Where, when, and which root types do respond to local phosphorus supply?

doi:10.21203/rs.3.rs-3231354/v1

Download PDF

Research Article

Where, when, and which root types do respond to local phosphorus supply?

https://doi.org/10.21203/rs.3.rs-3231354/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background and aims To maximize the effectiveness of local phosphorus (P) application, it is necessary to understand under which soil conditions plant growth is improved and where, when, and which root types respond to these conditions.

Methods We developed an experimental system that combines root-box method with AI-based image analysis to monitor spatio-temporal root response. Wheat plants (Triticum aestivum L. cv. Norin 61) were grown in boxes with varying soil P content between patches, while keeping P amount constant among treatments.

Results Significant improvement in root and shoot growth was observed only in treatments with substantial differences in soil P content between patches. Among root types, only lateral roots in P-rich patches increased but decreased in P-poor patch from 4th leaf stage, despite no difference in total roots per plant. If plants couldn’t sufficiently absorb P for growth, lateral roots primarily responded to local soil P content, whereas if plants could, their development would be systematically controlled by the entire plant.

Conclusion We concluded that lateral root distribution in response to local P supply from 4th leaf stage promoted shoot growth, only in cases where substantial differences in P content existed between patches. Seminal and nodal roots likely contribute to spreading root distribution, while lateral roots response plastic to local soil P conditions.

Local fertilizer placement

Rhizobox

Root foraging

Root plasticity

Root proliferation

Triticum aestivum L. (wheat)

Phosphorus (P) is an essential macronutrient for plant growth. P is also a resource that is threatened with depletion (Cordell and Neset 2014). Thus, there is an urgent need to improve P use efficiency. However, P mobility in soils is low and thus crops can absorb only 10–20% of the total P applied to the field (Holford 1997). Thus, synchronizing root-zone P supply with crop demands spatially and temporally at an optimal level in the rhizosphere is important for maximizing the efficiency of the root/rhizosphere in P mobilization and acquisition (Shen et al. 2013). Local fertilizer supply is one of the way to synchronize root-zone nutrient supply with crop demands (Freiling et al. 2022; Nkebiwe et al. 2016; Randall and Hoeft 1988). Under the local fertilizer supply conditions, plants can adjust the placement of their new roots to capture the resources in response to the soil environment, which is known as root plasticity (reviewed in Fitter 1994; Hodge 2004, 2006; Hutching and de Kroon 1994; Robinson 1994; Schneider and Lynch 2020; Yamauchi et al. 1996). A meta-analysis of the published 407 datasets from 37 studies on different local P placement methods revealed that wheat grain yields was higher with band in 76% of the studies, but not in 24% (Freiling et al. 2022). Whether root plasticity is adaptive or maladaptive depends strongly on the soil conditions where the plant grows (Schneider and Lynch 2020). Thus, to maximize the effect of local P supply method, it is necessary to understand the local P supply conditions, under which root plasticity is activated, and the spatio-temporal response of roots to these conditions.

Only location, size, and concentration can be controlled as fertilizer application methods. As regards location, both of the two meta-analyses of the published studies on diverse local fertilizer placement techniques have demonstrated that P application in the deep band (>10 cm) leads to the most yield advantages when compared to the other techniques (e.g. broadcast, seed placement, knife, and shallow band) (Freiling et al. 2022; Nkebiwe et al. 2016). However, there has been limited examination of patch size and concentration, despite a wealth of studies on the plant root response to nutrient patches (Hodge 2006). Hodge (2006) hypothesized that if a patch is highly concentrated with nutrients but very small, it may not be detected by the root system. In fact, optimal plant growth has been observed when patch size is half or a quarter of the container, with growth decreasing as patch size diminishes (Kume et al. 2006；Wijesinghe and Hutchings 1997; Yano and Kume 2005). On the other hand, Hodge (2006) argues that nutrient concentration is likely more significant than size, as a large patch with low concentration may not stimulate a plastic root response effectively. However, there are even fewer studies examining concentration than size. Sun et al. (2002) evaluated the effects of localized P supply at different concentrations on root segments and found that the number and density of 2nd-order lateral roots increased up to the appropriate concentration but then sharply decreased. It should be noted that the total amount of supplied P was different among the treatments in this experiment. Gersani and Sachs (1992) demonstrated evidence of tight developmental co-ordination amongst roots. As the solution suppling one half of the root system became more concentrated, progressively more lateral primordia were visible. Whereas, fewer primordia were visible in the other half of the root system supplied with a uniformly nutrient-rich solution. This suggested that the initiation of primordia is dependent not only on the local conditions but also on the nutrient conditions of the counterpart of root system. However, there have been no studies assessing the impact of changes in nutrient concentrations in both high and low patches on plants. For efficient fertilizer use in agricultural practices, it is essential to vary only the distribution of fertilizers while keeping the total applied amount and to study how plant growth responds to these distribution patterns.

