Excluding environmental variables, weed competition is the major constraint to crop production for many agroecosystems worldwide. Besides serious yield losses, various damages including the increased cost of cultivation, reduced quality of field products, propagation of pests and diseases, secretion of phytotoxins or reduced irrigation efficiency may be attributed to weed infestation (Holm et al., 1991). Investigation of the mechanisms and outcomes of crop–weed competitive interactions becomes increasingly important for designing more efficient and economical weed management strategies in agricultural areas (Craine & Dybzinski, 2013; Lehoczky et al., 2013). Most studies tend to focus only on the aboveground symptoms of plant interference, even if roots play a key role in competition for place, water, and nutrients (Rajcan & Swanton, 2001; Trinder et al., 2013). The insufficient information on belowground phenomena can be obviously attributed to the hidden character of plant roots associated with the cumbersome root sampling procedures.

Measurement of electrical capacitance (EC) in a plant–soil system is capable to provide a rapid assessment about the extension of root systems without damaging plants. The method was developed by Chloupek (1972) using numerous crop species under greenhouse and field conditions. By applying a capacitance meter (LCR-instrument) with alternating current, EC can be measured between a ground electrode embedded in the soil and a plant electrode fixed on the plant stem. The magnitude of EC proved to be positively correlated with root weight, root surface area or root length. First Dalton (1995) presented a conceptual model for describing the plant root–soil system in which root surface area was considered to be the surface area of a group of circular cylinders having the same average diameter as the cellular system constituting the roots (Fig. S1 [suppl.]). According to the root–soil–electrode network model, xylem and phloem sap in plant roots forms a low-resistance electrical conduit which is separated from a low-resistance external soil or nutrient solution by insulating root membranes. The capacitance of this root–soil interface is proportional to the charge accumulated on the membrane surfaces; thus the membrane polarized by an alternating current plays the role of a dielectric in a capacitor.

The EC can be specified by its magnitude and phase angle (Ф, in degrees); the latter shows the phase difference between current and voltage, indicating the “capacitive strength” as a tendency of plant root to behave as a capacitor. The root–soil system is a relatively poor capacitor. The phase angle is –90° in ideal capacitors irrespective of current frequency, but in the root–soil systems, Ф is much lower in general; furthermore, it depends on the applied current frequency and also on soil properties and plant age (Aubrecht et al., 2006; Cseresnyés et al., 2013a). Phase angle is the dielectric component of the plasma membrane, and is basically determined by the physicochemical structure of the root tissue. Though Ф in itself does not indicate root extension and activity, its magnitude and temporal changes can inform us on the age-dependent alteration of root structure and functionality.

The main limitation of the EC method is the sensitivity of electrical properties to soil water saturation, ionic status, soil texture and to the distance of plant electrode from substrate surface (Ozier-Lafontaine & Bajazet, 2005; Dietrich et al., 2012). Nevertheless, under well-defined conditions (constant soil moisture content, soil salinity and consistent electrode placement) and by the application of an effective current frequency, the method can provide a good estimation of the root system size. On the one hand, the technique can be used for relative comparison of root development; on the other, it is possible to formulate calibration relationship with reasonable predictive ability to assess the root properties for a given plant in a given soil (McBride et al., 2008). Reliability of the technique has been demonstrated by various pot experiments and field studies focused on crop plant genotypes (Beem et al., 1998; Svačina et al., 2014; Heřmanská et al., 2015) and young tree cultivars (Preston et al., 2004; Cao et al., 2010; Pitre et al., 2010) in most cases. The great advantage of EC method over other in situ techniques that it can simply study the functional aspect of the plant root system, providing information about the absorptive root surface area (Cseresnyés et al., 2013b).

The aim of this methodological work presented was to test the applicability of EC measurement for in situ monitoring of root development and biomass loss of crop and weed plants subjected to competitive interactions in a pot experiment with two weeds native to southern Asia which (due to their broad ecological adaptation) have become widely distributed in various agricultural areas of the Northern Hemisphere during the last century. They are considered to be among the most troublesome yield-limiting weeds in many temperate and tropical crops including maize, rice, cotton, sorghum, soybean, cassava and vegetables (Warwick & Black, 1988; Holm et al., 1991; Hyvönen & Ramula, 2014).

