The nitrogen fixation process in peanut (Arachis hypogaea L.) is essential for providing nitrogen (N) for the crop throughout the season, as it delivers from 50-80% of the total N required by the plant (Boddey et al., 1990). The rates of N fixation for peanut range from 100 to 190 kg N/ha (Boddey et al., 1990), and are quite comparable to other N-fixing crops such as soybean [Glycine max (L.) Merr.; 85 – 155 kg N/ha], chickpea [Cicer arietinum L.; 103 kg N/ha], or pigeon pea [Cajanus cajan (L.) Mills.; 168 – 280 kg N/ha) (Elkan, 1995). The process of biological N fixation (BNF) involves symbiotic bacteria that reduce atmospheric N₂ to ammonia, a form of N that is available for plant nutrition (Nievas et al., 2012). The BNF process can account for as much as 65% of the N that is used in agriculture worldwide, with the largest quantity of fixed N originating from the symbiotic association between rhizobia and legumes (Dupont et al., 2012). For example, intercropped peanut has the capacity to provide over 10% of the total N that is accumulated in the accompanying crop during a growing season (Chu et al., 2004). Therefore, assessing the process of nodulation in peanut could have important implications for evaluating and developing sustainable cropping systems globally.

The process of infection of roots by N fixing bacteria is fairly unique in peanut and occurs through a process known as “crack entry” (Boogerd et al., 1997). The process involves the invasion of rhizobia between the epidermal cells at the base of emerging lateral roots, and the eventual proliferation of bacteria into the root cortical cells (Bogino et al., 2011; Nievas et al., 2012). The majority of nodules in peanut are formed on the first order lateral roots near the basal region (top 8 cm) of the taproot (Tajima et al., 2006). Peanut generally lacks root hairs on the taproot because the surface cell layers are shed, so nodules form at the base of lateral roots where rosette-type root hairs up to 4 mm in length are located (Akasaka et al., 1998). Peanut typically forms determinate type nodules (Akasaka et al., 1998) that are characterized by having no permanent meristem, a regular spherical shape, and a homogeneous central tissue within the nodule that is composed of tightly packed bacteroids responsible for the actual N fixation process (Desbrosses and Stougaard, 2011; Dupont et al., 2012). Alternatively, legumes such as pea (Pisum sativum L.) or clover (Trifolium spp.) form indeterminate nodules that have a permanent meristem and take on a more cylindrical shape (Dupont et al., 2012). The two primary bacterial genera that commonly colonize peanut nodules are Bradyrhizobium as well as Rhizobium (Nievas et al., 2012; Angelini et al., 2011). However, rhizobia within single peanut nodules can be very diverse with no real dominant strain (Angelini et al., 2011) and often include non-rhizobial endophytes (Ibáñez et al., 2009).

Aside from biological factors such as rhizobial strain or host specificity among crop cultivars, environmental conditions and management practices within typical field production settings can impact the formation of nodules and the subsequent process of N fixation. These factors include: soil temperature; water availability; soil acidity; soil salinity; intercropping; crop rotation; fertilization regime; inoculation practices; and tillage practices (Sprent, 1972; Hungria and Vargas, 2000; van Kessel and Hartley, 2000; El-Akhal et al., 2013). Reduced tillage and cover crops prior to a legume crop often increase nodulation, as is the case for soybean nodulation (Hughes and Herridge, 1989) and rates of N fixation (Kihara et al., 2011) for this crop. Increased N fixation has also been noted under reduced tillage for chickpea and faba bean (Vicia faba L.; López-Bellido et al., 2011). One typical system where tillage plays a critical role and impacts the performance of the subsequent peanut crop is bahiagrass (Paspalum notatum Flueggé) sod systems. Bahiagrass sod is an important economic component in many Florida peanut production systems either through inclusion of bahiagrass and livestock in rotation with peanut (Katsvairo et al., 2006) or because many producers may rotate long-term bahiagrass fields with peanut to economically benefit their production system. Tillage systems that can be utilized to break up the bahiagrass sod may be limited due to its extensive and fibrous root system. Conventional, deep tillage is primarily chosen, but some growers are considering the economic and environmental benefits of reduced tillage systems (such as strip tillage) that are able to reduce fuel use and protect soil resources. Therefore, the tillage operation chosen to remove the bahiagrass sod could have serious impacts on the nodulation pattern and N fixation of the following peanut crop.

While it is acutely important to assess the impact of tillage and other field management decisions on nodulation and N fixation rates, evaluation and quantification of nodulation patterns in the field is difficult and partially explains the paucity of studies that have successfully achieved these measurements. Methods typically involve the destructive harvest of roots through coring, trenching, or uprooting large quantities of soil – all techniques that are extremely labor intensive, hugely destructive within a research plot area, and most often preclude the ability to perform repeated sampling within a growing season (Gray et al., 2013). Recently however, Gray et al. (2013) successfully demonstrated the efficacy of using the minirhizotron technique for the study of nodulation patterns in field-grown soybean under elevated CO₂ combined with drought stress. The minirhizotron technique involves the installation of clear plastic tubes within and parallel to the crop row, usually at a 30 or 40° angle which allows for repeated imaging of the root system through the season (Milchunas, 2012). A camera is inserted within the tube and roots with associated nodules can be imaged along the length of the tube. Images can then be digitally analyzed for nodule number, size, and density as well as other root traits through the use of several software programs. Because the tube normally has a locking mechanism interfaced with the camera, the user can repeatedly image the exact same locations along the length of the tube over time.