The root system consists of different root types, each under distinct genetic control, that are different in emergence, morphology, anatomy, and function (Hodge et al. 2009; Yamauchi et al. 1996). The wheat root system is composed of seminal roots (the radicle and three to five basal roots) and shoot-borne nodal roots (Freschet et al. 2021; Robertson et al. 1979; Tanaka 1974; Yamauchi et al. 1996) as well as lateral roots on these axial roots. Under heterogeneous soil conditions, it is highly probable that each of these root types within the root system will exhibit distinct responses. Schnerder and Lynch (2020) discussed to better understand and interpret plasticity, we need a comprehensive understanding of the utility of individual root phenes. Hodge (2004) argued that though root proliferation has been used to describe various morphological responses of roots to nutrient patches, lateral root initiation is more likely to do so locally. On the other hand, Schnerder and Lynch (2020) speculated that the nodal roots originating from tillers, and its plastic response would be important for edaphic stress tolerance in monocot species. Fitter (1994) discussed that root systems will maximize acquisition of resources in a patch environment by preferential allocation of root growth to rich patches, which involves two stages: locating and exploiting. The former will be achieved by fast-growing, long-lived root systems and the latter will require generation of a high root length density in a short time. According to this hypothesis, thicker seminal and nodal roots will be involved in locating, while thinner lateral roots will be involved in exploiting. However, since there are few studies of long-term cultivation under local fertilizer conditions, little is known when and where each root type responds to local fertilizer application and contributes to the overall plasticity of the root system.

A few previous studies have examined the spatio-temporal response of root system to heterogeneous nutrient conditions by destructive sampling methods. For example, Passioura and Wetselaar (1972) showed that the process of nitrification and the spatio-temporal response of root system architecture and suggested that ammonium seemed to be the main stimulus for wheat roots. Drew and Saker (1978) demonstrated that the higher accumulation of root dry matter and P uptake rates in the P-enriched zone, even before lateral roots emerged. Non-destructive investigations can detect more detailed changes of root plastic response (Robinson 1994). Recently, X-ray computed tomography (CT) has been used for non-destructively imaging and quantifying the spatio-temporal response of roots to local fertilizer (Ahmed et al. 2016; Blaser et al. 2020; Flavel et al. 2012, 2014; Gao et al. 2019; Hou et al. 2022). Gao et al. (2019) succeeded to precisely observe and quantify the spatio-temporal response of roots to the localized P supply. However, there is an inevitable trade-off between image resolution and pot size at a fixed size of detector (Hou et al. 2022): the better the image resolution, the smaller the pot diameter. Thus, the pot size used in the study of Gao et al. (2019) was too small (just 25 mm in height and 70 mm in diameter) to cultivate the plants more than 21 days and they could not observe an increased P uptake and plant growth in the localized P supply treatment as compared to the uniform P treatment. Further, due to the trade-off between image resolution and pot size, it is possible that the thin roots such as lateral roots would not be detected. Thus, there is a need for experimental systems that can assess the response of each root type, including lateral roots, to local P supply over time in larger plants.

Root box, also known as a rhizobox, commonly made of transparent polyvinyl chloride (PVC) or glass plate has been widely used to observe roots of relatively large plants (Araki et al. 2000, 2001; Böhm 1979; Kono et al. 1987; Koyama et al. 2021; Neumann et al. 2009; Shuurman and Goedewaagen 1971; Yamauchi et al. 1987). Visible root growth on the outer surface of the root box was recorded by scanning or taking pictures of the root box surface or manually drawing the visible roots on transparent sheets (Böhm 1979). Using a special root box with a nylon mesh placed between the substrate and the transparent face of the box, Boukcim et al. (2006) succeeded to show the dynamic response of a tap root and lateral roots, separately, to local nitrate or ammonium supply by tracing the roots manually. However, even after the development of root tracing software for digital images around the 1990s, the tracing was still very labor intensive and time consuming. In recent years, advances in image analysis technology and artificial intelligence (AI) have led to the development of software that can segment roots from the soil, enabling fully automated analysis of roots in the two-dimensional soil images (Smith et al. 2022).

In this study, we established an experimental system that integrates the root-box method with AI-based image analysis to monitor the spatio-temporal response of roots to local P supply. Using this system, the objectives are to assess: how much sampled root length can be estimated by the segmented root images, under which local P supply conditions plants growth is improved, and where, when, and which root types respond to local soil P conditions.

Growth conditions

A 7 cm cylindrical partition wall was placed in the center of the transparent root boxes (L 250 mm × W 400 mm × H 30 mm). Within this cylinder, 40 g of dry volcanic ash soil (Shibanome-tsuchi, Shidara Co. Ltd., Tochigi, Japan) without fertilizer was filled. The region around the cylinder was divided into four equal patches using partition walls. Each patch was filled with 320 g of dry soil that was applied with the same amount of ammonium sulfate (1.26 g N kg^-1 dry soil) and potassium chloride (1.49 g K kg^-1 dry soil) and a following different amount of superphosphate. Total amount of superphosphate (0.26 g P) was applied to each root box, distributed among the four patches according to the following ratios: 25:25:25:25 (control), 0:33:33:33 (P0), 50:17:17:17 (P50), 75:8:8:8 (P75), and 100:0:0:0 (P100) as shown in Fig. 1. After filling the soil in the root boxes, the partition walls were removed. Hereafter, the patch that was filled with the soil applied with the higher or lower amount of P was referred to as P-rich or poor patch, respectively (Fig. 1). The root boxes were arranged in a randomized complete block design with four replicates.

