PLoS OnePLoS One440plosone101285081plosPLoS ONE1932-6203PLOSPMC7540849PMC7540849.1754084975408493302728910.1371/journal.pone.0239673PONE-D-20-208991Research ArticleBiology and Life SciencesOrganismsEukaryotaPlantsLegumesBiology and Life SciencesOrganismsEukaryotaPlantsGrassesBiology and Life SciencesAgricultureAgricultural MethodsMonoculture CroppingBiology and Life SciencesAgricultureAgrochemicalsFertilizersResearch and Analysis MethodsMathematical and Statistical TechniquesStatistical MethodsMultivariate AnalysisPrincipal Component AnalysisPhysical SciencesMathematicsStatisticsStatistical MethodsMultivariate AnalysisPrincipal Component AnalysisBiology and Life SciencesDevelopmental BiologyPlant Growth and DevelopmentBiology and Life SciencesPlant SciencePlant Growth and DevelopmentResearch and Analysis MethodsResearch DesignExperimental DesignBiology and Life SciencesGeneticsPhenotypesHigh-throughput, image-based phenotyping reveals nutrient-dependent growth facilitation in a grass-legume mixtureImage-based phenotyping reveals nutrient-dependent growth facilitation in a grass-legume mixturehttp://orcid.org/0000-0001-9632-0874BallKirsten RaeConceptualizationData curationFormal analysisFunding acquisitionMethodologyProject administrationSupervisionVisualizationWriting – original draftWriting – review & editing12¤*PowerSally AnneConceptualizationSupervisionWriting – review & editing1http://orcid.org/0000-0003-0581-1817BrienChrisData curationFormal analysis3WoodinSarahConceptualizationSupervisionWriting – review & editing2JewellNathanielData curationFormal analysis3http://orcid.org/0000-0003-1195-4478BergerBettinaMethodologyResourcesSupervisionWriting – review & editing3PendallEliseConceptualizationMethodologySupervisionWriting – review & editing1 Hawkesbury Institute for the Environment, Western Sydney University, Penrith, New South Wales, Australia Institute of Biological & Environmental Sciences, University of Aberdeen, Aberdeen, United Kingdom Australian Plant Phenomics Facility, The Plant Accelerator, School of Agriculture, Food and Wine, University of Adelaide, Urrbrae, South Australia, AustraliaPapaRobertoEditorUniversità Politecnica delle Marche, ITALY

Competing Interests: The authors have declared that no competing interests exist.

Current address: Environmental Sciences, University of Arizona, Tucson, Arizona, United States of America.

* E-mail: kirsten.ball@email.arizona.edu710202020201510366205e02396738720201092020071020201910202001112024© 2020 Ball et al2020Ball et alhttps://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.10.1371/journal.pone.023967310.1371/journal.pone.0239673

This study used high throughput, image-based phenotyping (HTP) to distinguish growth patterns, detect facilitation and interpret variations to nutrient uptake in a model mixed-pasture system in response to factorial low and high nitrogen (N) and phosphorus (P) application. HTP has not previously been used to examine pasture species in mixture. We used red-green-blue (RGB) imaging to obtain smoothed projected shoot area (sPSA) to predict absolute growth (AG) up to 70 days after planting (sPSA, DAP 70), to identify variation in relative growth rates (RGR, DAP 35–70) and detect overyielding (an increase in yield in mixture compared with monoculture, indicating facilitation) in a grass-legume model pasture. Finally, using principal components analysis we interpreted between species changes to HTP-derived temporal growth dynamics and nutrient uptake in mixtures and monocultures. Overyielding was detected in all treatments and was driven by both grass and legume. Our data supported expectations of more rapid grass growth and augmented nutrient uptake in the presence of a legume. Legumes grew more slowly in mixture and where growth became more reliant on soil P. Relative growth rate in grass was strongly associated with shoot N concentration, whereas legume RGR was not strongly associated with shoot nutrients. High throughput, image-based phenotyping was a useful tool to quantify growth trait variation between contrasting species and to this end is highly useful in understanding nutrient-yield relationships in mixed pasture cultivations.

The Australian Plant Phenomics Facility received grant funding from the Australian Government through the National Collaborative Research Infrastructure Strategy (NCRIS). KB received a Postgraduate Internship Award from the Australian Plant Phenomics Facility towards the completion of this project”.pmc-status-qastatus0pmc-status-liveyespmc-status-embargonopmc-status-releasedyespmc-prop-open-accessyespmc-prop-olfnopmc-prop-manuscriptnopmc-prop-legally-suppressednopmc-prop-has-pdfyespmc-prop-has-supplementyespmc-prop-pdf-onlynopmc-prop-suppress-copyrightnopmc-prop-is-real-versionnopmc-prop-is-scanned-articlenopmc-prop-preprintnopmc-prop-in-epmcyespmc-license-refCC BYData AvailabilityData are available via Figshare. DOI: (10.25909/12895121).Data Availability

Data are available via Figshare. DOI: (10.25909/12895121).

