COMPARISON OF THE MORPHOLOGY, FLOWERING AND PHENOLOGY, CYCLE TYPE PLANTS OF LIFE IN GRINDELIA LANCEOLATA FROM CEDAR (ASTERACEAE) AND GLADES MIDDLE TENNESSEE IN COMMON NORTHERN ALABAMA: A GARDEN STUDY Adams^ Christopher M.Baskin CBaskin A. Jerry Carol Department of Biology Department of Biology Deportment of Biology Lexington, KY 40506, U.S.A. Lexington, KY 40506, U.S.A. Lexington, KY 40506, U.S.A. KY Lexington, 40546, U.S.A. ABSTRACT The life cycle type of Grindelia lanceolala Nutt. has been described as biennial, short-lived mono- and carpic perennial, (polycarpic) perennial in the taxonomic literature. Plants of this species in middle Tennessee cedar glades clearly are monocarpic. However, observations suggested field that those from glades in northern Alabama are at least dicarpic, and further that they differ morpho- logically and flower later than those in Tennessee glades. The purpose of this study was to deter- mine differences in morphology, flowering phenology and/or cycle type of Tennessee and Ala- if life bama plants are retained when grown from seeds "common garden" nonheated in a - in a i.e., greenhouse in Lexington, Kentucky Morphological differences (all statistically significant) between nam rosette and stem leaves; (2) ber of secondary basal stems; (3) height of primary stem; (4) number number and of capitula per plant; (5) of ray of disk flowers per capitulum; (6) diameter of capitulum; and (7) length of ray flower corolla Tennessee plants began flowering about month eariier than 1 . Alabama plants, and none of them produced basal rosettes after they flowered once (in their 2"'' year), confirming that they are strictly monocarpic. Alabama plants also flowered first in their 2"'' 66% them year; however, of have produced basal rosettes (which bolted and flowered) for five con- secutive growing seasons, confirming that they are polycarpic. Also, individual Tennessee plants many potentially can produce twice as seeds as individual Alabama plants during a single flower- ing/fruiting period. These results strongly suggest genetic differences exist in vegetative and floral We lanceolala. speculate that these differences could be associated with different ancestral geographic origins: Tennessee plants from a monocarpic race in the Ozarks and Alabama plants from a peren- campo observaciones de sugieren que aquellas de zor la ), mas que morfologicamente que icas; y alia, difieren y f loi alcareos de Tennesssee. El proposito de este estudio fue d fenologia floral y/o tipo de ciclo de vida de plantas de T{ rmando que so embargo, sin ( INTRODUCTION an herbaceous grows Grindelia lanceolata Nutt. (Asteraceae) species that in is open habitats on shallow soils underlain by limestone (Steyermark 1934, 1937; Baskin and Baskin geographical range extends from the Ozarks of 1979). Its Missouri and southeastern Kansas, south through eastern Oklahoma and north- ern and western Arkansas, and into northeastern and central Texas (Steyermark GPFA Nesom and Smith Johnston 1934, 1999; Correll 1970; 1986; 1988; 1990). Nesom (1990) also reports the species from the Monterrey area in Nuevo Leon, Mexico. Disjunct populations occur in the Central Basin of Tennessee (Chester 1997) and in northern Alabama (Small 1933; Harper 1944), where they are et al. & associated closely with open cedar glades (Baskin Baskin 1979, 1996; Baskin Grindelia lanceolata has been reported from Louisiana (Rydberg et 1995). al. 1932; Small 1933); however, Gandhi and Thomas (1989) do not list the species in their recent treatment of the Asteraceae of Louisiana. The species also has been reported from a single county in southeastern Ohio (Jones 1943; Fisher 1988), where, apparently, has been introduced (Porter 1956). it Nesom In his taxonomic treatment of Texas species of Grindelia, (1990) recognized three varieties of G. lanceolata: lanceolata, texana (Scheele)Shinners, known and greenei (Steyermark) Nesom, the only from Mexico. However, latter Correll and Johnston (1970) did not recognize any separate taxonomic entities and thus included only G. lanceolata in their treatment. Julian A. Steyermark listed two forms of G. lanceolata in his Flora of Missouri (1999): lanceolata and known latijolia Steyerm., the latter from only one county in Missouri. Small GPFA Fernald Gleason Gleason and Cronquist (1933), (1950), (1963), (1986), and (1991), Chester et (1997) do not recognize any