Seeds of wheat (Triticum aestivum L. cv. Norin 61) were allowed to absorb water on a petri plate covered with enough water for 1 day. The soil was irrigated to the field capacity before sowing. In general, wheat develops seminal roots (the radicle at germination, followed by the emergence of three to five basal roots) (Freschet et al. 2021; Robertson et al. 1979; Tanaka 1974; Yamauchi et al. 1996). In order to make these roots to access equally to each of all four patches, one germinated seed was sown in the center of the root box (Figs. 1 and S1). Each plant was grown for 39 days in a greenhouse at 25°C. The root boxes were irrigated every day to maintain soil moisture content of 40–50% (w/w).

Non-destructive imaging and quantification of each root type

The bottom surface of root box was scanned with a A3 (L 297 mm × W 420 mm) scanner (24-bit Color mode, 600 dpi; DS-G20000, Seiko Epson Corp., Nagano, Japan) from day 2 to day 39 after seeding (Fig. S1). To increase the number of roots emerging at the bottom of the root box, the soil layer was thinned to 30 mm. A total of 760 TIFF images (20 root boxes × 38 days) were acquired. The pixels of these images were H 6838 and W 4758. Root segmentation from the images was done with the software RootPainter (Smith et al. 2020). First, a training dataset of 500 images (H 600 pixels; W 600 pixels) was created by cropping the scanned images using ImageJ software (Rasband 1997). From this dataset, 200 images were randomly selected and partially annotated, with a focus on annotating root and background edges, using the software's built-in neural networks for segmentation model training. The resulting model was then applied to all the 760 images to perform batch segmentation, resulting in segmented PNG images. Since image analysis system of WinRHIZO (Win-RHIZO Pro 2017, Regent Instruments, Inc., Quebec City, QC, Canada) cannot read PNG images, the segmented PNG images were converted to 8-bit grayscale TIFF images using ImageJ. When analyzing the response of roots to local soil P conditions, the region of the cylinder was excluded from the TIFF images and the remaining images were divided into four patches using ImageJ. These images were analyzed using the WinRHIZO. In this study, the cm of root length visible per cm² of the bottom surface of root box is termed rooting intensity (Böhm 1979).

Further, to analyze the response of different root types to local soil P conditions, both seminal roots and nodal roots were manually traced in the segmented root images using image editing software (MediBang Paint Pro, MediBang Inc., Japan). The traced images were subsequently analyzed using WinRHIZO software. Lateral root length was determined by subtracting the combined length of seminal and nodal roots from the total root length.

Root length, shoot dry weight, shoot P concentration, and soil available P

At 39 days after sowing (DAS), roots were sampled in each patch. The roots were spread out on a Plexiglas tray filled with enough tap water to cover the surface of the roots and scanned with the A3 scanner (400 dpi; DS-G20000, Seiko Epson Corp., Nagano, Japan) equipped with the scanner transparency unit (ESA3FLU3, Seiko Epson Corp., Nagano, Japan). The images were saved in TIFF format. Then, the root length in each patch was analyzed using the WinRHIZO software. The shoots were dried at 80 °C for 72 h, weighed, ground with a mill, and digested with a mixture of nitric acid/perchloric acid. P concentration in the digests was determined with the molybdenum blue method. P uptake was calculated by multiplying the measured P concentration with the shoot dry weight. Soil available P in each patch after local P supply treatments was extracted according to Truog (1930) and measured colorimetrically. Weighted rooting intensity was calculated using the following equation.

Weighted rooting intensity

Statistical analysis

All statistical analyses were performed by statistical software (JMP v. 12.2, SAS Institute Inc., NC, USA). The data were analyzed by one-way analysis of variance (ANOVA) at significant levels of P < 0.001, 0.01, and 0.05 and by Tukey’s HSD test at P < 0.05. Simple linear and quadratic regression models were used to analyze correlations among parameters at significant levels of P < 0.001, 0.01, and 0.05.

Comparison of rooting intensity and root length

Before analyzing spatio-temporal response of roots to local soil P conditions, we verified how much sampled root length can be estimated by rooting intensity calculated from the segmented root images (Fig. 2). The linear regression of rooting intensity and root length showed significantly positive correlation (r = 0.99, P < 0.001), allowing the estimation of root distribution from the segmented root images. The slope of the regression line indicated that the rooting intensity amounted to 30% of root length determined by the sampled roots.

Shoot growth and P uptake

We investigated under which local P supply conditions P uptake and plant growth were improved. Shoot dry weight increased with increasing difference in soil available P content between the P-rich and poor patches, and P50, P75, and P100 was significantly higher than the control. There was no significant difference between P50, P75, and P100 (Fig. 3a). Similar trends were observed in P uptake by plant, which increased with the difference in soil available P content. P50, P75, and P100 showed significantly higher P uptake than the control. There was no significant difference between P75 and P100 (Fig. 3b).

