RESEARCH Searls Prairie Clover (Dalea searlsiae) for Rangeland Revegetation: Phenotypic and Genetic Evaluations Kishor Bhattarai, B. Shaun Bushman,* Douglas A. Johnson, and John G. Carman K. Bhattarai and J.G. Carman, Dep. of Plants, Soils, and Climate, Utah ABSTRACT State Univ., Logan, UT 84322-4820; B.S. Bushman and D.A. John- Few North American legumes are available for son, USDA-ARS Forage and Range Research Lab., Utah State Univ., use in rangeland revegetation in the western Logan, UT 84322-6300. Mention of a proprietary product does not USA, but Searls prairie clover [Dalea searlsiae constitute a guarantee or warranty of the product by USDA, Utah State (A. Gray) Barneby] is one that holds promise. University, or the authors and does not imply its approval to the exclu- Commercial-scale seed production of this spe- sion of other products. Received 1 July 2010. *Corresponding author: cies could address the issues of unreliable seed ([email protected]) availability and high seed costs associated with Abbreviations: ADF, acid-detergent fi ber; AFLP, amplifi ed fragment its wildland seed collection. To evaluate its util- length polymorphism; AMOVA, analysis of molecular variance; CP, ity for revegetation, we collected Searls prairie crude protein; DMY, dry-matter yield; Dp, Dalea purpurea; Ds, Dalea clover at 20 locations across Utah and Nevada. searlsiae; MCMC, Markov chain Monte Carlo; NDF, neutral-detergent Amplifi ed fragment length polymorphisms fi ber; NJ, neighbor-joining; PC, principal component (AFLP) and morphological and phenotypic traits (measured in common-garden plots) were used he spread of invasive weeds in the Great Basin region has to clarify the role of evolutionary forces respon- Tresulted in its identifi cation as the third most endangered sible for its genetic structure. Collections were ecosystem in the USA (Stein et al., 2000). A large portion of evaluated for dry-matter yield, infl orescence rangeland in this region burns each year (4.97 × 105 ha, the aver- weight, number of infl orescences, plant height, age from 1998 to 2007; Mike Pellant, personal communication, foliage diameter, fl owering date, acid-detergent fi ber, neutral-detergent fi ber, and crude protein 2009), and if left untreated, weed invasions will probably acceler- at two common-garden locations in northern ate (Jessop and Anderson, 2007). Postfi re revegetation requires Utah. Collections from southern Utah and east- large amounts of seed. For example, in 2007 the Bureau of Land ern Nevada exhibited high phenotypic values, Management purchased more than 3 × 106 kg of seed for revege- whereas collections from western Nevada and tation. A mix of species is preferred in seed mixtures for revegeta- northwestern Utah had low phenotypic values. tion because functionally diverse plant species are more likely to Collections from northwestern Utah were genet- minimize the risk of weed invasion and increase ecosystem func- ically differentiated from those of southern Utah tion (Pokorny et al., 2005; Sheley and Carpinelli, 2005; Walker and Nevada via AFLP markers. Strong isolation and Shaw, 2005). North American legumes are of particular inter- by distance between collections suggests that est for inclusion in seed mixtures because they fi x nitrogen, thus genetic drift and gene fl ow are important factors enhancing the nitrogen content of forage (van der Heijden et al., in determining population structure in Searls 2006), as well as improving wildlife forage and habitat (Madison prairie clover. Published in Crop Sci. 51:716–727 (2011). doi: 10.2135/cropsci2010.07.0387 Published online 10 Jan. 2011. © Crop Science Society of America | 5585 Guilford Rd., Madison, WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher. 716 WWW.CROPS.ORG CROP SCIENCE, VOL. 51, MARCH–APRIL 2011 and Robel, 2001). Seeds of North American legumes for the climatic variables of source collection sites, suggest the use in revegetation programs are usually harvested from possibility of local adaption (Endler, 1986). Climatic variables wildlands, a practice that results in unreliable seed avail- that correlate with common-garden phenotypes can be used ability and high seed prices, in turn limiting the use of to delineate regional seed sources, as has been done for tree these legumes (Walker and Shaw, 2005). Commercial species (Johnson et al., 2004). Neutral DNA markers (e.g., seed production could reduce seed costs and make such amplifi ed fragment length polymorphisms, or AFLPs) mea- species more readily available for rangeland revegetation. sure overall genetic diversity due to evolutionary