Visualizing and measuring root growth in situ is extremely challenging. We present a customizable rhizobox method to track root development and proliferation over time in response to nutrient enrichment. This method is used to analyze maize genotypic differences in root plasticity in response to an organic nitrogen source.
Roots are notoriously difficult to study. Soil is both a visual and mechanical barrier, making it difficult to track roots in situ without destructive harvest or expensive equipment. We present a customizable and affordable rhizobox method that allows the non-destructive visualization of root growth over time and is particularly well-suited to studying root plasticity in response to localized resource patches. The method was validated by assessing maize genotypic variation in plasticity responses to patches containing 15N-labeled legume residue. Methods are described to obtain representative developmental measurements over time, measure root length density in resource-containing and control patches, calculate root growth rates, and determine 15N recovery by plant roots and shoots. Advantages, caveats, and potential future applications of the method are also discussed. Although care must be taken to ensure that experimental conditions do not bias root growth data, the rhizobox protocol presented here yields reliable results if carried out with sufficient attention to detail.
Although often overlooked compared to their aboveground counterparts, roots play a critical role in plant nutrient acquisition. Given the substantial carbon cost of root construction and maintenance, plants have evolved mechanisms to develop roots only where foraging is worth the investment. Root systems can thus efficiently and dynamically mine resource patches by proliferating in hotspots, upregulating rates of uptake, and rapidly translocating nutrients to the phloem for further transport1. Plasticity responses can vary widely among plant species or genotypes2,3 and depending on the chemical form of the nutrient involved4,5. Variation in root plasticity should be explored further, as understanding complex root responses to heterogeneous soil resources could inform breeding and management strategies to increase nutrient use efficiency in agriculture.
Despite its necessity and relevance for understanding plant systems, visualizing and quantifying root plasticity at relevant scales poses technical challenges. Excavating the root crown from the soil ("shovelomics"6) is a common method, but fine roots exploit small pores between soil aggregates, and excavation inevitably leads to some degree of loss of these fragile roots. Furthermore, destructive harvest makes it impossible to follow changes in one root system over time. In situ imaging methods such as X-ray computed tomography allow direct visualization of roots and soil resources at high spatial resolution7, but are expensive and require specialized equipment. Hydroponic experiments avoid constraints associated with extracting roots from soil, but root morphology and architecture differ in aqueous media as compared to the mechanical constraints and biophysical complexity of soils8,9. Finally, rhizosphere processes and functions cannot be integrated with developmental plasticity in these artificial media.
We present a protocol for the construction and use of rhizoboxes (narrow, clear-sided rectangular containers) as a low-cost, customizable method to characterize root growth in soil over time. Specially designed frames encourage roots to grow preferentially against the back panel due to gravitropism, increasing the accuracy of root length measurements. Rhizoboxes are commonly used to study root growth and rhizosphere interactions10,11,12, but the method presented here offers an advantage in simplicity with its single-compartment design and inexpensive materials, and is designed to study root responses to localized nutrients. However, the method could also be adapted to study a range of other root and rhizosphere processes such as intra/interspecies competition, spatial distribution of chemical compounds, microbes, or enzyme activity. Here, we investigate genotypic differences among maize hybrids in response to patches of 15N-labeled legume residue and highlight representative results to validate the rhizobox method.
1. Preparation of the Front and Back Panels, and Spacers
Figure 1: Layout of drilled holes. Holes are drilled 1.3 cm from the side edges at 2.5, 19, 38, and 53.3 cm from the top, and 1.3 cm from the bottom edge at 2.5, 20.3, and 38 cm from the left margin. Please click here to view a larger version of this figure.
2. Installation of a Strip of Polyester Batting at the Bottom of the Box
NOTE: This will prevent soil and water from leaking through joints between spacers.
Figure 2: Assembled rhizobox with batting. A narrow strip of batting at the bottom of the rhizobox prevents soil and sand from leaking out. Please click here to view a larger version of this figure.
3. Assembly of the Rhizoboxes
Figure 3: Patch spacers. Screws inserted through the center of HDPE strips keep them from falling into the box. The rhizobox is filled with soil around the spacers, the soil is wetted, and the spacers are removed in order to leave empty treatment and control patches. Please click here to view a larger version of this figure.
