Nondestructive Techniques

There has been a great resurgence recently of interest in observational, nondestructive techniques for studying root-related processes. Several review volumes present detailed discussion of minirhizotron and rhizotron usage, including Taylor (1987), Box and Hammond (1990), and Fahey et al. (1999). In essence, the rhizotron approach involves installing a large glass plate in an observation gallery and then measuring the growth of roots against the glass over time (Fogel and Lussenhop, 1991) (Fig. 2.2). Using this technique, one can follow a large part of a given root population visible through the glass over various time periods. The disadvantages are that the soil profile must be recreated and re-tamped to an equivalent bulk density, or mass per unit volume of soil that closely approximates the density of the surrounding soil. It is also only a small fraction of an entire field or forest.

Minirhizotrons, on the other hand, have a smaller amount of surface area in one place, being tubular (5-7 cm in diameter), and are placed, as are the rhizotrons, at a 20-25° angle from the vertical (Fig. 2.3). However, being light and readily handled, they enable extensive replication in any given plot, experimental treatment, or entire field site. Tubes may be of either glass or a durable plastic such as polycarbonate. For example, Cheng et al., (1990) used 12 minirhizotron tubes in each replicate and two replicates per treatment (conventional tillage and no-tillage) in a study of sorghum root growth and turnover in a southeastern United States agroecosystem. Other studies have followed the dynamics of soil mesofauna, namely collembola, in fields under various crops in Michigan agroecosystems (Snider et al., 1990). Anumber of precautions should be employed in the usage of minirhizotrons, so as to avoid artifacts of placement. For example, total root biomasses can be underestimated in the top 7-10cm if inadequate care is taken to shield

FIGURE 2.2. University of Michigan Soil Biotron. (a) View of tunnel and aboveground laboratory from the south. (b) View from the west. Note white pine stump left after logging and burning in about 1917. (c) Interior of tunnel showing window bays covered with insulated shutters. (d) Close-up of glass, wire-reinforced, 6-mm-thick windowpane. Note fungal rhizomorphs. The wire grid is about 2 cm by 2 cm (from Fogel and Lussenhop, 1991).

FIGURE 2.2. University of Michigan Soil Biotron. (a) View of tunnel and aboveground laboratory from the south. (b) View from the west. Note white pine stump left after logging and burning in about 1917. (c) Interior of tunnel showing window bays covered with insulated shutters. (d) Close-up of glass, wire-reinforced, 6-mm-thick windowpane. Note fungal rhizomorphs. The wire grid is about 2 cm by 2 cm (from Fogel and Lussenhop, 1991).

the top of the minirhizotron tubes from transmitted light. Also, adequate soil-tube contact needs to be ensured by careful drilling and smoothing of the bored hole (preferably using a hydraulic coring apparatus), as noted by Box and Johnson (1987). If there is some open space between the outer tube surface and the soil, roots may respond as if this is a major soil crack and preferentially grow along it (van Noordwijk et al., 1993). To handle the large amounts of data and images obtained using minirhizotrons, it is necessary to use image analysis programs such as those described by Smucker et al. (1987), Hendrick and Pregit-zer (1992), and Pregitzer et al. (2002). With the advent of digital analysis techniques and image storage on CD-ROM, the literally millions of bits of information per soil-root image can be manipulated and analyzed reasonably promptly and efficiently. Caution must be taken, how-

FIGURE 2.3. An auger jig system used to install angled minirhizotron tubes (from Mackie-Dawson and Atkinson, 1991).

ever, to ensure that the material used for the tube has a minimal effect on the roots being observed. Withington et al. (2003) compared minirhi-zotron data for glass, acrylic, and butyrate tubes in an apple orchard, and acrylic and butyrate tubes in a study with six forest tree species. Root phenology and morphology were generally similar among tubes. Root survivorship varied markedly between hardwood and conifer species, however, probably because of hydrolysis by fungi interacting with the plastic tubes. Comparison of data from cores of root-standing crops with data from cores of minirhizotron-standing crops showed a closer match with the acrylic than the butyrate data. Glass was consid ered to be the most inert, but one-third of the glass tubes were lost as a result of breakage during the winter in the Pennsylvania site.

Frank et al. (2002), using minirhizotrons, measured significant increases in root growth in nine higher-elevation (1635-2370 meters) mixed-grass grazing lands in Yellowstone National Park. They found that large migratory herds of elk, bison, and pronghorn, by their grazing, stimulated aboveground, belowground, and whole-grassland productivity by 21%, 35%, and 32%, respectively. This feedback effect, which was demonstrated earlier by Dyer and Bokhari (1976) and McNaughton (1976), will be addressed further in system-level effects considered in Chapter 5.

In a study of seven minirhizotron data sets, Crocker et al. (2003) substituted root numbers for root lengths using a regression technique. Linear regression models were fitted between root length and root number for production and mortality of a wide range of tree species from subtropical to boreal conditions. Treatments yielded r2 values ranging from 0.79 to 0.99, indicating that changes in root numbers can be used to predict root-length dynamics reliably. Slope values for mean root segment length (MRSL) ranged from 2.34 to 8.38 mm per root segment for both production and mortality. Crocker et al. (2003) caution that the quantitative relationship between root lengths and numbers must be established for a particular species-treatment combination, but it will save on time needed to quantify root dynamics.

Additional approaches to calculation of fine root production and turnover continue to appear in the scientific literature. Gaudinski et al. (2001) used one-time measurements of radiocarbon (14C) in fine roots (<2 mm in diameter) from three temperate forests in the eastern United States—a coniferous forest in Maine, a mixed hardwood forest in Massachusetts, and a loblolly pine plantation in South Carolina. Roots were sampled as either mixed (live and dead) or live. Using accelerator mass spectrometry (AMS) to analyze very small samples, Gaudinski et al. (2001) found that root tissues are derived from recently fixed carbon, and the storage time prior to allocation to root growth is less than 2 years and more likely less than 1 year. Live roots in the organic horizons contain carbon fixed 3-8 years ago, versus roots in mineral B horizons with carbon of 11-18 years mean age. This spatial component to root age has not been measured before, and has important implications in calculating more realistic carbon budgets for terrestrial ecosystems. This assessment of mean root ages in forest tree roots is in marked contrast to the more rapid turnover times as noted above in the studies using minirhizotrons and other direct means of observation. It does not negate the findings of the observational studies, but emphasizes the need to be aware of the wide range of age of fine roots in the entire soil profile.

0 0

Post a comment