Introduction

Farm management plays a crucial role in global food security as it affects food production and the environmental impacts of agricultural practices [1]. Generally, fertilisation increases crop yield; however, mineral fertilisers are costly and can have detrimental effects on the wider environment [2]. Nitrogen (N) fertilisation can accelerate soil carbon turnover and gaseous emissions [3, 4], and assuming zero growth in agricultural production, world rock phosphate (P) reserves are only expected to last approximately 260 years [5]. As global grain demand is projected to increase by around 50% in the next 30 years, environmental impacts from fertiliser applications will become unsustainable, highlighting the need to improve agricultural nutrient use efficiency [1, 5–7].

Optimising crop growth under nutrient application can be achieved via several different practices. These include preferential planting of ‘environmentally appropriate’ species [8], precision nutrient application to match plant demand [9], genetic engineering [10] and intercropping of complementary species [11]. Plants have two overarching growth strategies in response to N and P availability: (i) increasing nutrient uptake and directing them towards greater or more rapid growth, or (ii) conserving and storing nutrients and slowing or maintaining growth-rate [12, 13]. Plant growth strategies involve trade-offs between increasing both productivity and resource acquisition (interpreted herein as an acquisitive growth strategy), and reducing productivity and conserving resources (interpreted herein as a conservative growth strategy) [12]. Grasses require substantial amounts of mineral N, mainly owing to rapid growth rates and a lengthy growing season [14], and legumes are slower growing and less reliant on mineral fertilisers [15, 16]. Legumes can provide N to other species via a symbiotic relationship with nitrogen fixing bacteria but require moderately high levels of P to support biological nitrogen fixation (BNF) [17, 18]. Ultimately, plant growth strategies will determine growth rates, nutrient requirements and interactions with other plants [12], and species growth strategies drive variations to primary productivity in agricultural systems [19].

Interspecific interactions between plants can be negative, in the case of competition for resources [20], neutral, where complementarity ensures that species do not compete for the same resources [11], and positive, where facilitation leads to higher performance of a species when grown in mixture than in monoculture [21]. In mixed species pastures, niche complementarity and facilitative relationships can be beneficial. If facilitation is occurring, the total production of a species is likely to be significantly greater in a mixture than its average production in a monoculture—referred to as “overyielding” [22–24]. Mechanisms behind overyielding have been attributed to belowground processes such as enhanced mobilisation of soil nutrients [25] and aboveground interactions including microclimate improvement and light partitioning [26]. Measurements of overyielding combined with examinations of temporal plant growth dynamics can assist us to determine how net primary productivity is likely to vary under particular environmental conditions [27, 28]. The most traditional assessment of plant growth–absolute growth (AG)—reports the total increase in biomass per unit of time and is usually obtained using final harvest weights [29]. Relative growth rate (RGR), a lesser used but potentially more informative assessment of growth, describes the increase in size relative to the size of the plant at the beginning of a time interval [29]. RGR is strongly determined by a plants metabolic requirement for N and P [30], and hence analyses of growth rate diversity in pasture mixtures can assist in detecting potential for nutrient competition or assessing facilitation.

Calculating growth dynamics can be labour- and cost-intensive and often impractical when reliant on destructive harvest techniques [31, 32]. Additionally, calculations of relative growth obtained from weight and height measurements of different species may introduce bias and inaccurate comparison [33, 34]. The recent emergence of high throughput (HT) imaging techniques that allow rapid temporal assessment of plant phenotype by environment interactions, and are scalable to the farm scale, are contributing significantly to relieving this research bottleneck [35, 36]. High throughput phenotyping (HTP) is an automated technique providing quantification of plant traits without the need for destructive harvest [37]. In multiple-image based systems, cameras operating at different angles allow derivation of a mathematical relationship between several two-dimensional (2D) red-green-blue (RGB) images to quantify plant size [38]. Shoot architectural traits such as canopy biomass, density and leaf area are primarily extracted from RGB images and are used to calculate AG and RGR. Additional information can be acquired using hyperspectral imaging to calculate traits related to foliar nutrient content [25, 39, 40]. All HTP methods require calibration to improve our ability to relate image information to plant growth dynamics, in order that these datasets may be used to measure phenotypic variation in biological systems of interest [25].