The results show the ability of electrical capacitance measurement to investigate root mass in situ and to show changes over time in relation to various competitive interferences. The rate of decrease in root EC caused by plant interaction is therefore a promising estimator for the strength and time-course of competition through the indication of the loss of above- and belowground biomass. In this respect, results obtained fit in with differences in competitive strategies evolved by the two weed species as reported in the literature: Echinochloa crus-galli is known to be an aggressive belowground competitor even in the early growth stage (Martinková & Honěk, 1998; Bajwa et al., 2015). Its fibrous root system overlays the crop roots quickly, resulting in effective water and nutrient extraction from the soil (Perera et al., 1992; Chauan & Abugho, 2013). Growth inhibition in crops is not only due to the decreased water and nutrient content, but rather to the reduced ability of crop roots to extend and reach resources (the phenomena is undoubtedly intensified by space limitation in pots). In our experiment, common nutrient and light deficiency symptoms, including thin stem, general chlorosis and enhanced senescence in older leaves of maize plants became visible first on 16 and 20 DAP in Z-E5 and Z-E3 groups, respectively (5 and 8 days earlier than in the maize–A. theophrasti competition). The EC and biomass data supported concurrently the higher competitive capacity of E. crus-galli over crop in comparison with the impact of A. theophrasti, especially when weeds were grown with maize in low density (Fig. 4A and 4B). Echinochloa crus-galli proved to be less sensitive to intraspecific competition at medium density (E3) than A. theophrasti, and maize competitiveness was weaker over E. crus-galli than against A. theophrasti (Figs. 4C and 4D).

Abutilon theophrasti shows its competitive ability during the mid and late vegetative phase until the beginning of flowering (Sattin et al., 1992; Nagy et al., 2014). Weed causes maize yield loss principally through competition for light by overtopping the crop, and by producing allelopathic chemicals (Kazinczi et al., 2001; Traore et al., 2002; Bonifas & Lindquist, 2006). The outcome of maize–A. theophrasti competition can be determined by water and nutrient (mostly the nitrogen) availability and by the relative emergence time of weed and crop (Lindquist et al., 2007; Qaderi et al., 2015). In our study, the optimal water and the limited nitrogen supply with the simultaneous maize–weed emergence did not allow A. theophrasti the rapid early growth and crop overtopping by the end of experiment. Thus, the decrease in root EC (likely inhibition in maize root growth) was detected later and only at high (Z-A5) and medium (Z-A3) weed density with the appearance of nutrient deficiency symptoms from 21 and 28 DAP in these groups, respectively. Abutilon theophrasti proved to be more sensitive both to intraspecific competition and to maize interference than E. crus-galli. When A. theophrasti grew without crop, partitioned proportionally more biomass to shoot versus root, irrespective of density. Competitive interaction with maize inhibited mainly the shoot development of weed, leading to a significant increase in root/shoot ratio.

Temporal changes in the magnitude and phase angle (Ф) of EC in maize refer to the altering functionality and physicochemical structure of the root system (Rajkai et al., 2005; Cseresnyés et al., 2013a). Both parameters increased continually to a maximum, thereafter showed a downward trend (Fig. 1). The decrease is thought to be caused by root death and by the reduction in root activity due the deposition of the impermeable, electrically insulating suberin in endodermal cell walls (Dalton, 1995; Čermák et al., 2006). The increasing weed competition (density) provoked an additional decrease in the Ф of maize root. The rising interference of weed roots is a strengthening constraint to crop root extension and proliferation that leads to an increasing proportion of older root segments to the younger ones. Therefore, the enhanced integrated root system age associated with the higher ratio of suberized root segments decreased the uptake activity (EC) and the capacitive strength (Ф) of the whole root system (Gao et al., 1998). Since Ф is less sensitive to the altered root growth than the magnitude of EC, the change was significant solely at high weed densities (Z-E5 and Z-A5).