Spatio-temporal response of each root type to local soil P conditions

The segmented root images showed the spatio-temporal response of each root type to local P supply (Fig. 4). At 27 DAS around 5th leaf stage (Fig. S2), P0 and P100 showed greater root growth in the P-rich patch than the P-poor patch. This root response was more pronounced at 39 DAS around 9th leaf stage. Among root types, seminal and nodal roots elongated randomly towards each patch, whereas lateral roots clearly responded to local soil P conditions (Fig. 4). Specifically, only the seminal roots that elongated in the P-rich patch of P0 and P100 branched obviously more lateral roots than in the P-poor patch, but all the seminal roots of control developed lateral roots similarly. These responses are more clearly shown in Video S1.

From these images, we assessed where, when, and which root types respond to local soil P conditions. Fig. 5a-c showed total rooting intensity including all root types. Interestingly, there was no significant difference in total rooting intensity per plant among the local P supply treatments, but the differences among the treatments became distinct when analyzed the total rooting intensity separately in the P-rich and poor patches. The total rooting intensity in the P-rich patch increased by the local P supply treatments, whereas that in the P-poor patch decreased. ANOVA revealed significant effects of the local P supply treatments on total rooting intensity in the P-rich patch from 16 DAS around 4th leaf stage onwards, except for 17 DAS (Figs. 5b and S2). Total rooting intensities in P50, P75, and P100 were higher than the control, but there was no significant difference between P0 and the control (Table S1). In contrast, total rooting intensity of the P-poor patch in P0, P50, P75, and P100 tended to be lower than the control from about 16 DAS. ANOVA detected significant effects of the local P supply treatments on total rooting intensity in the P-poor patch at 8 DAS and from 16 DAS onwards (Fig. 5b).

Fig. 5d-l demonstrated the effect of local P supply treatments on different root types. The response of seminal and nodal roots differed from that including all root types, whereas lateral roots exhibited a similar response. The rooting intensities of seminal roots were significantly different among the treatments during the early growth stage, but no differences were observed when sampled at 39 DAS (Fig. 5d-f). Nodal roots did not respond directly to local soil P conditions, but their response to local P supply treatments was similar to shoot dry weight and P uptake (Figs. 2 and 5g-i). After around 22 DAS around 5th leaf stage, ANOVA detected a significant effect of the local P supply treatments on the rooting intensity of nodal roots (Figs. 5g-i and S2). The rooting intensity of nodal roots of P75 and P100 increased significantly compared to the control, as did shoot dry weight and P uptake. Only lateral roots exhibited the similar response to the total rooting intensity (Fig. 5a, b, c, j, k, l). The rooting intensity including all root types and that of lateral roots showed a significant difference from around 16 DAS around 4th leaf stage.

Fig. 6 showed the relationship between soil available P and the rooting intensity of each root type. Significantly quadratic correlations were only observed for the rooting intensity including all root types (y = -0.0003x² + 0.068x - 0.330, R² = 0.857, n = 34, P < 0.001) and that of lateral roots (y = -0.0003x² + 0.071x - 0.753, R² = 0.884, n = 34, P < 0.001) (Fig. 6a, d). However, no significant correlations were observed between soil available P and the rooting intensities of seminal roots or nodal roots, respectively (Fig. 6b, c).

Contribution of efficient root distribution in response to soil available P to shoot growth

Weighted rooting intensity, the rooting intensity multiplied by the soil available P content of the patch in which the roots grew, was calculated to determine whether the efficient root distribution in response to soil available P improved shoot growth. The linear regression of shoot dry weight and weighted rooting intensity showed close positive correlation (r = 0.914, P < 0.001) (Fig. 7b). Though there was also a significant positive correlation between shoot dry weight and rooting intensity (r = 0.620, P < 0.01) (Fig. 7a), the correlation coefficient was higher for weighted rooting intensity than rooting intensity. Thus, efficient root distribution in response to soil available P improved shoot growth under local P supply conditions.

We investigated how much sampled root length can be estimated by the segmented root images, under which local P supply conditions plants growth is improved, and where, when, and which root types respond to local soil P conditions. The slope of the regression line indicated that 30% of the sampled root length could be estimated accurately using the rooting intensity determined by the segmented root images (Fig. 2). Shoot dry weight increased as the difference in soil available P content between the P-rich and poor patches increased, and P50, P75, and P100 were significantly higher than the control (Fig. 3a). P uptake showed a similar trend to shoot dry weight (Fig. 3b). Although there was no significant difference in total rooting intensity per plant among the local P supply treatments during the experiment, we observed a significant effect of the treatments on both the total root intensities in the P-rich and poor patches from around 16 DAS around 4th leaf stage (Figs. 5a-c and S2). The total rooting intensity in the P-rich patch increased with the local P supply treatments, whereas that in the P-poor patch decreased. Among the root types, only lateral roots responded to local P supply treatments and soil available P content as same as total root (Figs. 5 and 6). Additionally, we found a significantly positive correlation between shoot dry weight and the root length weighted by soil available P (r = 0.914, P < 0.001). From these results, we concluded that lateral root distribution in response to soil available P from the 4th leaf stage promoted shoot growth, only when there was a substantial difference in soil available P between the P-rich and poor patches due to local P supply treatments.