The minirhizotron technique may allow for the detailed and comprehensive examination of the natural history of peanut nodule formation by soil depth and across time that has not been previously documented. In addition, the effect of alternative tillage systems on nodulation, particularly those utilized in a bahiagrass sod rotation can also be examined. To investigate the suitability of the minirhizotron imaging technique in determining differences in peanut nodulation, a field experiment was established in Citra, FL using alternative tillage operations in a bahiagrass sod cropping system as a test case. In situ nodulation patterns along the root system over the growing season were imaged directly through the use of minirhizotrons.

Nodule formation was successfully imaged using the minirhizotron technique, allowing for the sequential measurement of the number and size of nodules through the growing season. It was possible to return to the same imaging frame at each sampling date which allowed the phenological study of individual nodules over time. An interesting observation was noted: many nodules were surrounded by air spaces which may be important for adequate gas exchange that must occur during the N fixation process (Figure 1).

Utilizing the minirhizotron system in the bahiagrass sod test case, the pattern of nodulation within the different tillage systems could be assessed with the technique. In this test case, the individual tillage operations had no impact on the total number of nodules at all soil zone depths and image dates (Table 2). However, the date when the tillage operations were performed (approximately 30 days prior to planting – Date 1, or at plant – Date 2) did influence the number of nodules produced, with Date 2 having more nodules than the earlier tillage date. There were no differences in nodule number among the four imaging dates. This indicates that the number of nodules formed was fixed relatively early in crop development, at least by the second image data (9 July, 61 DAP), and stayed constant through the season with few new nodules formed or older nodules deteriorated. This is illustrated in Figures 2 and 3 where groups of nodules became apparent by 9 July (61 DAP) and persisted until the last image date on 17 August (100 DAP). This indicates that nodules are fully formed by the crop’s peak pod fill (60 DAP) and persist through peak physiological performance (90 DAP). There was one instance where nodules were noted to have senesced and disappeared (Figure 4), but based on the analysis results, this was a fairly rare occurrence. Nodule number differed among soil depth zones (Table 2), with the majority of the nodules formed within soil depth zones 1 and 2 (0-20 cm in depth, Figure 5). The deepest single nodule in the study was noted at 90 cm below the soil surface in the Strip/HRC treatment and persisted through to the 17 August imaging date (Figure 6).

To illustrate general nodule density and the pattern of density development by depth across the primary root area examined, nodule density was calculated as the number of nodules within the reading frame area (13 X 18 mm = 234 mm² area) and graphed along the rooting depth imaged (roughly 90 cm below the soil surface). Figure 7 illustrates the general pattern of nodule density by soil depth across all tillage and tillage date treatments for three imaging dates. This illustrates how nodules were clustered within the top 20 cm of soil and that the nodule density became fairly constant by 9 July. Nodule density below 40 cm was only noted for 9 July and 17 August, likely reflecting the overall pattern of root extension to deeper soil depths by 9 July.

Nodule size expressed as surface area differed among tillage treatments with Strip and Strip/HRC both having larger nodules than Conventional tillage (Table 2). This was the case regardless of when the tillage operation was performed (Date 1 or 2) resulting in average nodule surface areas of 1.9, 2.7, and 2.5 mm² for Conventional, Strip, and Strip/HRC, respectively (Figure 8). There were differences among the image dates with the maximum nodule surface area being reached on 1 August; the surface area values at this date were significantly different than the first image date on 12 June. Visually, the expansion of nodule size is illustrated by images in Figure 9 where the enlargement of a single nodule is shown. Nodule surface areas were different among soil depth zones with larger nodules congregated in the top 20 cm (Figure 10). There was an interaction between tillage and depth zone (Table 2) driven by the trend of larger nodules in the Strip/HRC in zone 2 more so than the other two tillage treatments (Figure 10).

There are several important contributions from this study: 1) the general observational and quantitative assessment of the natural history and progression of peanut nodule formation, both temporally and spatially; 2) the evaluation of the minirhizotron system in a test case examining the impact of tillage on this natural history; and 3) the demonstration of the distinctive and expansive research applications that the minirhizotron technology can afford when studying belowground processes. In general for most crop species, the pattern of root nodule formation in relation to different soil depth zones has rarely been reported (Tajima et al., 2006). The minirhizotron system is perfectly suited for this type of study, but this technique has not been commonly utilized to date for studying nodulation patterns (but see Gray et al., 2013), and the current study is certainly the first for peanut.