This study used image-based HTP to 1) predict absolute and relative growth rates, and from those predictions detect facilitation in a mixed pasture cultivation; and 2) interpret variations in RGR and nutrient uptake between grass-legume cultivations under contrasting N and P inputs (factorial combination of low (L) and high (H) inputs of N and P). As the general growth strategies and responses of grass and legume species to fertilisation and intercropping are well known, we applied HTP to test the following:

RGR, AG and nutrient uptake in grasses would be primarily influenced by increased N availability owing to their rapid, nutrient acquisitive growth strategy. RGR and AG in legumes would be preferentially influenced by P availability as P is required in biological nitrogen fixation. Legumes are also expected to display a slower, more nutrient conservative growth strategy and their growth is expected to be less tightly correlated with soil and shoot nutrients.

Overyielding (greater yield in mixture compared to monoculture, indicating facilitation) would be present if N or P, but not both, were limiting. This is because P is required by legumes to support biological nitrogen fixation (BNF) and induce facilitation, and under N replete conditions grasses are expected to grow faster.

Where overyielding is occurring, because of their slow, conservative growth strategy legume RGR is expected to decline, but their shoot N and P concentrations would remain stable.

Materials and methodsExperimental design and species

Our experiment was conducted between the 24th July 2018 and 5th October 2018 at the Australian Plant Phenomics Facility at the University of Adelaide (-34.971298, 138.639627) in a Lemnatec Imaging System, during which time we obtained plant images on 43 consecutive days. The experiment investigated the effects of four fertiliser treatments derived by factorial combination of two levels of nitrogen (LN and HN) and two levels of phosphorus (LP and HP) on one grass (Phalaris aquatica) and one legume (Trifolium vesiculosum) pasture species grown in monoculture and mixture. The twelve treatments were arranged in a latinized, resolved incomplete-block design with ten replicates, for a total of 120 pots. The two Australian pasture cultivars, Holdfast GT (P. aquatica) and Cefalu (T. vesiculosum), were chosen as they have a similar upright growth form, are both winter and early spring active and are suited to a wide range of sandy clay loam soils. Seeds were obtained from Heritage Seeds Australia, along with the appropriate group C rhizobial inoculant required for T. vesiculosum. The experimental design was obtained using CycDesigN [41] and randomized using the dae package in R [42, 43] (Fig 1).

10.1371/journal.pone.0239673.g001Main experimental design.

Within each replicate (indicated by the solid red box) the four nutrient combinations (LNLP, HNHP, HNLP and LNHP) were assigned to 4 main units arranged in a grid according to the experimental design. The three species combinations [grass only (Gra), legume only (Leg) and mixture (Gra Leg)] were randomized to the three consecutive pots within each main unit. Coloured blocks indicate 1 pot containing two halves.

Growing conditions

We used pasteurized, unfertilised potting mixture consisting of pre-mixed (0.33:0.33:0.33 by volume) sand, clay loam and coco peat at pH 6.3. Three kg (dry weight) of soil was potted into 198 mm diameter x 149 mm high (4587 cm3) drainage-drilled pots, each seated on a 200 mm round dish ensuring no water or nutrient loss occurred from the system. Eight seeds were planted in each pot (Day after planting; DAP 0), with two seedlings on each half of the pot after thinning (DAP 16). In mixtures, one half comprised grass while the other half comprised legume. The plants were physically separated aboveground by a white plastic divider, set 5 cm into the soil and oriented north–south on the conveyor system in the greenhouse for consistency with respect to imaging orientation and solar exposure. The divider allowed intermingling of roots belowground (Fig 2).

10.1371/journal.pone.0239673.g002Pot-level separation of plant biomass and experimental design to improve image differentiation.

a) Cart layout and pot level separation using above-ground dividers (DAP 35); b) digitally enhanced RGB image (top view) of a mixed-species pot, showing centre divider (white), grass (natural green colour, right; eastern half of pot) and legume (orange highlights, left; western half of pot); c) oblique view of a mixed-species pot at final harvest (DAP 70) with centre divider removed; d) digitally processed RGB image (side view) of a mixed-species pot, showing only the grass component in the eastern half (artificial green; superimposed white lines correspond to the enclosing wire frame, which was painted blue for maximum visual contrast).

To promote microbial activity, we added a field soil microbial wash generated from unfertilised (>30 years) old-field pasture soil obtained from the Hawkesbury Forest Experiment in Richmond, NSW (-33.611672, 150.740172). Briefly, 1 g of field soil was mixed with 100 ml of DI water and molasses in a ratio of 100:1 (ml) with the recommended amount of Group C rhizobial inoculant to obtain ~ 3,000,000 rhizobia. The 100 ml of microbial wash was added to every experimental pot on DAP 16 after the seedlings had established. Effective nodulation and belowground biomass were assessed on DAP 35 prior to fertilization in non-experimental pots (n = 6) and both were considered adequate to assert that the inoculation was successful [44, 45] and intermingling of roots was occurring. Ongoing, plants were watered once daily, and soil water content was maintained at field capacity (22% (w/w) gravimetric water content) by watering to weight. Commencing DAP 29, weighing and watering-to-weight was incorporated into the imaging sessions. Twice-daily watering was performed near the end of the experiment (DAP 66–69), when water consumption was high. The greenhouse maintained average temperatures of 27°C/16°C on a day/night cycle and an average day length between 07:00–19:00 under natural light conditions for a total of 70 days.