forces, such Plant materials for rangeland revegetation are consid- as gene fl ow, founder eff ects, selection, and genetic drift. As ered to be appropriate if they are adapted to and compatible a result, genetic diversity data obtained from DNA markers with the local revegetation environment, do not pose the indicate the evolutionary potential of populations (McKay and risk of gene transfer to related species that do not already Latta, 2002). In addition, DNA-marker analyses provide esti- occur naturally, are nontoxic to herbivores (because a large mations of inbreeding depression based on diversity within a portion of western rangelands are used for livestock graz- collection, as well as group diff erentiation based on variation ing), and are amenable to commercial seed production to among collections (Ouborg et al., 2006; Bonin et al., 2007). ensure a reliable quantity and quality of seeds. Searls prai- When the distribution of populations is beyond the rie clover [Dalea searlsiae (A. Gray) Barneby] is a perennial gene-fl ow capability, isolation by distance can occur. Iso- legume that naturally occurs in the southern Great Basin, lation by distance is usually detected using the Mantel test the southwestern Colorado Plateau, and northern Arizona (Mantel, 1967), by a positive correlation between genetic (USDA-NRCS, 2010). It is diploid (2n = 14, or rarely, 16) and geographic distances among individuals. Strong isola- and is primarily insect pollinated (Barneby, 1977; Jim Cane, tion by distance indicates that populations are at equilib- personal communication, 2009). Its natural distribution rium with respect to gene fl ow and genetic drift (Slatkin, within the southern Great Basin ameliorates gene-transfer 1993). However, other factors, such as precipitation, tem- concerns in that region because hybridization with related perature, elevation, or disturbance, can infl uence isolation species would have occurred naturally, if at all. Compounds by distance (Kittlein and Gaggiotti, 2008); therefore, a that are toxic to livestock and wildlife and are found in partial Mantel test is often used to account for the infl u- some other legumes were below detectable levels in Searls ence of other factors when isolation by distance is esti- prairie clover (unpublished data). Additionally, its relatively mated (Telles and Diniz-Filho, 2005). upright growth habit makes Searls prairie clover a potential The objective of our study was to evaluate phenotypic candidate for commercial seed production. and DNA-marker-based variation of 20 wildland collec- Local germplasm has been recommended for use tions of Searls prairie clover. Forage production and qual- in revegetation to reduce the possibility of maladaption ity, fl owering date, infl orescence weight, and growth habit or outbreeding depression (Lesica and Allendorf, 1999; were assessed in two common gardens, and genetic diversity McKay et al., 2005). However, disturbances and global among and within collections was determined using AFLP changes on rangelands may alter the ecosystem such that markers. With these data, the role of genetic drift and gene local germplasm is no longer optimal (Broadhurst et al., fl ow among populations were evaluated. We also identi- 2008), and maintaining and producing many local germ- fi ed potentially adaptive phenotypic traits and regional seed plasm accessions for seed production can be diffi cult for sources that might minimize the possibility of outbreeding seed growers (McKay et al., 2005; Broadhurst et al., 2008). depression and maladaptation while providing a tractable The latter issue may lead to increased seed costs because number of seed sources for germplasm development. the demand for local germplasm is unpredictable, given the sporadic nature of wildfi res and subsequent variable MATERIALS AND METHODS need for local germplasm. Using phenotypic and genomic Plant Material Collection assessments to pool seed sources for native species into Seed of Searls prairie clover was collected from 20 sites in Utah larger regional levels can minimize both the number of and Nevada during the summer of 2005 (Fig. 1). Elevation, lati- seed sources required for use in revegetation and resto- tude, and longitude data were obtained for each collection site ration programs and the risk of maladaptation and out- (Table 1). Mean annual temperature and precipitation (average breeding depression. from 1961 to 1990) of the collection sites were obtained from The suitability of regional seed sources can be deter- the Moscow Forestry Sciences Laboratory (2009). A 5- to 15-g mined by