Based on the current results, the nodulation pattern over time indicated that nodules were formed relatively early in the growing season and the number was set by 9 July (61 DAP) corresponding to peak pod fill (Boote, 1982). New nodule formation or nodule senescence past that date was relatively rare. This pattern may be due to the suppressive ability of early-formed nodules on subsequent nodule development (Kossiak and Bohlool, 1984; Danso and Bowen, 1989). This timing of nodule production for peanut is similar to studies on other species noting that primary nodulation is attained by crop flowering (Voisin et al., 2003). As far as nodule pattern within the root system, the current study showed most nodules were formed within the top 20 cm below the soil surface, a pattern that is also similar to other studies (Danso and Bowen, 1989; Hardarson et al., 1989; Tajima et al., 2006). However, the current study did note an increase in nodule surface area over time indicating that, while nodule number may have been set early in crop development, nodule expansion was certainly possible through the season.

The variability among different soil depth zones in the nodule density and size found in this study has important implications regarding the rhizobial population that was likely infecting the different root segments. Native rhizobial populations are normally widely distributed throughout the soil profile (Bogino et al., 2011). This is unlike the patterns of distribution for rhizobial strains used in commercial inoculants applied in-furrow or directly on the seed that normally remain near the soil surface (0-10 cm); this is because movement of these rhizobial strains through the soil profile is very slow (Hardarson et al., 1989; McDermott and Graham, 1989; Bogino et al., 2011). Native rhizobial populations are also highly competitive against commercial strains and will normally interfere with the effectiveness of infection by these applied inoculants (Bogino et al., 2011). In the current study, the zones where nodule density and size were high occurred in two zones, 0-10 cm and 10-20 cm. This indicates that perhaps the applied inoculant in the 0-10 cm and native rhizobial strains in the 10-20 cm and deeper soil zones were both active and effective. This is strictly speculative, but testing to confirm and quantify the contribution of native rhizobia seems important in evaluating the effectiveness of commercially applied inoculant. The soil depth zone position of nodules is also important because it is known to directly impact N fixation activity (Hardarson et al., 1989). Roots near the surface derive a higher proportion of their N from the symbiosis with rhizobia (Lory et al., 1992) while deeper nodules have relatively low levels of N-fixation (McDermott and Graham, 1989; Tajima et al., 2006), sometimes due to inhibitory levels of incorporated N fertilizer. However, conservation tillage systems that provide limited soil mixing may have lower nutrient levels at deeper depths which may actually help increase the N-fixing activity of nodules within these regions because the nutrient concentrations would rarely reach inhibitory levels (Virginia et al., 1986).

Nodule size is known to have a direct impact on N fixation (Voisin et al., 2003; Tajima et al., 2007) particularly under stress conditions (King and Purcell, 2001). Large nodules may have amplified N fixation activity in comparison to small nodules due to a greater phloem supply (Purcell et al., 1997). It has been estimated for soybean that, by volume, 25% of a 2 mm diameter nodule is infected, while a 4 mm diameter nodule has 60% infected tissue indicating greater relative N fixation ability (King and Purcell, 2001). In general, peanut has smaller nodules than soybean, but higher overall N fixing activity (Tajima et al., 2007). This size range was similar to that found in the current study where most peanut nodules were approximately 2 mm² in surface area, translating to a 1.4 mm diameter.

This case study was successful in evaluating the efficacy of the minirhizotron system for evaluating the impact of tillage practices on the natural history of nodulation including responses in nodule number and size. Previous studies have found a promotion of nodulation under no-tillage systems and that there was a close positive relationship between nodule number and the N-content of roots within conservation tillage systems (Hughes and Herridge, 1989; Sidiras et al., 1999; Kihara et al., 2011; López-Bellido et al., 2011). Tillage system in the current case study had no effect on nodule number but did significantly influence nodule surface area (a surrogate for overall nodule size). The current results indicate that nodule surface area was enhanced in both conservation tillage treatments (Strip and Strip/HRC), regardless of whether the tillage operation was performed one month prior or at planting. Because of the strong relationship between nodule size and N-fixation rates (Purcell et al., 1997; King and Purcell, 2001; Voisin et al., 2003; Tajima et al., 2007), this result indicates a possible advantage of conservation tillage systems through enhancement of N-fixation rates as a whole. However, because the data was relatively limited in scope and primarily aimed at evaluating the use of the minirhizotron technique in different tillage systems, these results require further study and confirmation.

Another contribution from this study is the illustration of the application of the minirhizotron technique in a myriad of possible studies focused on belowground processes. Aside from root measurements and nodulation patterns, the minirhizotron images captured in this study illustrate additional assessments of belowground community dynamics that could be possible using the technique. Other soil organisms were captured in the images from the current study including arthropod egg masses (Figure 11). Studies concentrating on populations of soil organisms could also be digitally imaged using this technique. In the particular case study, further examination of the interaction between the endophytic fungi of bahiagrass and rhizobia on subsequent colonization in peanut could also be studied using minirhizotrons. The current study and similar applications of the minirhizotron technique show the potential of this method to provide novel and important imaging of phenomena that have previously been unobservable. This was certainly the case for the current study’s documentation of nodulation patterns and processes and opens the door for additional unique investigations.