Environmental Engineering Reference
In-Depth Information
at the farm gate for $40-45/Mg (Perrin et al. 2008). Assuming a switchgrass farm-gate cost of
$40-54/Mg and conversion of 0.329 L ethanol/kg of switchgrass, the farm-gate feedstock cost
would range from $0.12 to 0.16/L.
22.7
BreedInG and selectIon
22.7.1 E
cotypic
v
ariation
The presence of obvious ecotypic variation has been a strong driver of many switchgrass activities
because many of the original collections were made from prairie remnant plants. Although the
upland and lowland types form an obvious visual dimorphism, there is a wealth of ecotypic variation
within both upland and lowland types. Early plant collectors recognized morphological variations
among remnant prairies and sought to collect accessions that would capture much of this variation.
Many of the early switchgrass cultivars were simple seed increases of random plants collected
from a single source-identified prairie remnant, what are today termed “natural-track” cultivars. A
modest amount of selection occurred in determining which of these accessions would be elevated to
cultivar status and released to the public. This was accomplished by common-garden experiments in
which numerous accessions were evaluated for many years, sometimes followed by more extensive
testing at multiple locations. In a few cases, some selection within the strain occurred for vigorous
plants and/or traits that would allow for adequate seed production.
These early selection efforts utilized genetic variability among accessions or strains of switchgrass,
which is relatively easily observed in replicated experiments (Eberhart and Newell 1959; Hopkins
et al. 1995b; Casler 2005). Taken together, these natural-track cultivars provide a fairly effective
representation of the morphological and physiological diversity associated with the diverse geography
and climates found in the eastern two-thirds of the United States (Table 22.1). Obvious from Table
22.1 and from investigation of switchgrass germplasm resources, such as those present in the USDA
National Plant Germplasm System, Germplasm Resources Information Network (GRIN; http://
www.ars-grin.gov/npgs/index.html) is the relative difference in frequency of upland vs. lowland
accessions, the latter being relatively rare in prairie remnants (Hultquist et al. 1997; Casler 2005).
Lowland ecotypes of switchgrass tend to be more southern adapted than upland ecotypes, as
a group. Because of their late heading and ability to maintain photosynthetically active tissue
through the longer growing season, lowland ecotypes generally have higher biomass yield than
upland ecotypes at southern locations located in USDA hardiness zones 6-9 (Sanderson et al.
1999; Fike et al. 2006b). Lowland ecotypes had 44% higher biomass yield than upland ecotypes
under a one-harvest management, whereas this advantage was reduced to 13% under a two-cut
management (Fike et al. 2006b). The upland ecotypes used in this study were more closely matched
in phenology to the photoperiods of these test locations. Because they were near anthesis at the
time of first harvest, there was sufficient time to allow a second period of vegetative development,
unlike the lowland ecotypes for which first harvest occurred before their first period of vegetative
development was completed (Fike et al. 2006b). Even though the lowland ecotypes had the highest
biomass yields, their use in a geographic region considerably north and east of their origin is likely
responsible for their more inconsistent responses to harvest management and other increased inputs
compared to upland ecotypes (Fike et al. 2006a).
With the recent expansion of breeding and selection activities on switchgrass in the United
States, upland-lowland intermediate types have been found with increasing frequency. The exact
origin and nature of these accessions is not known. In some cases, they occur sympatrically with
obvious upland ecotypes, suggesting their possible hybrid origin (Hultquist et al. 1997). Cave-
in-Rock appears to be an intermediate ecotype, sharing traits of both classic upland and lowland
phenotypes. Interestingly, Cave-in-Rock is an octaploid, suggesting a fairly complex genetic history
if it indeed arose as a hybrid between tetraploid upland and lowland plants. Analysis of DNA from
“Miami” switchgrass also provides conflicting data for classification, with cpDNA suggestive of the