Fertilisation treatments

Nitrogen (N) and phosphorus (P) were prepared in DI water in the forms of ammonium nitrate (NH4NO3) and a pH 6.3 balanced mixture of disodium phosphate (Na2HPO4) and sodium dihydrogen phosphate (NaH2PO4). Nutrients were added on a dry-weight, mg kg-1 of soil basis. The low N-low P (LNLP) treatment nutrients (33 mg N, and 11 mg P, N:P ratio; 3:1) were added to all pots prior to imaging on DAP 16. On DAP 35, nutrients were added to increase the total amount of N and/or P to desired treatment levels. In total, for the high N-high P (HNHP) treatment, we added 99 mg N and 33 mg P (N:P ratio 3:1). For the HNLP treatment, we added 99 mg N and 11 mg P (N:P ratio 9:1). For the LNHP treatment, (33 mg N and 33 mg P, N:P ratio; 1:1). All pots received other macro- and micronutrients at the following rates (mg kg-1 dry soil): K2SO4, (75); CaCl2 2H2O, (75); MgSO4 7H2O, (45); CuSO4 5H2O, (2.1); ZnSO4 7H2O, (5.4); MnSO4 H2O, (6.4); CoCl2 6H2O, (0.33); Na2MoO4 2H2O, (0.18); H3BO3, 0.3 and FeEDTA, (0.4). Plant available macronutrients measured in plant-free pots are detailed in S1 File.

Biomass harvest

On DAP 70, all above ground shoot biomass (AGB) was harvested and soil samples collected at the pot level to determine total AGB, shoot nutrient concentrations and soil extractable nutrients (N and P). AGB was assessed by carefully separating the plants between the two sides of the pot and cutting at soil level to separate from belowground biomass. AGB was dried at 70°C and weighed and the reported AGB is the total weight of the two individual plants from each side of the pot, here forward referred to as “half-pot”. Where “whole-pot” values are reported, this represents the total dry weight of all plant biomass within one pot. All plants were deemed viable for harvest, except a single pot in the grass only LNLP treatment which did not survive. Clerical error was suspected for one half-pot dry-weight observation in the LNLP grass monoculture and LNHP grass monoculture and these observations were removed from analysis. The final dataset included n = 236 useable half-pot observations. All nutrient analysis observations for these replicates were also removed.

Soil and shoot nutrient analyses

Soil samples were extracted within 12 h of biomass harvest. Extractable N in soil was determined by shaking 40 ml of 2 M potassium chloride (KCl) solution with 4.0 g soil (< 2 mm) at 170 rpm for 1 hour and then filtering through a 2.5 μm ashless filter (Grade 42, Whatman PLC, Kent, U.K). Extractable P was determined by mixing 4 g soil in 40 ml of 0.5 M NaHCO3 and shaking for 16 hours [46]. Soil extracts were stored at -20°C until colorimetric analysis in a discrete analyser (AQ2, SEAL Analytical, Ltd., Milwaukee, WI USA and EPA135 method). For total carbon (C) and N shoot nutrients, a subsample (~3 g) of AGB from each plant was finely ground and homogenised with an MM 400 mixer mill (Retsch, GmbH, Haan, Germany) and an approximately 5 mg subsample was taken for combustion analysis. C and N concentrations were estimated using an Elementar Vario El Cube Carbon/Nitrogen analyser (Elementar Analysersysteme GmbH, Langenselbold, Germany). Phosphorus concentration of foliar samples was obtained after digesting approximately 55 mg of plant material in concentrated H2SO4 and H2O2 in a microwave digester, and colorimetric analysis following an ammonium molybdate reaction [47]. Measurement error was suspected in two replicates for total C, N and P; one in the HNLP legume monoculture and one in the LNHP grass monoculture and were removed from the analysis.

Imaging

Each pot was individually imaged daily using four different cameras, comprising one top view camera and three side view cameras at different angles. Of the four images obtained, three (top view and two side views) were used for image analysis and estimation of projected shoot area (PSA). These images had the right orientation to separate the images along the midline, enabling a separation of the two halves of the pot along the plastic divider (Fig 2). LemnaGrid software (LemnaTec GmbH, Aachen, Germany) was used to separate the images into two halves and measure the pixels corresponding to plants in each half. For this, a nearest neighbor colour classification was used to separate foreground and background, followed by noise removal steps. It was decided not to exclude any imaging data as outliers for analysis purposes.