intraspecific taxa in G. al, lanceolata. With regard to life cycle type, G. lanceolata has been reported to be bien- & (Rydberg monocarpic nial 1932), short-lived perennial (Baskin Baskin 1979; & GPFA 1986; Gleason Cronquist and (presumably polycarpic) perennial 1991), and Nesom (Small 1933; Correll Johnston 1970; Enquist 1987; 1990). In their study on and the autecology population biology of G. lanceolata in the limestone ce- Basm dar glades of the Central of Tennessee, Baskin and Baskin (1979) found that plants were short-lived monocarpic perennials plants lived for a few (i.e., years before they bolted and flowered once, and then died). The youngest plants to flower in the Baskins' study were in their third growing season (2+ years old). Other plants in the study flowered in their fourth or fifth year; plants died all flowering once monocarpic). after (i.e., However, during field studies in the limestone cedar glades in northern Alabama, and Alabama Baskin Baskin (pers. obs.) noticed that plants of G. lanceolata differed in morphology, flowering time, and cycle type. In con- life trast to plants in Tennessee, rosettes were present on those with dead flowering Alabama, stalks in suggesting that this species at least dicarpic in the cedar is glades of northern Alabama. The purpose of this study was to determine there if are distinct measurable or observable differences in vegetative morphology, flo- ral morphology, flowering phenology, and/or cycle type between G. life lanceolata plants from Tennessee and Alabama cedar To determine glades. if morphology and environmen- differences in cycle type are genetically-or life tally—based, both Tennessee and Alabama plants were grown from seed in a common common environment, garden experiment. a i.e., AND METHODS MATERIALS Growth Conditions All plants used in this study were grown from seeds in a nonheated greenhouse m autumn in Lexington, KY. Seeds were collected 1996 from G. lanceolata popu- growing and Alabama and sown lations in cedar glades in Tennessee (sepa- rately) on soil in metal flats. The greenhouse soil mix was a 3:1 (v/v) mixture of and limestone-derived topsoil river sand. Following germination, juveniles were num- transplanted to 15-cm-diameter plastic pots in spring 1997 and assigned a ber Morphological and cycle features of 103 Alabama plants and of 88 Ten- life num- nessee plants were monitored in this study; plants of both groups were bered consecutively. Temperatures in the nonheated greenhouse were recorded continuously with an thermograph electric for the duration of the five-year study period. From mean maximum minimum and these recordings, daily temperatures for ± January 6,95 ± February 10.95 ± March 14.48 ± 21.92 April ± May 28.6 ± 30.62 ± July 32.13 ± August 31.68 ± September 28.18 ± ± October 20.05 12.6 ± 3.03 month each for each of the five years were determined. These temperatures then maximum minimum mean and monthly were used to calculate (± SE) (±SE) photon teinperatures the study (Table Average daily photosynthetic (°C) for 1). LI-COR nonheated measured with irradiance at plant level in the greenhouse, a model LI-1000 data logger and three LI-190-SA quantum sensors, ranged from m 6 mol - d"l on overcast days to 25 mol m"^ d"^ on clear days during the grow- ing season (March to October) (Snyder et 1994). al. and Morphology Vegetative Floral To determme morphology, various char- there are differences in vegetative leaf if were compared between two groups Length, width, oven- acters the of plants. dry mass, and specific leaf area were determined for leaves collected from the rosette and from the lower-, middle-, and upper (just below the terminal ca- pitulum) portions of the main stem. One leaf each from these four regions of the shoot was removed and measured/ weighed every plant in the study Leaf for mm length and width were measured to the nearest using a standard metric ruler The width measurement was taken the widest part of the Leaves at leaf. were dried in an oven at 70'C for 24 hours, and their dry mass determined with an analytical balance. Average values for each leaf character were calculated made plants in the study Leaf prints were using Diazo-type paper for lor all am monia and was determined by weight developing, leaf area (one side of leaf) was of paper/area print relationships. Specific leaf area (SLA) calculated using SLA where the following equation: = Aieaf/Wieal Aieaf is leaf area (one side only) SLA and leaf dry weight. was determined for one leaf from each of the Wieai is four shoot levels for each plant in the study, and then average values were cal- number each secondary stems culated leaves of position. In addition, of basal for and height each were determined and means of plant averages calculated. All were compared by t-tests (P=0.05). Number of capitula the main stem and secondary stems were for for all m counted and averages calculated for each plant the study. Floral measure- ments were made on 15 Tennessee plants and on 15 Alabama plants selected randomly random-numbers Number using and a table. of ray disk flowers, length and width of ray flower corolla, and diameter were measured, to the near- mm, est for the terminal capitulum of the main stem of each of the 30 plants. The terminal capitulum is always the first to flower on a plant. Capitulum di- ameter and corolla length and width were measured using a standard metric ruler and means calculated. All means were compared by t-tests (P=0.05). Flowering Phenology The terminal capitula of the main stems were monitored for flowering. The flow- ering period in this study extended from the beginning of flowering (indicated by the first anther of the first disk flower to shed pollen; ray flowers are pistil- late) in the first plant and ended with the beginning of flowering in the last plant. Beginning of pollen shed was determined by brushing a finger across the anthers and observing clumps of the bright yellow pollen adhered to if it. Seed Production Potential number To assess the reproductive effort of G. lanceolata, potential of seeds per was number individual plant calculated. Potential of seeds per individual for Alabama and was number for Tennessee plants calculated as follows: Potential of seeds produced per plant = (Avg. no. capitula per main stem x Avg. no. ray flowers per capitulum) + (Avg. no. capitula per main stem x Avg. no. disk flow- ers per capitulum) + (Avg. no. capitula per secondary stem x Avg. no. secondary stems X Avg. no. ray flowers per capitulum) + (Avg. no. capitula per secondary stem X Avg. no. secondary stems x Avg. no. disk flowers per capitulum). and Morphology Vegetative Floral SLA Length, width, dry weight, and of leaves from the rosette, and from the and upper main lower-, middle-, portions of the stem, differed significantly between Tennessee and Alabama plants of G. lanceolata (Table Only rosette 2). dry weight and length of leaves on the lower portion of the stem were nonsig- from Alabama had and nificant. In general, plants longer, wider, heavier leaves, a higher SLA, and were than Tennessee However, Tennessee plants taller plants. produced significantly more secondary stems than Alabama plants (Table 2). Tennessee plants produced significantly more capitula per plant on both main stems and secondary, basal stems than did Alabama plants (Table Av- 3). erage number of capitula per main stem and per secondary stem was signifi- Alabama cantly higher for Tennessee plants. plants produced larger capitula SLA (cm7g) ± 0.057 0.004 Secondary Stems No. greater diameter) and more ray flowers per capitulum than did Tennessee (i.e., Number between plants. of disk flowers, however, did not differ significantly the two groups. Ray flowers from Alabama plants had longer petals than Ten- m was no width between two nessee plants, but there difference petal the groups. Flowering Phenology All plants from both groups bolted and flowered in their second year Flower- ing in Tennessee plants began on July 1998, and all terminal flowers of main 1 stems had flowered by August 1998 flowering Alabama 5 (Fig. In contrast, in 1). plants began and ended on 26July and 7 September, respectively For both groups, capitula elsewhere on the main stem and on secondary stems continued to from Tennessee and Alabama./ ?olata plants number produced by of capitula plants., all = whereas ficantly different test, P 0.05), N! (t Te^e.. Alabama Character Signif. Level Produced Total No. Capitula ": sf Secondary Stems Produced No. Capitula Main Stem ^ :::^;.3 Secondary Stems : Is : No. Ray Flowers m ± 34 NS 252 ± 71 Capitulum Diameter (cm) ± ± 4.4 0.2 5.7 0.5 main flower 1-2 weeks following the beginning of anthesis of the last terminal smgle After completion of flowering, Tennessee plants (n=88) died; not a all many one produced a new basal rosette. However, Alabama plants continued new to produce rosettes and to bolt and flower over the course of the five-year One-hundred and Alabama and study period three plants bolted (Figs. 2). 