagrass cultivars grown in the transition zone. - International Turfgrass Society Research Journal, 12: 391-404. BEARD J.B., SIFERS S.I., 1996 - New cultivars for Southern putting greens. - Golf Course Management, 64: 58-62. BLUM A., 1988 - Plant breeding for stress environments. - CRC Press, Boca Raton, Florida, USA, pp. 114-132. CROCE P., DE LUCA A., MOCIONI M., VOLTERRANI M., BEARD J.B., 2004 - Adaptability of warmseason turfgrass species and cultivars in a mediterranean climate. - Acta Hor- ticulturae, 61: 365-368. DE WET J.M., HARLAN J.R., 1970 - Biosystematics of Cyn- odon L.C. Rich (Gramineae). - Taxon, 19: 565-569. GIOLO S., MACOLINO M., BAROLO E., RIMI F., 2013 - Sto- lons reserves and spring green-up of seeded bermudagrass cultivars in a transition zone environment. - HortScience, 48(6): 780-784. GÓMEZ DE BARREDA D., REED T.V., YU J., McCULLOUGH P.E., 2013 - Spring establishment of four warm-season turf- grasses after fall Indaziflam applications. - Weed Techn., 27(3): 448-453. LULLI F., GUGLIELMINETTI L., GROSSI N., ARMENI R., STEFANINI S., VOLTERRANI M., 2011 - Physiological and morphological factors influencing leaf, rhizome and stolon tensile strength in C4 turfgrass species. - Functional Plant Biology, 38: 919-926. LULLI F., VOLTERRANI M., GROSSI N., ARMENI R., STE- FANINI S., GUGLIELMINETTI L., 2012 - Physiological and morphological factors influencing wear resistance and recovery in C3 and C4 turfgrass species. - Functional Plant Biology, 39: 214-221. MACOLINO S., SERENA M., LEINAUER B., ZILIOTTO U., 2010 - Preliminary finding on the correlation between water-soluble carbohydrate content in stolons and first year green-up of seeded bermudagrass cultivars. - HortTechnol- ogy, 20: 758-763. MORRIS K.N., SHEARMAN R.C., 2007 - NTEP turfgrass evaluation guidelines (last update: March 2014). - National Turfgrass Evaluation Program. NIKOLOPOULOU A.E., NEKTARIOS P.A., AIVALAKIS G., VOLTERRANI M., CHRONOPOULOS I., 2012 - Effects of rootzone CO 2 and O 2 levels on seed germination and stolon growth of Cynodon dactylon. - Acta Agriculturae Scandi- navica, 62(1): 53-61. PATTON A., RICHARDSON M., KARCHER D., TRAPPE J., 2009 - 2007 NTEP Bermudagrass trial-year 3 results. Ar- kansas Turfgrass Report 2009. - Ark. Ag. Exp. Stn. Res. Ser., 579: 20-24. PATTON A.I., RICHARDSON M.D., KARCHER D.E., BOYD I.W., REICHER Z.J., FRY J.D., McELROY J.S., MUN- SHAW G.C., 2008 - A guide to establishing seeded bermu- dagrass in the transition zone. - Online Applied Turfgrass Science. Doi, 10.1094 /ATS-2008-0122 -01MD. POMPEIANO A., GROSSI N., VOLTERRANI M., 2012 - Veg- etative establishment rate and stolon growth characteristics of 10 Zoysiagrasses in Southern Europe. - HortTechnology, 22: 1-7. POMPEIANO A., VOLPI I., VOLTERRANI M., GUGLIELMI- NETTI L., 2013 - N source affects freeze tolerance in bermu- dagrass and zoysiagrass. - Acta Agriculturae Scandinavica, 63: 341-351. ROCHE M.B., LOCH D.S., 2005 - Morphological and devel- opmental comparisons of seven greens quality hybrid ber- mudagrass (Cynodon dactylon (L.) Pers. X C. transvaalensis Burtt-Davy) cultivars. - International Turfgrass Society Re- search Journal, 10: 627-634. SHEARMAN R.C., 2006 - Fifty years of splendor in the grass. - Crop Sci., 46: 2218-2229. TALIAFERRO C.M., 1992 - Out of Africa. A new look at Afri- can Bermudagrass. - USGA Green Section Record, 30(4): 10-12. TALIAFERRO C.M., 2003 - Bermudagrass [Cynodon (L.) Rich.], pp. 235-256. - In: CASLER M.D., and DUNCAN R.R. (eds.) Turfgrass biology, genetics, and breeding. Wiley & Sons, Hoboken, New Jersey, USA, pp. 384. VOLTERRANI M., GROSSI N., GAETANI M., POMPEIANO A., 2010 - Zoysiagrass cultivar establishment rate and turf quality in central italy. - Acta Horticulturae, 881: 313-316. VOLTERRANI M., GROSSI N., LULLI F., GAETANI M., 2008 - Establishment of warm season turfgrass species by trans- plant of single potted plants. - Acta Horticulturae, 783: 77-84. VOLTERRANI M., MAGNI S., 2004 - Species and growing me- dia for sports turfs in Mediterranean area. - Acta Horticul- turae, 661: 359-364. VOLTERRANI M., POMPEIANO A., MAGNI S., GUG- LIELMINETTI L., 2012 - Carbohydrate content, character- ization and localization in bermudagrass stolons during es- tablishment. - Acta Agriculturae Scandinavica, 62(1): 62-69.