Data processing and growth calculations from image-based data

Image-based data were processed using the multi-step method of smoothing and extraction of traits (SET) described by [48] with the aid of growthPheno [49], an R package [43]. Firstly, the half-pot projected shoot area (PSA) was defined as the sum of plant pixels visible in the three half-images in each half-pot on each imaging day, yielding n = 240 observations per day. The whole-pot projected shoot area (PSA) was defined as the sum of plant pixels visible in the six half-images in each whole-pot on each imaging day, yielding n = 120 observations per day. In this paper, the whole-pot data was only used to compare between measured AGB vs PSA predicted values. The raw PSA (kpixels) data exhibited a high degree of day-to-day variation and so the next step was to smooth the PSA data to produce what is herein referred to as sPSA; natural cubic smoothing splines with df set to 5 were used (moderate smoothing) [50]. Thirdly, the time points DAP 35, 40, 50, 60 and 70 were chosen for further analysis based on changes in growth pattern observed from sPSA plots. Only measurements taken after application of nutrients (DAP 35–70) are presented. Finally, RGR (day-1) was calculated for the intervals between successive pairs of time points using Eq 1: RGR(t1,t2)=log[sPSA(t2)/sPSA(t1)] where t1 and t2 are the DAPs defining an interval.

We calculated image-derived AG and overyield from sPSA on DAP 70 and related these values to DAP 70 harvested AGB using correlation analyses.

Correlations of image-based biomass estimates to harvested aboveground biomass

We used the ‘lm’ function from Base R [43] to obtain correlation coefficients between our absolute growth values (sPSA DAP 70) and harvested dry weight biomass (AGB) in grams. All model assumptions of normality and homoscedasticity were met.

Statistical analysis of growth and nutrient concentrations

To obtain growth predictions from the image-based data, and to compare shoot nutrient concentrations between treatments a mixed-model analysis was performed using the R package ASReml-R [51] and asremlPlus [52] packages for the R statistical computing environment [46]. We analysed the half-pot responses based on a maximal mixed model including terms for the treatment differences, spatial effects and residual error variation. The model was of the following form; y=Xβ+Zu+e where y is the response vector of values for the trait being analysed; β is the vector of fixed effects; u is the vector of random effects; and e is the vector of residual effects. X and Z are the design matrices corresponding to β and u respectively. The fixed-effect vector β was partitioned as β⊤=[μ βR⊤ βSi⊤ βH⊤ βSp⊤ βG⊤ βL⊤] where μ is the overall mean and the first three β subvectors correspond to the effects of replicates (R), greenhouse sides (Si, east or west) and pot halves (H, east or west) that capture systematic spatial variation within the greenhouse; βSp⊤ incorporates the two species main effects; βG⊤ contains parameters for the 3 main effects, the 3 two-factor interactions and the three-factor interaction of the factors cultivation type, nitrogen and phosphorus for the grass (G) species; and βL⊤ contains the same parameters for the legume (L) species. The random-effects vector u was partitioned as [uR:MuR:M:P]where uR:M is the vector of main-unit (M) random effects within each replicate (R) and uR:M:P is the vector of random effects for pots (P) within each main-unit (M); these vectors captured any non-trend spatial variation. The residuals e were assumed to be normally distributed with their variance allowed to vary with both species and nitrogen. If the data y are ordered by species (labels G and L) followed by fertiliser level (labels L and H) and then observations with the combinations of species and fertiliser level, then the residuals are modelled as: N(0240,[σGL2I60060060060060σGH2I60060060060060σLL2I60060060060060σLH2I60]), where I60 and 060 denote identity and zero matrices respectively. For each trait, residual likelihood ratio tests with α = 0.05 were used to determine whether the variance model can be simplified by removal of the nitrogen level variance difference and/or species variance difference. The model was modified to reflect the results of these tests and residual-versus-fitted values plots and normal probability plots confirmed that model assumptions were met. Wald F-tests with α = 0.05 were conducted for the fixed effects within each species to determine a model for describing how cultivation, nitrogen and phosphorus affect the response for each species. Testing began with the three-factor interaction for a species and, only if it was not significant, proceeded to test the two-factor interactions; the main effects were only tested if that factor had not occurred in a significant interaction. Thus, the estimated marginal means (Lenth et al., 2019), the means for the selected model, were obtained using the R packages ASReml-R [51] and asremlPlus [52].