Schnerder and Lynch (2020) discussed that the determination of whether plasticity is adaptive or maladaptive depends strongly on its temporal expression. Plants grow exponentially, so that initial growth differences increase as growth progresses. Thus, it is necessary to understand the timing of the onset of plastic root response. However, there were no non-destructive methods for measuring roots in soil, making it difficult to determine dynamic root response to localized nutrient application. Gao et al. (2019) for the first time succeeded to report the detail spatio-temporal response of roots to localized P application in soil using X-ray CT, but no differences in nutrients uptake and shoot growth were shown between the localized and uniform P supply treatments. That is because of the pot size constraints to maintain a certain degree of resolution of X-ray CT. On the other hand, by using recent image analysis techniques (Smith et al. 2022), we have succeeded in presenting detailed spatio-temporal responses of each root type, as well as nutrient uptake and growth responses, by a simple experimental system that does not require expensive X-ray CT equipment.

There have been limited reports on the timing of the onset of plastic root response to local nutrient applications (Robinson 1994). In the case of local P application, the relative growth rate (RGR) of root length increased four-fold by local P application within 1 d, compared with the RGR of roots treated with water (Jackson and Caldwell 1989). Drew and Saker (1978) reported a significant increase in local root dry matter within at most five days. However, in our study, a significant difference in rooting intensity due to local P supply treatments was observed from around 16 DAS (Table S1). Roots were already present in the patches before nutrient enrichment in the studies conducted by Jackson and Caldwell (1989) and Drew and Saker (1978), whereas in our study, the roots were not present initially. These differences in experimental setup makes the difference in the timing of the onset of plastic root response to local nutrient application. The distance from the seeds to the nutrient patch in this study was 3.5 cm, which is similar to the distance reported by Gao et al. (2019) as 4 cm. In both this study and Gao et al. (2019), a significant increase in root length due to local P application was observed from around 15 DAS. These findings suggest that root foraging for local nutrient application is influenced by the ability to search for the patch and the ability to respond to that specific patch. Recently, P uptake and growth in rice plants were enhanced by combining genetic traits of RSA and localized P application, i.e. a line that has shallower root growth angle and P-dipping (placing P near the roots at transplanting) (Oo et al. 2021). This is a good practice example of reducing carbon costs for the ability to search for the patch. In contrast, the ability to respond to that specific patch is a genetic trait that cannot be compensated for by local fertilizer application methods. However, studies investigating this specific trait are still lacking, highlighting the need for further research in this area.

We presented the spatio-temporal responses of seminal, nodal, and lateral roots for the first time and found that among them, only lateral roots responded dynamically to local P supply treatments (Figs. 4 and 5). In contrast, Schnerder and Lynch (2020) suggested that the plastic response of nodal roots could be important for edaphic stress tolerance in monocot species. Actually, Kume et al. (2006) demonstrated that maize plants in pots preferentially developed nodal roots into P-enriched patch, though this was only observed in plants cultivated in pots with large patch size. However, our data did not support this finding, as nodal roots did not response to local P supply treatments (Fig. 6c). Instead, nodal roots increased as growth improved. These differences in nodal root responses may be attributed to variations in plant species or growth conditions, particularly patch size. Regardless of the specific responses of nodal roots, our study clearly demonstrated that lateral roots were the primary responders to local P supply treatments. This finding aligns with the concept discussed by Fitter (1994), who proposed that locating the patches is achieved by fast-growing, long-lived root systems, while exploiting the patches requires the generation of a high root length density within a short period of time to maximize acquisition of resources in a patch environment. Our data of spatio-temporal response of each root type (Figs. 4 and 5) strongly indicated that seminal and nodal roots are assumed to play a role in spreading the root distribution and searching the nutrient rich patches, while lateral roots are responsible for generating a high root length density within those patches.

Plants allocate their resources efficiently responding to soil nutrient conditions. In this study, the rooting intensity in the P-rich patch increased by the local P supply treatments, while that in the P-poor patch decreased, despite no difference in total rooting intensity per plant between the local P supply treatments (Fig. 5). It indicates that growth differences are not caused by changing the total amount of roots, but by changing their distribution. How do plants control this distribution? This study demonstrated the significant quadratic relationship between soil available P and total or lateral rooting intensity (Fig. 6a, d). Interestingly, the data from both the P-poor and rich patches aligned neatly on the same regression equation. It suggests that the plastic root response is primarily influenced by soil available P. The optimal soil available P was 113.3 mg kg^-1 for the total rooting intensity and 118.3 mg kg^-1 for the lateral rooting intensity. Beyond these thresholds, both the total and lateral rooting intensity tended to decrease. The treatment containing the soil available P of 103.8 mg kg^-1 was P75 and that containing the soil available P of 141.5 mg kg^-1 was P100. The shoot dry weight of P100 tended to be smaller than that of P75 (Fig. 3a), consistent with the observed decrease in rooting intensity. These results indicate that lateral roots primarily respond to local soil available P until plants can absorb sufficient P for growth. Once their P requirements are met, the development of lateral roots would be systematically controlled by the plant as a whole.