Regression analyses (Fig. 5) proved the feasibility of EC measurement for the accurate determination of root dry mass for each plant species applied. Considering EC a reliable indicator to active root surface and water uptake rate (Aubrecht et al., 2006), it points to the plant specific differences in our case. Monocotyledonous maize and E. crus-galli seemed to maintain similar root uptake rate per root mass (nearly overlapping regression lines). The steeper regression line achieved for A. theophrasti refers to the higher root uptake rate, which may be linked to a higher partitioning of biomass to shoot, leading in turn to a higher transpiration activity (particularly without competition effect). The positive intercept for the regressions is due to the electrode polarization (may occur up to a few kilohertz) and to the EC of plant stem between the soil surface and plant electrode (Ozier-Lafontaine & Bajazet, 2005). This is why the rate of decrease in root EC is consistently lower by a few percent than the relative biomass loss in the same comparison (Fig. 4). Nevertheless, the relative decline of root EC caused by competition seems to be a quite good indicator of the biomass loss. Though there was only one harvest at the end of this experiment, we formerly verified the close correlation between root biomass and EC in maize of different ages grown under different conditions by repeated plant harvest (Cseresnyés et al., 2013b).

Our results illustrate that EC measurement in a plant–soil system is a promising and useful method for studying in situ the effect of competitive interferences on root extension, root function and on rate of change in biomass loss. The instrumental process is very simple, inexpensive, time-saving and principally non-destructive, thus is suited for continuous plant monitoring, approaching to the functional side of root development. By applying the EC method, not only the final outcome of plant interactions can be detected, but it is possible to obtain information about how the competitive growth inhibition be realized in time (decrease in root EC became significant 3–5 days before the appearance of nutrient deficiency symptoms). Competition experiments with EC measurements can be carried out by growing various crop cultivars under different environmental variables (e.g. water regime, nutrient supply, temperature, light conditions, elevated CO₂ concentration, herbicide treatment) in order to study the effect of these factors on plant competitiveness. Nevertheless, it is important to emphasize, that measurement data are comparable only among plants of the same species, grown in the same substrate at the same moisture level in the same time frame (Chloupek et al., 2010). A drawback to the method against conventional procedures (i.e. core sampling or rhizotrones) that EC does not provide information about the relative depth distribution and morphology of the roots, which plays important role in competition.

Though several studies demonstrate the relevance of the process in the field (Beem et al., 1998; Preston et al., 2004; Chloupek et al., 2006; Ellis et al., 2013b; Svačina et al., 2014; Nakhforoosh et al., 2014), EC method seemed to be reliable under relatively homogeneous soil conditions (e.g. sandy soils) with adequate water supply. Temporal changes in soil water content usually preclude the possibility of the continuous monitoring of root activity under real field conditions. In these cases, different treatments can be compared by using their single-time root EC data collected within one day without rain (Chloupek et al., 2006, 2010), provided that soil water contents were statistically equal around all plant roots studied. It is advised to complete EC measurement simultaneously with the detection of soil moisture content (with a TDR instrument) in the root zone. If the investigation is carried out on a relatively large area, a systematic soil sampling (for assessing ionic status, texture and organic matter content) would be needed. Future experiments are worth completing with intermediate biomass sampling or with surveys focused on the relation of EC to soil texture or moisture. Variability of soil temperature may also affect the measured data. Field conditions are expected to require a higher number of replicates to cover the greater heterogeneity of plant population. Due to the limitations described above, the capacitance method is therefore appropriate for pot experiments carried out in growth chambers, greenhouses and outdoor with somewhat controlled external factors.

The EC measurement may be of interest for future applications, particularly in pot experiments. It can partially substitute or integrate the intrusive and time-consuming techniques commonly used or may be combined with other physiological investigation methods. The root system EC can become an important part of the biochemical or genetic data set for plant stress physiology researches (though it cannot be used for continuous drought stress as measurements are dependent on soil moisture). These functional approaches will be helpful tounderstand and elucidate weed life strategies and competitive mechanisms for limiting resources in relation to both intra- and interspecific interactions.