Calculating overyielding from sPSA DAP 70 values

Overyielding calculations in mixtures were conducted using methods described by [22, 24]. All assumptions including ensuring consistent planting densities and the allowance of sufficient time for below-ground community interactions to develop, were met [53]. Using our absolute growth (sPSA DAP 70) values we obtained relative yield totals (RYT, Eq 2) for each treatment from the sum of the relative yield (RY) of each species (i): RYT=∑i=1sRYi where s is the number of species i and RY is RYi=YmixYmon where Ymix is the observed yield of species i in mixture and Ymon is the observed yield of species i in monoculture expressed as a ratio (Ygra(Mix):Ygra and Yleg(Mix):Yleg). Ratios that contribute to the analysis of overyielding: Ygra:Yleg yield of grass vs legume in monoculture; Ygra(Mix):Yleg(Mix) yield of grass vs legume in mixture; Ygra,leg:Ŷgra,leg total yield in mixed cultivation compared with expected cumulative yield of constituent species according to monocultures; Ygra(Mix):Ygra yield of grass in mixture vs monoculture (RYT); Yleg(Mix):Yleg yield of legume in mixture vs monoculture (RYT). Where overyield was detected from calculations, we determined the response to be statistically significant where the mixed model results revealed that the factor ‘cultivation’ significantly influenced species yield (sPSA DAP 70), and we obtained multiple comparisons using the ‘emmeans’ package [54] within R [43].

The relationship between species-specific growth rates and nutrient uptake

Principal components analysis (PCA) using “prcomp” with varimax rotation in R [43] was conducted to examine relationships between our image-derived growth (RGR) and DAP 70 nutrient parameters (Soil extractable N and P and shoot N and P concentrations). Assumptions of sampling independence, normality, linear relationships between variable pairs and moderate correlations were confirmed. The optimal number of explanatory factors from each PCA was chosen based on eigenvalues greater than 1 [55]. Varimax rotation was applied so that information explained by one factor was independent of information in the other factors to achieve simple structure [56], and loadings <0.30 were not considered significant. S1 Table in S1 File details the eigenvalues and varimax loadings for all our principal components analyses.

ResultsAbsolute growth (smoothed projected shoot area on DAP 70) is correlated with aboveground biomass in mixed pasture cultivations

Aboveground biomass (DAP 70) was significantly correlated with sPSA (DAP 70) at the whole pot level (r2 0.77, p<0.001, n = 120) with half-pot level correlation coefficients differing between grasses (r2 0.85, p<0.001, n = 239) and legumes (r2 0.67, p<0.01, n = 240; S1a-S1c Fig in S1 File). There were no differences in these relationships between plants grown in monoculture or mixture.

RGR in grasses is driven by increased N availability and by presence of legumes, and in legumes by P availability and the presence of grasses

Significant treatment effects for RGR are shown in Table 1. RGR in grasses began higher than legumes and declined more rapidly, and growth ceased in grasses grown in monoculture without N after DAP 60 (Fig 3b). Between DAP 35–40 grass in the HNHP and HNLP treatments grew most rapidly regardless of cultivation, but between DAP 40–50, grasses in mixture under the HNHP and HNLP treatments achieved the highest RGR. Between DAP 60–70 a three-way interaction (cultivation by N by P) emerged. Here, RGR in grass under N addition were comparable between cultivations, whereas under LN, grass RGR was greater in mixture. Grass RGR under LN with P addition (LNHP) was higher in mixture compared to monoculture.

10.1371/journal.pone.0239673.t001P-values of the Wald F-statistics derived from linear mixed effects models for the main and interacting effects.
DAP35–4040–5050–6060–70
TreatmentRGR (kpixels day-1) Grass
C*N*P0.4430.7830.517<0.001
C*N0.1270.183<0.001nr
C*P0.7040.3540.715nr
N*P0.1630.0630.46nr
Cultivation0.201<0.001nrnr
Nitrogen<0.001<0.001nrnr
Phosphorus0.6660.2570.074nr
RGR (day-1) Legume
C*N*P0.6580.9310.0420.024
C*N0.0920.185nrnr
C*P0.5580.043nrnr
N*P0.8470.678nrnr
Cultivation0.168nrnrnr
Nitrogen<0.0010.073nrnr
Phosphorus0.656nrnrnr

Cultivation (C), nitrogen (N) and phosphorus (P) on the half-pot level trait RGR in grass and legume. Growth intervals are indicated by DAP (days after planting). Significant effects are indicated at (α = 0.05). nr indicates ‘not reported’ (suppression due to interaction effects).

10.1371/journal.pone.0239673.g003Tables of estimated marginal means.

a) Estimated marginal means for AG (sPSA DAP 70, kpixels) and b) RGR (day-1) by species (Eq 1), cultivation and nutrient treatment with half-least significant difference (5%) error bars. Non-overlapping error bars indicate significant differences. Purple = LNLP, Yellow = HNHP, Green = HNLP, Blue = LNHP.

Initial RGR in legumes (DAP 35–40) was increased by N addition (HNLP and HNHP treatments), but this effect did not persist for the remainder of the experiment (Fig 3b). Between DAP 40–60, RGR was increased by P addition in monocultures only. Between DAP 60–70, legume RGR was unaffected by nutrient treatment in monoculture but was altered in mixture where legumes benefited most under LNHP conditions. The lowest legume RGR was associated with N and P co-limitation in the LNLP mixed treatment.