Hodge (2006) hypothesized that if a patch is highly concentrated with nutrients but very small, it may not be detected by roots. Actually, the growth-promoting effect of local fertilizer application cannot be observed unless there is a certain patch size (Kume et al. 2006; Wijesinghe and Hutchings 1997; Yano and Kume 2005). Given that growth-promoting effects were observed in this study, it is likely that the patch size was large enough. On the other hand, Yano and Kume (2005) demonstrated that maize plants grown in smaller P-patches, where more roots were distributed outside the patches (excessive proliferation), experienced a biomass depression compared to plants grown in larger P-patches. This highlights the need to examine whether the response of wheat plants to local fertilizer application varies with patch size. Additionally, it is important to investigate whether the proliferation of roots incurs a cost if nutrients are lost from the patch during the absorption process by plants. We presented the proliferation process of each component root in each patch (Fig. 5). Using these values, if a model is constructed that takes into account the process of nutrient loss and the process of root flourishing in the patch, it can be possible to understand under which conditions roots become a cost. This should be studied in the future.

Hodge (2006) argues that nutrient concentration is likely more important than the size of the patch, as a large patch of low concentration may be insufficiently fertile to evoke a plastic root response. In our study, significant differences in growth were observed, only when the differences in soil available P between the patches were above a certain level (Fig. 3a). Therefore, it may be possible to increase the likelihood of success of local P supply in field conditions by identifying in advance the soil available P of the target cultivation area and determining the amount of P fertilizer required to create a substantial difference in soil available P. Further research is needed to validate this approach in field experiments.

Our experimental system provides several advantages compared to the use of X-ray CT for monitoring the temporal response of roots to local P supply. In terms of accuracy, a recent study using X-ray CT (Gao et al. 2019) quantified only 26% and 40% of the sampled root length in maize and faba bean, respectively. In contrast, our system allows for the estimation of approximately 30% of the sampled root length from the scanned images (Fig. 2), though the root diameter of wheat was thinner than maize and faba bean. Further, pot diameter of recent studies using X-ray CT has been limited less than 110 mm due to the trade-off with spatial resolution (Hou et al. 2022). Consequently, plants can only be grown for relatively short periods, typically around 25 days (Flavel et al. 2012, 2014; Gao et al. 2019). In contrast, using a root box of the same size as our study, summer crops such as rice and soybean can be grown for longer periods, from 30 to 52 days, and winter crops such as wheat and rye can be grown for up to 130 days (Kono et al. 1987; Koyama et al. 2021; Yamauchi et al. 1987). However, our method does have certain limitations. To eliminate the influence of root age and gravity and maximize the visibility of root response to local nutrient conditions, we cultivated plants in the middle of the root box with a thin layer of soil. Due to the shallow soil layer, the root system is physically restricted by the root box and cannot represent the root system architecture as it would in a natural environment. As a result, it is not feasible to test the effects of root gravitropism and gravity transfer of nutrient water on roots using our system. Therefore, it is essential to effectively combine this method with field trials in the future to propose optimal fertilizer application methods.

Acknowledgements

We would like to thank Koichi Hoshino for his assistance with chemical analysis, Dr. Takayuki Kashiwagi for his seed supply, and Dr. Chie Hayakawa for her pre-submission review. We acknowledge the following funding sources that supported this work: JSPS KAKENHI Grant (No. 19K15822 and 22K05595) from Japan Society for the Promotion of Science.
Funding

This work was partly supported by JSPS KAKENHI Grant (No. 19K15822 and 22K05595) from Japan Society for the Promotion of Science.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Author contributions

