Factors Affecting Reproductive Success in a Rare Grass, Calamagrostis porteri Subsp. Insperata

Abstract

Calamagrostis porteri A. Gray subsp. insperata (Swallen) C. W. Greene is a rare grass that is currently treated as a of concern by the U.S. Fish and Wildlife Service. Production of viable seed had never been observed, and little was known about the factors affecting this reproductive failure. In this study we examined the effects of maternal genotype, pollen source, and light intensity on caryopsis (the single-seeded fruit produced in grasses) production. Maternal genotype and pollen source (self vs. outcross) significantly affected the number of caryopses initiated, while light intensity had no significant effect. However, nearly all of the caryopses initiated aborted late in development. In the 2000+ florets we examined, only one fully filled, viable caryopsis was found. The conditions under which this caryopsis was produced suggest future courses of management. Calamagrostis porteri A. Gray subsp. insperata (Swallen) C. Greene is a rare grass currently known from approximately 80 populations in five Midwestern states: Illinois, Indiana, Kentucky, Missouri, and Ohio. It was discovered in 1934 in Jackson Co., Ohio, and was originally described as C. insperata (Swallen, J. Wash. Acad. Sci. 25: 413. 1935). It was formerly a Category 2 federal candidate for listing as endangered or threatened ue in part to its extremely patchy distribution throughout its range (Drewry, 1993), and is now, like all former Category 2 candidates, called a of concern by the U.S. Fish and Wildlife Service (Drewry & Sayers, 1996). It grows primarily in forest openings and along edges of upland woods (Spooner et al., 1983). Calamagrostis porteri subsp. insperata was reported by Greene (1980) to be octoploid (2n = 56), although a related species, Calamagrostis canadensis, also 2n = 56, is reported to be tetraploid and has allozyme banding patterns typical of those expected in tetrasomic inheritance (Macdonald & Lieffers, 1991). Diploids (2n = 14) are not known in Calamagrostis, indicating that the genus is quite old and polyploidization occurred early in its evolutionary history (Greene, 1980). Although apomixis is prevalent in the genus, Greene (1984) postulated that Calamagrostis porteri subsp. insperata reproduced sexually due to the presence of sexual megagametophytes produced from reduced megaspores, as well as normal microsporogenesis and pollen formation. However, it has never, to our knowledge, been observed to set viable seed, which probably contributes to its rarity. Several hypotheses have been proposed to explain this reproductive failure. First, encroachment of woody vegetation and subsequent shading may inhibit flowering. Flowering tillers are very rare in most populations and are most often seen on plants in relatively sunny areas (Greene, 1984; Bittner & Gibson, 1993). Second, C. porteri subsp. insperata may be selfincompatible (Greene, 1980), and if so, reproduction may be limited by lack of compatible mates. Many populations consist of only a few plants, which could be vegetative clones, given that C. porteri can reproduce asexually via rhizomes (Greene, 1984). In addition, many plants lack viable pollen, as estimated by -stainability with cotton blue (Greene, 1980). Information regarding reproduction of many rare or endangered taxa, including C. porteri subsp. insperata, is scanty, but vitally important for management and recovery plans. The purpose of this study was to investigate the effects of light intensity, pollen source, maternal genotype and pollen viability on reproductive success in Calamagrostis porteri ' This study was supported by funds from the Missouri Botanical Garden and the Center for Plant Conservation. The authors thank Craig Greene, S. Ellen Macdonald, George Yatskievych, and Brian Stone for their thoughtful comments on the project and/or the manuscript. 2 Author for correspondence. Missouri Botanical Garden, P0. Box 299, St. Louis, Missouri 63166, U.S.A. Present address: Chicago Botanic Garden, 1000 Lake Cook Road, Glencoe, Illinois 60022, U.S.A. 3 Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri 63166, U.S.A. ANN. MISSOURI BOT. GARD. 85: 64-68. 1998. This content downloaded from 173.161.55.241 on Wed, 22 Jan 2014 15:46:28 PM All use subject to JSTOR Terms and Conditions Volume 85, Number 1 Havens & Holland 65 1998 Reproductive Success in Calamagrostis Table 1. Provenance (county and state of collection, voucher number and location, and Missouri Botanical Garden accession number), treatment, and pollen viability for the four genotypes used. Genotype Provenance Treatment Pollen viability 1 Jackson Co., OH, Subpopulation C Sun-Female 0% Voucher: Weiland 29 (MO) Accession # 901734 Shade-Female 95% 2 Jackson Co., OH, Subpoputation B Sun-Female 98% Voucher: Weiland 29 (MO) Accession # 901733 Shade-Female 20% 3 Vinton Co., OH Sun-Female 99% Voucher: Rogers 345 (MO) Accession # 872392 Shade-Female 98% 4 Texas Co., MO Sun-Male 99% Voucher: Summers 3631 (MO) Accession # 902331 subsp. insperata. We examined caryopsis (singleseeded fruits produced in grasses; functionally analogous to seeds in most other species of flowering plants) production and viability in two different light regimes and with three pollination treatments: self-pollination, natural crossing, and hand-outcrossed pollinations. MATERIALS AND METHODS Specimens of Calamagrostis porteri subsp. insperata were field-collected in 1987 and 1990 from Jackson and Vinton Cos., Ohio, and Texas Co., Missouri (Table 1). The plants are maintained in protective cultivation at the Missouri Botanical Garden as part of its cooperative agreement with the Center for Plant Conservation. The Center for Plant Conservation is an organization dedicated to preserving rare plants in the United States primarily through ex situ cultivation and seed banking. Only four genotypes were included in this experiment. These were the only genotypes that flowered in the ex situ collection. Since viable seed production has never been observed in natural populations, growing plants from seed was not an option. Three sets of two genetically identical clones were used as pollen recipients (hereafter females; it should be noted that although plants are hermaphroditic, we are referring only to their functional gender in this study). The female plants were collected from three different populations or subpopulations in Ohio. The three clumps of tillers were divided and potted separately to produce the pairs of clones. The pollen donor (hereafter male) was collected in Missouri and is presumably genetically distinct from all of the females. All plants were grown in a shaded lath house until flowering culm initiation. Upon flowering, in mid June 1994, one member of each pair of clones was placed in full sun, the other member emaining in the shaded lath house for the duration of the experiment. Inflorescence number varied on the female plants from 1 to 65. Two to four of the inflorescences on each female were bagged with nylon stockings prior to anthesis to test for self-compatibility. Approximately half of the remaining inflorescences were hand-pollinated by shaking one of the male's inflorescences over the female inflorescence. The remaining inflorescences were not manipulated and allowed to outcross or self-pollinate naturally. Hence, each female plant received three treatments, except one plant from genotype #1 that had only one flowering culm, and it received the handpollination treatment. Inflorescences were harvested 22 August 1994 after they had turned brown. Approximately 100 florets were randomly selected and removed from each inflorescence. These florets were scored for the presence or absence of caryopses and, when present, their developmental stage. To assess viability, caryopses were placed in petri dishes on moist filter paper in an incubator at 220C under long day conditions (16/8 photoperiod) and monitored for germination for one month. Half' of the caryopses from each treatment were given a cold treatment (12 weeks at 4VC) prior to incuba-