Grasses and legumes contribute to overyield in all nutrient treatments

As species in our study were separated aboveground with a divider, we discuss interspecific interactions from the standpoint that belowground interactions were occurring. Overyield (sPSA DAP 70) was present in all nutrient treatments, supporting the widely reported benefit of mixed grass-legume cultivations on total pasture growth (Fig 3a, Table 2; p<0.05). The ratio Ygra,leg:Ŷgra,leg reveals the effect of mixed cultivation on absolute growth (AG). At the whole pot level, productivity in mixtures increased in all treatments. On average, the LNHP treatment increased by 27%, followed by LNLP which increased 21%, then HNLP which increased 20% and finally HNHP which was 10% more productive in mixture (Table 2; p<0.05). The ratios Ygra(Mix):Ygra and Yleg(Mix):Yleg determined individual species yield expected in mixtures compared with those observed in monocultures (Table 2). Overyield was contributed to by both grasses and legumes.

10.1371/journal.pone.0239673.t002Estimated marginal means of yield (sPSA DAP 70) and calculations of overyield (RYT, <xref ref-type="disp-formula" rid="pone.0239673.e008">Eq 2</xref>), between grasses (Gra) and legumes (Leg) across nutrient treatments.
FertilisersPSA± SE GrasPSA± SE Gra (Mix)sPSA± SE LegsPSA± SE Leg (Mix)Ygra:YlegYgra(Mix):Yleg(Mix)Ygra,leg:Ŷgra,legYgra(Mix):Ygra (RYT)Yleg(Mix):Yleg (RYT)
HNHP527.9 (24.4) de589.5 (30.7) ef548.6 (15.5) c596.3 (18.0) d0.960.981.101.111.08
LNHP200.8 (16.8) a359.4 (17.9) c548.6 (15.5) c596.3 (18.0) d0.370.601.271.781.08
HNLP493.3 (24.4) d639.9 (30.7) f436.6 (15.5) a484.2 (18.0) b1.121.321.201.291.10
LNLP200.4 (16.8) a289.8 (17.9) b436.6 (15.5) a484.2 (18.0) b0.450.591.211.441.10

Ygra:Yleg = yield of grass vs legume in monoculture; Ygra(Mix):Yleg(Mix) yield of grass vs legume in mixture; Ygra,leg:Ŷgra,leg yield in mixed cultivation compared with expected cumulative yield of constituent species according to calculations monocultures; Ygra(Mix):Ygra yield of grass in mixture vs monoculture (RYT); Yleg(Mix):Yleg yield of legume in mixture vs monoculture (RYT). All means (kpixels) and ratios were calculated from sPSA (DAP 70) values. Different letters represent significant treatment (nutrient and cultivation) differences at α = 0.05 within species.

Grass overyield was dependent on N or P addition, but not the two in combination (p<0.001). Grass under LNLP achieved 44% more biomass in mixture (p<0.001); in HNLP mixture grass achieved a 29% increase in biomass compared with monoculture (p<0.001). HNHP treatment overyield was not detected for grasses (Fig 3a; Table 2). The highest overyield was detected in the LNHP treatment, where grass was 78% more productive in mixture than in monoculture (p<0.001). The reliance of grasses on mineral N or by co-occurring legumes, was further evidenced by the fact that in LNHP and LNLP, grass monocultures could not maintain growth after DAP 60 (Fig 3a).

Legumes overyielded in all mixtures (Table 2), the highest attributed to P addition treatments (Fig 3a). Growth in mixed HP treatments (HNHP, LNHP) was 8% higher compared with monoculture, and although LP treatments (HNLP, LNLP) achieved lower yields than the HP treatments they were 10% more productive in mixture compared with monoculture.

Nitrogen and phosphorus interact to influence shoot nutrient concentrations in grasses and legumes between cultivations

Grasses increased shoot N (mg g-1) in response to N addition and being grown in mixture, where the effect of N addition was augmented (p<0.001, Table 3). Grass shoot N under HNLP and HNHP treatments was 32% higher in mixture, and in LNLP and LNHP treatments was 31% higher. Legume shoot N (mg g-1) was only increased by P addition, with no additional effect in mixture (p<0.001).