1, m m flowered 1998, 78 1999, 74 in 2000, 73 in 2001, and 68 in 2002 (Fig. 2). m m Flowering period duration was 43 days in 1998, 48 1999, 43 2000, 39 in m and 44 2002 Length flowering period 1998 was 36 days 2001, in (Fig. of for 1). and Tennessee plants (Fig. Flowering (both terminal all other capitula) for 1). both groups had been completed by mid-September. Seed Production Potential An individual Tennessee plant had the potential to produce an average of 3,128 Alabama had produce only seeds, while a single plant the potential to 1,568 seeds. DISCUSSION and Baskm Alabama appeared Observations by Baskin that G. lanceolata plants and from to be distinct in morphology, flowering time, life cycle type those in Alabama Tennessee were supported by the results of this study. plants are taller One more and have larger leaves than Tennessee plants. of the striking differ- number ences between the two groups the of secondary basal stems. Tennes- is many see plants produce more than two times as secondary basal stems as do Alabama giving them suffrutescent-like appearance. Indeed, Tennessee plants, a IS g i ^ b: r^ £r 3 B I lllfl IJ Rosettes Produced No. Plants Flower in Number of seeds per plant produced during a single reproductive event m was monocarpic considerably higher the perennial ("biennial") plants (Ten- nessee) than in the polycarpic plants (Alabama). Salisbury (1942) reported that many biennials in Great Britain produced an average of more than four times as Duffy monocarpic seeds as did polycarpic perennials. (1999) predicted that et al. species should maximize the number of meristems branches) devoted to (i.e., many which would reproduction, allocate as reserves as possible to reproduc- tive effort since there only one reproductive event. Conversely, they predicted is that polycarpic perennials should maximize the number of growth meristems premium on and (but not reproductive ones) since there a fitness longevity is many competitive ability. Tennessee plants produced 2.2 times as secondary which Alabama As basal stems, all ot flowered, as did plants. a result, Tennes- see plants produced, on average, significantly more capitula and had the po- many Alabama tential to produce almost twice as seeds as did plants. Thus, Tennessee plants obviously devoted more energy to reproductive meristems than did Alabama which devoted more energy growth height and plants, to in Whereas ma- perennation. all Tennessee plants died after flowering once, the jority of Alabama plants continued to flower each year for several years. By pro- new Alabama ducing fewer flowers, plants have reserve energy production of for rosettes perennation), which also should be considered growth meristems. (i.e., number Thus, there a tradeoff between of seeds produced per reproductive is and an event lifespan of individual. A very interesting difference between Tennessee and Alabama plants is life cycle type. In the nonheated greenhouse, both Tennessee and Alabama plants bolted and flowered in their second year However, whereas 88 Tennessee all Alabama plants died after they reproduced once, only 13 of 103 plants did so. 100% 13% Alabama Thus, of the Tennessee plants, but only of the plants, be- haved Another 12% Alabama were 12% were as biennials. of the plants dicarpic, and 1% dying and tricarpic, tetracarpic, after their third, fourth, fifth years, re- Alabama spectively. Sixty-six percent of the original 103 plants have flowered in their second, third, fourth, fifth, and sixth years, and it is expected that most of them could survive and flower for the next several growing seasons. Apparently, differences in morphology, flowering time, and cycle type life Alabama between Tennessee and plants are genetically-based, since they were m common grown maintained plants from seeds environment. However, in a the advantage, any, for G. ianceolata behaving as a short-lived monocarpic if perennial in Tennessee glades, and of behaving as a polycarpic perennial in it m the cedar glades northern Alabama, not known, Floristically and is Alabama vegetationally, cedar glades of northern are very similar to those in & Baskm Baskm central Tennessee (Baskin et al. 1995; 1999, in press). Further,