4. Building PVC Frames to Support the Rhizoboxes at an Angle
NOTE: When the box is placed at an angle, gravitropism will encourage the roots to grow against the back panel so that all roots are visible for tracing. The polyvinyl chloride (PVC) dimensions in Figure 4 result in a frame that maintains the rhizobox at an approximately 55° angle to the bench.
Figure 4: Frame to support rhizoboxes. The lightweight frame is constructed from PVC cut to the specified lengths and connected using the joint types indicated. Please click here to view a larger version of this figure.
5. Sewing Protective Cases to Reflect Light and Heat
NOTE: Roots avoid light, so these cases exclude light in order to make sure that observed root plasticity responses are driven by the nutrient source in the patches and not by light avoidance. Light deprivation fabric also reduces the temperature inside the rhizoboxes, helping to avoid heat stress.
6. Preparation of 1:1 (V/V) Soil : Sand Substrate to Fill the Rhizoboxes
7. Substrate Preparation for the Treatment and Control Patches
8. Loading Rhizobox with Substrate, and Establishing Treatment and Control Patches
9. Even Watering to 60% Water-Holding Capacity
NOTE: This amount of soil moisture was found to prevent plants from experiencing drought stress while preventing the development of anoxic conditions or algal growth.
10. Seed Germination and Transplantation
11. Plants Growth
12. Harvesting Shoots, and Obtaining Root and Soil Samples for Analysis
13. Validation of Tracings and Estimation of Relative Root Growth Rates
14. Analysis of 15N partitioning among root, shoot, and treatment soil samples
Roots grew preferentially against the back of the box, as anticipated. Total traced root length on the back of the box ranged from 400 to 1,956 cm, as compared to 93-758 cm on the front of the box. Pairwise Pearson correlation coefficients were calculated between scanned root length and traced root length on the front of the box, back of the box, and the sum of front and back was used to determine whether tracing accurately reflected total root length (n = 23, as the plant in one box died during the experiment). Scanned total root length was significantly correlated with traced root length on the back of the box (Figure 5A, p = 0.0059), front of the box (Figure 5B, p = 0.022), and sum of back and front (Figure 5C, p = 0.0036). Tracing only the back of the box is thus validated as giving a representative measure of root growth while halving the time required to trace roots. It should be noted, however, that tracing captures only 21.6-54.6% of total root length. While roots do grow preferentially against the surface of the rhizobox, fine lateral roots in particular may not be visible for tracing. Tracing is well-suited to relative comparisons of root length over time, especially early in development, but harvesting and scanning root systems is preferable if the goal is to accurately quantify total root length.
Figure 5: Correlations between traced and scanned root length data. A) Traced root length was significantly correlated with scanned root length when only the front of the box was traced. B) Tracing roots on the back of the box, where the majority of roots were visible, improved the R2 value of the regression against scanned root length over tracing the front of the box; the correlation was again significant. C) The most accurate method is to trace roots on both sides of the box, as shown by the highest R2 value of the three methods and the significant correlation with scanned root length. Please click here to view a larger version of this figure.
Root growth rates over time were similar among boxes, as shown by consistent slopes when plotting the natural log of total root length against time (Figure 6). While slight variability is to be expected, consistent growth rates indicate that experimental conditions were uniform for all boxes. Dramatically different slopes would indicate that plants were growing at different rates, suggesting the need to check for differences in variables such as temperature or moisture.
Figure 6: Root growth rates over time. Similar slopes of root length vs. time among rhizoboxes indicate that roots grew at equal rates. Non-uniform slopes could indicate that experimental conditions vary among rhizoboxes. Please click here to view a larger version of this figure.
Roots of all maize genotypes proliferated in patches containing 15N-labeled cover crop residue. Two-way ANOVA with patch type and genotype as fixed factors (n = 23) revealed that root length density was higher in treatment than control patches using scanned root data (Figure 7a, p = 0.013) as well as traced root data (Figure 7b, p = 0.005). RLD was not significantly different among genotypes in either case.
Figure 7: Root length density by genotype and root data type. a) Scanned root data showed that all genotypes (A-F) proliferated in the treatment (T) patch, and genotypic differences were not significant. b) Harvested and scanned root data confirmed the significant effect of legume residue, but not genotype (A-F), on root length density in patches. Letters A-F represent six different genotypes and error bars represent standard error. C: control; T: treatment. Please click here to view a larger version of this figure.