Masaaki Hashimoto, Shun Murakami, and Takuya Koyama contributed to the conception and design of the study. Masaaki Hashimoto, Shun Murakami, Hiromitsu Aoki, and Takuya Koyama performed the research and collected the data. Masaaki Hashimoto and Takuya Koyama analyzed and interpreted the data and prepared the manuscript. All authors read and approved the final manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ahmed S, Klassen TN, Keyes S, Daly M, Jones DL, Mavrogordato M, Sinclair I, Roose T (2016) Imaging the interaction of roots and phosphate fertiliser granules using 4D X-ray tomography. Plant Soil 401:125–134. https://doi.org/10.1007/s11104-015-2425-5
Araki H, Hirayama M, Hirasawa H, Iijima M (2000) Which roots penetrate the deepest in rice and maize root systems? Plant Prod Sci 3:281–288. https://doi.org/10.1626/pps.3.281
Araki H, Iijima M (2001) Deep rooting in winter wheat: rooting nodes of deep roots in two cultivars with deep and shallow root systems. Plant Prod Sci 4:215–219. https://doi.org/10.1626/pps.4.215
Blaser SRGA, Koebernick N, Spott O, Thiel E, Vetterlein D (2020) Dynamics of localised nitrogen supply and relevance for root growth of Vicia faba ('Fuego') and Hordeum vulgare ('Marthe') in soil. Sci Rep 10:15776. https://doi.org/10.1038/s41598-020-72140-1
Böhm W (1979) Methods of studying root systems. Springer-Verlag, Berlin
Boukcim H, Pages L, Mousain D (2006) Local NO³⁻ or NH⁴⁺ supply modifies the root system architecture of Cedrus atlantica seedlings grown in a split-root device. J Plant Physiol 163:1293–1304. https://doi.org/10.1016/j.jplph.2005.08.011
Cordell D, Neset T-SS (2014) Phosphorus vulnerability: a qualitative framework for assessing the vulnerability of national and regional food systems to the multi-dimensional stressors of phosphorus scarcity. Glob Environ Chang 24:108–122. https://doi.org/10.1016/j.gloenvcha.2013.11.005
Drew MC, Saker LR (1978) Nutrient supply and the growth of the seminal root system in barley. III. Compensatory increases in growth of lateral roots, and in rates of phosphate uptake in response to a localised supply of phosphate. J Exp Bot 29: 435–451. https://doi.org/10.1093/jxb/29.2.435
Fitter AH (1994) Architecture and biomass allocation as components of the plastic response of root systems to soil heterogeneity. In: Caldwell MM, Pearcy RW (ed) Exploitation of environmental heterogeneity of plants. Academic Press, New York, pp 305–323
Flavel RJ, Guppy CN, Tighe M, Watt M, McNeill A, Young IM (2012) Non-destructive quantification of cereal roots in soil using high-resolution X-ray tomography. J Exp Bot 63:2503–2511. https://doi.org/10.1093/jxb/err421
Flavel RJ, Guppy CN, Tighe MK, Watt M, Young IM (2014) Quantifying the response of wheat (Triticum aestivum L) root system architecture to phosphorus in an Oxisol. Plant Soil 385:303–310. https://doi.org/10.1007/s11104-014-2191-9
Freiling M, von Tucher S, Schmidhalter U (2022) Factors influencing phosphorus placement and effects on yield and yield parameters: A meta-analysis. Soil Till Res 216:105257. https://doi.org/10.1016/j.still.2021.105257
Freschet GT, Pagès L, Iversen C, Comas L, Rewald B, Roumet C, Klimešová J, Zadworny M, Poorter H, Postma JA et al (2021) A starting guide to root ecology: strengthening ecological concepts and standardizing root classification, sampling, processing and trait measurements. New Phytol 232:973–1122. https://doi.org/10.1111/nph.17572
Gao W, Blaser SRGA, Schlüter S, Shen J, Vetterlein D (2019) Effect of localised phosphorus application on root growth and soil nutrient dynamics in situ – comparison of maize ( Zea mays ) and faba bean ( Vicia faba ) at the seedling stage. Plant Soil 441:469–483. https://doi.org/10.1007/s11104-019-04138-2
Gersani M, Sachs T (1992) Developmental correlations between roots in heterogeneous environments. Plant Cell Environ 15:463–469. https://doi.org/10.1111/j.1365-3040.1992.tb00997.x
Hodge A (2004) The plastic plant: Root responses to heterogeneous supplies of nutrients. New Phytol 162:9–24. https://doi.org/10.1111/j.1469-8137.2004.01015.x
Hodge A (2006) Plastic plants and patchy soils. J Exp Bot 57:401–411. https://doi.org/10.1093/jxb/eri280
Hodge A, Berta G, Doussan C, Merchan F, Crespi M (2009) Plant root growth, architecture and function. Plant Soil 321: 153–187. https://doi.org/10.1007/s11104-009-9929-9
Holford ICR (1997) Soil phosphorus: its measurement and its uptake by plants. Aust. J. Soil Res 35: 227–239. https://doi.org/10.1071/S96047
Hou L (Helen), Gao W, van der Bom F, Weng Z (Han), Doolette CL, Maksimenko A, Hausermann D, Zheng Y, Tang C, Lombi E et al (2022) Use of X-ray tomography for examining root architecture in soils. Geoderma 405:115405. https://doi.org/10.1016/j.geoderma.2021.115405
Hutchings MJ, de Kroon H (1994) Foraging in plants: the role of morphological plasticity in resource acquistion. Adv Ecol Res 25:159–238. https://doi.org/10.1016/S0065-2504(08)60215-9
Jackson RB and Caldwell MM (1989) The timing and degree of root proliferation in fertile-soil microsites for three cold-desert perennials. Oecologia 81:149–153. https://doi.org/10.1007/BF00379798
Kono Y, Yamauchi A, Nonoyama T, Tatsumi J, Kawamura N (1987) A revised system of root-soil interaction for laboratory work. Environ Cont Biol 25:141–151. https://doi.org/10.2525/ecb1963.25.141
Koyama T, Murakami S, Karasawa T, Ejiri M, Shiono K (2021) Complete root specimen of plants grown in soil‑filled root box: sampling, measuring, and staining method. Plant Methods 17:1–13. https://doi.org/10.1186/s13007-021-00798-3
Kume T, Sekiya N, Yano K (2006) Heterogeneity in spatial P-distribution and foraging capability by Zea mays: Effects of patch size and barriers to restrict root proliferation within a patch. Ann Bot 98: 1271–1277. https://doi.org/10.1093/aob/mcl216
Neumann G, George TS, Plassard C (2009) Strategies and methods for studying the rhizosphere-the plant science toolbox. Plant Soil 321: 431–456. https://doi.org/10.1007/s11104-009-9953-9
Nkebiwe PM, Weinmann M, Bar-Tal A, Müller T (2016) Fertilizer placement to improve crop nutrient acquisition and yield: A review and meta-analysis. Field Crop Res 196:389–401. https://doi.org/10.1016/j.fcr.2016.07.018
Oo AZ, Tsujimoto Y, Mukai M, Nishigaki T, Takai T, Uga Y (2021) Synergy between a shallow root system with a DRO1 homologue and localized P application improves P uptake of lowland rice. Sci Rep 11:1–10. https://doi.org/10.1038/s41598-021-89129-z
Passioura JB, Wetselaar R (1972) Consequences of banding fertilizers in soil. II. Effects, on the growth of wheat roots. Plant Soil 36:461– 473. https://doi.org/10.1007/BF01373498
Randall GW, Hoeft RG (1988) Placement methods for improved efficiency of P and K fertilisers: a review. J Prod Agric 1:70–79. https://doi.org/10.2134/jpa1988.0070
Rasband WS (1997) ImageJ. US National Institutes of Health, Bethesda, MD https://imagej.nih.gov/ij/
Robertson BM, Waines JG, Gill BS (1979) Genetic variability for seedling root numbers in wild and domesticated wheats 1. Crop Sci 19:843–847. https://doi.org/10.2135/cropsci1979.0011183X001900060024x
Robinson D (1994) The responses of plants to non‐uniform supplies of nutrients. New Phytol 127:635–674. https://doi.org/10.1111/j.1469-8137.1994.tb02969.x
Schneider HM, Lynch JP (2020) Should Root Plasticity Be a Crop Breeding Target? Front Plant Sci 11:1–16. https://doi.org/10.3389/fpls.2020.00546
Shen J, Li C, Mi G, Li L, Yuan L, Jiang R, Zhang F (2013) Maximizing root/rhizosphere efficiency to improve crop productivity and nutrient use efficiency in intensive agriculture of China. J Exp Bot 64:1181–1192. https://doi.org/10.1093/jxb/ers342
Smith AG, Han E, Petersen J, Olsen NAF, Giese C, Athmann M, Dresbøll DB, Thorup-Kristensen K (2022) RootPainter: deep learning segmentation of biological images with corrective annotation. New Phytol 236:774–791. https://doi.org/10.1111/nph.18387
Sun H, Zhang F, Li L, Tang C (2002) The morphological changes of wheat genotypes as affected by the levels of localized phosphate supply. Plant Soil 245:233–238.
Shuurman JJ, Goedewaagen MAJ (1971) Methods for the examination of root systems and roots. Centre for Agricultural Publications and Documentation, Wageningen. https:// edepot. wur. nl/ 218769.
Tanaka N (1974) The study of crop roots. Jpn J. Crop Sci 43:291–316. (in Japanese) https://doi.org/10.1626/jcs.43.291
Truog E (1930) The determination of the readily available phosphorus in soils. J. Am. Soc. Agron 22:874–882.
Wijesinghe DK, Hutchings MJ (1997) The effect of spatial scale of environmental heterogeneity on growth of a clonal plant: an experimental study with Glechoma hederacea. J Ecol 85:17–28. https://doi.org/10.2307/2960624
Yamauchi A, Kono Y, Tatsumi J (1987) Comparison of root system structures of 13 species of cereals. Jpn J. Crop Sci 56:618–31. https://doi.org/10.1626/jcs.56.618
Yamauchi A, Pardales Jr JR, Kono Y (1996) Root system structure and its relation to stress tolerance. In: Ito O, Johansen C, Adu-Gyamfi JJ, KatayamaK, Kumar Rao JVDK, Rego TJ (ed) Roots and nitrogen in cropping systems of the semi-arid tropics. Japan International Research Center for Agricultural Sciences (JIRCAS), Tsukuba, pp 211–234.
Yano K, Kume T (2005) Root morphological plasticity for heterogeneous phosphorus supply in Zea mays L. Plant Prod Sci 8:427–432. https://doi.org/10.1626/pps.8.427