Root diameter can be used to make inferences about root function and turnover. Fine roots are more likely to be lateral roots that rapidly develop and proliferate in response to resource hotspots, while larger coarse roots are more likely to be long-lived, slow-to-respond axial roots. Scanned root systems were analyzed for the proportion of roots in each diameter class: <0.2 mm, 0.2-0.4 mm, 0.4-0.8 mm, 0.8-1.6 mm, and >1.6 mm, and each size class was tested for genotypic differences. Genotypes with more fine roots in treatment patches might indicate a more effective proliferation response. One-way ANOVA with genotype as a fixed factor (n = 23) revealed that genotypes did not differ in root length in each size class for the root system overall (Figure 8a), treatment patches (Figure 8b), or control patches (Figure 8c). A majority of the roots were fine roots (<0.2 mm).
Figure 8: Proportions of roots in different diameter classes by genotype and location. a) In each rhizobox (excluding treatment and control patches), the majority of roots were fine (<0.2mm in diameter). Genotypes did not differ in the proportions of roots in each diameter class. b) In treatment patches, root length per size class likewise decreased with increasing diameter across genotypes. c) Control patches were characterized by the same patterns. Letters A-F represent six different genotypes and error bars represent standard error. Please click here to view a larger version of this figure.
Label N was higher in shoot than root samples across genotypes according to two-way ANOVA with sample type and genotype as fixed factors (n = 23, Figure 9), showing that 77-81% of N taken up from the treatment patch was translocated from roots to shoots during the experiment. One-way ANOVA (n = 23) showed that δ15N of root and shoot samples did not vary by genotype.
Figure 9: Nitrogen obtained from legume residue in roots and shoots at harvest. All genotypes were equally effective at taking up N from the patch containing 15N-labeled legume residue. The majority of N taken up from the patch was translocated from roots to shoots. Letters A-F represent six different genotypes and error bars represent standard error. Please click here to view a larger version of this figure.
The rhizoboxes described in this protocol can be used to answer varied questions in root and rhizosphere science, and have found diverse uses elsewhere10,20,21,22,23,24,25. Other researchers have captured time-lapse images of rhizoboxes21,25,26, some using automated systems22,27. These approaches may be used for quantitative analyses of root length and architecture not possible with tracing methods. Rhizoboxes have also been used to visualize microbial communities with techniques such as fluorescent in situ hybridization (FISH) and micro-autoradiography (MAR)21,22, or to capture spatially explicit patterns of water and nutrient resources with RGB imaging24 or extracellular enzyme activity with zymography11,30. The rhizoboxes presented here are unique from previous designs in that they are relatively large, making it possible to study species with extensive root systems; have a simple single-compartment design; use readily available, inexpensive materials; and are specially designed to study localized patches. The versatility of this rhizobox protocol could allow it to be customized for a range of other applications in root plasticity and rhizosphere interactions. Other nutrients could replace nitrogen in the patches. Immobile nutrients such as phosphorus would subject to lesser leaching, likely making them a good fit for this approach. The rhizoboxes are also well-suited to comparisons of bulk and rhizosphere soil, as the zone of root influence (a long-standing definition for the rhizosphere15) can be more clearly delineated than in pot studies and separated from bulk soil with a razor at harvest. Adapting this method to study rhizosphere processes opens up a broad new range of ways to extend the protocol, including study of both abiotic and biotic interactions23.
The rhizobox method presented here is well-suited to measuring relative differences among genotypes or species in root growth early in development, characterizing relationships among root traits, and exploring the effects of soil characteristics on root growth. Certain steps of the protocol are particularly critical because they affect factors with disproportionate influence on root growth: soil moisture, bulk density, and slope. Watering evenly and equally across boxes is critical given the influence of soil moisture on root growth patterns1,31. Pilot experiments showed that water delivered through drip emitters became evenly distributed after 24-48 h due to capillary action; however, variability among emitters in the volume of water delivered during a given irrigation period was too high to recommend this method using our arrangement. Hand watering slowly and evenly was the best of the techniques tested, but other irrigation methods are certainly possible. Watering to a previously calculated weight ensures that all boxes maintain the same soil moisture content, preventing variability in root growth due to water stress32. The weight of empty rhizoboxes can vary significantly, so calculating ideal weights for individual boxes is important.
As with moisture, establishing even bulk density of the substrate throughout the rhizobox is critical to measure root proliferation responses. Roots grow more extensively in less-dense soil33 and compaction may create channels for preferential water flow, further affecting resource distribution and root growth patterns34. Drying and sieving field soil, thoroughly mixing sand and soil, and moving the funnel back and forth slowly and evenly help create a homogeneous matrix for root growth.