Figs.S1S2.docx
Fig. S1 Non-destructive imaging of the bottom surface of root box with a A3 (L 297 mm × W 420 mm) scanner. Fig. S2 Response of leaf age to local P supply treatments. Values are means ± SE (n = 4). ×, control; △, P0; ◇, P50; □, P75; 〇, P100.
TablesS14.xlsx
Table S1 Statistical analysis for rooting intensity of total roots among treatments. Rooting intensity per plant (a), in P-rich patch (b), and in P-poor patch (c). The different letters indicate significant differences among the treatments (Tukey HSD test, P < 0.05). Table S2 Statistical analysis for rooting intensity of seminal roots among treatments. Rooting intensity per plant (a), in P-rich patch (b), and in P-poor patch (c). The different letters indicate significant differences among the treatments (Tukey HSD test, P < 0.05). Table S3 Statistical analysis for rooting intensity of nodal roots among treatments. Rooting intensity per plant (a), in P-rich patch (b), and in P-poor patch (c). The different letters indicate significant differences among the treatments (Tukey HSD test, P < 0.05). Table S4 Statistical analysis for rooting intensity of lateral roots among treatments. Rooting intensity per plant (a), in P-rich patch (b), and in P-poor patch (c). The different letters indicate significant differences among the treatments (Tukey HSD test, P < 0.05).
VideoS1.mp4
Video S1 Spatio-temporal response of each root type to local P supply treatments.

Download PDF

Version 1

posted

You are reading this latest preprint version

Where, when, and which root types do respond to local phosphorus supply?

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Declarations

References

Supplementary Files

Status:

Version 1