A number of components within this experiment will depend on the research questions of interest as well as the soil type and plant species utilized within the study. The ratio of sand to soil may need to be adjusted for soils of different texture, to ensure that the substrate wets evenly without clumping. Pilot experiments showed that a 1:1 (v/v) soil : sand mixture was superior to 1:2 or 1:3 mixtures for the soil used in this experiment, a very fine sandy loam. The dimensions of the rhizoboxes and duration of the experiment can be adjusted depending on the research question, root traits, and plant species of interest. Maize is a relatively fast growing plant species; we therefore selected larger rhizobox dimensions compared to other rhizobox studies20,35 to provide a sufficient volume of soil that allows the experiment to continue for the optimal duration as plants become increasingly rootbound over time. Finally, the fitting of statistical models to visible and traced root growth must be determined for each study and species of interest, perhaps with pilot experiments. Visible root growth may follow a saturating curve rather than a linear model25,36, and the slope of the regression of visible on harvested roots varies by plant species21.
Roots have an inherent tendency to pursue gravity and avoid light37, but care must be taken to ensure that gravitropism and light avoidance do not bias root growth data. If greenhouse benches are tilted slightly, for example, roots will grow down the slope rather than responding to nutrient patches. Similarly, light gradients in the greenhouse could cause shifts in shoot growth and roots to grow preferentially on one side if the rhizoboxes are not kept completely dark. The light deprivation cases presented in the protocol are an effective means of reducing incident light, but other methods such as wrapping the rhizoboxes in aluminum foil may also work.
The non-destructive rhizobox method presented here facilitates the tracking of roots in situ over time38 and can be implemented by a graduate student with a basic working knowledge of power tools. As such, it offers advantages over destructive harvesting methods such as shovelomics6, which are better suited to studying mature root architecture but do not allow repeated measurements of the same root system over time, and MRI or X-ray computed tomography39,40, which allow high-quality imaging of root-substrate interactions but requires expensive equipment. However, it is not without limitations. Construction of the rhizoboxes can be time-consuming, and ensuring that factors such as moisture, bulk density, slope, and light are as uniform as possible, as described above, is non-trivial. Nonetheless, given sufficient time and attention to detail, the rhizoboxes deliver reliable results and can be reused for many experiments.
The authors have nothing to disclose.
The authors would like to acknowledge anonymous reviewers for their feedback, as well as J.C. Cahill and Tan Bao for initial guidance on developing the rhizobox protocol. Funding was provided by the Foundation for Food and Agriculture Research, the US Department of Agriculture (USDA) National Institute of Food and Agriculture, Agricultural Experiment Station Project CA-D-PLS-2332-H, to A.G. and by the UC Davis Department of Plant Sciences through a fellowship to J.S.
1.27 cm diameter PVC pipe | JM Eagle | 530048 | 305 cm per box, cut into lengths as specified in the protocol |
PVC side elbows | Lasco | 315498 | 2 per box |
PVC 90-degree elbows | Charlotte | PVC 02300 0600 | 4 per box |
PVC T joints | Charlotte | PVC 02402 0600 | 4 per box |
Extruded acrylic panes | TAP Plastics | N/A | 2 per box, 0.64 cm thick x 40.5 cm wide x 61 cm long |
HDPE spacers (sides) | TAP Plastics | N/A | 2 per box, 0.64 cm thick x 2.5 cm wide x 57 cm long |
HDPE spacers (bottom) | TAP Plastics | N/A | 1 per box, 0.64 cm thick x 2.5 cm wide x 40.5 cm long |
HDPE spacers (patch) | TAP Plastics | N/A | 2 per box, 0.64 cm thick x 3.8 cm wide x 28 cm long |
Polyester batting | Fairfield | #A-X90 | 2.5 cm x 40.5 cm strip per box |
20-thread screws | N/A | N/A | 3.2 cm long, 0.64 cm diameter |
Washers | N/A | N/A | 0.64 cm internal diameter |
Hex nuts | N/A | N/A | sized to fit the screws |
Light deprivation fabric | Americover, Inc. | Bold 8WB26.5 | 1 piece 95 cm wide and 69 cm long per box |
Sand | Quikrete | No. 1113 | |
Field soil | N/A | N/A | |
Transparencies for tracing | FXN | FXNT1319100S | One per side of the box to be traced |