10.1371/journal.pone.0239673.t003Estimated marginal means of shoot nutrient concentrations (mg g<sup>-1</sup>).
Fertiliser TreatmentNut (mg g-1) ± SE GraNut (mg g-1) ± SE Gra (Leg)Nut (mg g-1) ± SE LegNut (mg g-1) ± SE Leg (Gra)
NITROGEN
HNHP14.34 (0.5) c18.96 (0.7) d31.06 (0.8) b31.06 (0.8) b
LNHP7.00 (0.2) a9.18 (0.2) b32.58 (1.0) b32.58 (1.0) b
HNLP14.34 (0.5) c18.96 (0.2) d29.17 (0.8) a29.17 (0.8) a
LNLP7.00 (0.2) a9.18 (0.2) b26.87 (1.0) a26.87 (1.0) a
PHOSPHORUS
HNHP3.48 (0.1) g3.75 (0.1) h1.98 (0.07) b1.98 (0.07) b
LNHP3.09 (0.1) e3.36 (0.1) f1.98 (0.07) b1.98 (0.07) b
HNLP1.17 (0.1) a1.45 (0.1) b1.27 (0.07) a1.27 (0.07) a
LNLP1.84 (0.1) c2.12 (0.07) d1.27 (0.07) a1.27 (0.07) a

Nutrient concentrations (mg g-1) were measured in shoot biomass harvested on DAP 70 for Grass (Gra) and Legume (Leg). Different letters represent significant treatment (nutrient and cultivation) differences within species, at α = 0.05.

Shoot P (mg g-1) in grass was influenced by an interaction between N and P (p<0.001), cultivation and P (p<0.01) and cultivation and N (p<0.01). Shoot P increased when N was added or in mixtures. Without N but with P addition (LNHP), grass shoot P was 8% higher in mixtures. The highest shoot P concentration was in grasses under the HNHP mixed treatment where it increased by 7% from monoculture. Shoot P (mg g-1) in legumes was influenced by P status (p<0.001) being higher when P was added but was unaffected by cultivation with grasses (Table 3).

Grass RGR parameters are related to shoot N in both cultivations, and legume RGR parameters diverge between monocultures and mixtures

Principal components analysis revealed differences in the relatedness of nine parameters pertaining to growth and nutrient use between grasses and legumes (S2 and S3a-S3b Figs in S1 File). Although DAP 35–70 RGR parameters were largely correlated in grasses, in legumes they were not and as such we decided to include all RGR parameters in our analysis to allow comparison between species.

Over the duration of the experiment, RGR in grass grown in both monoculture and mixture was strongly correlated with shoot N, and to a lesser degree, extractable N; it was uncorrelated (orthogonal) to shoot and extractable P concentrations (Fig 4a and 4b). Additionally, the correlation in RGR intervals suggests that RGR displayed consistency throughout the experiment for grasses.

10.1371/journal.pone.0239673.g004PCA loading plot of the first two principal components for a) grasses in monoculture and b) grasses in mixture.

Gradient colour scales and arrow length represent the percentage of variation explained by each variable along the principal component. RGR 35–70 = Relative growth rate (kpixels day-1), Ex P = Extractable P (mg g-1), Ex N = Extractable N (mg g-1), Shoot P = Shoot P (mg g-1), Shoot N = Shoot N (mg g-1).

Legume RGR in monoculture became increasingly correlated with nutrient availability, especially shoot P and extractable P, over the course of the experiment. Legume RGR in mixture showed a similar pattern, with a strong correlation to shoot and extractable P at the end of the experiment, but with a stronger correlation to extractable N and shoot N in the early stages of the experiment (Fig 5a and 5b).

10.1371/journal.pone.0239673.g005PCA loading plot of the first two principal components for a) legumes in monoculture and b) legumes in mixture.

Gradient colour scales and arrow length represent the percentage of variation explained by each variable along the principal component. RGR 35–70 = Relative growth rate (kpixels day-1), Ex P = Extractable P (mg g-1), Ex N = Extractable N (mg g-1), Shoot P = Shoot P (mg g-1), Shoot N = Shoot N (mg g-1).

DiscussionGrowth strategy is determined by nutrient availability and community interactions

Facilitation of N uptake [21] and growth in grass-legume mixtures is the primary motivation for intercropping in pastoral agriculture. N and P fertilization can have varying effects on this relationship [57], as well as the length of time that species interact under particular nutrient conditions [58]. We successfully measured growth facilitation using HTP, revealing a yield benefit for both grasses and legumes (Fig 3a). Grasses displayed a faster more acquisitive growth strategy, while legumes reduced growth rate, increased AG but maintained shoot nutrient concentrations (Fig 3a and 3b, Tables 1–3). Variation in species-specific growth strategies can shape resource use and primary productivity in agricultural systems as they determine requirements for N and P [59]. Using HTP, we accurately detected interspecific differences in response to nutrients, revealing higher AG and faster RGR in grasses responding to N, and a slower, more conservative growth strategy in legumes responding to P (Fig 3b). RGR in grasses was consistently related to shoot N concentrations in both cultivations, whilst in legumes, nutrient effects on growth were only detected towards the mid-to-late stages and were more pronounced in mixture (DAP 50–70), (Figs 4 and 5).

Mixed cultivation increases growth in both species and augments N and P uptake in grasses

Combining grass and legume is beneficial for pasture yield as legumes can increase grass growth via transfer of fixed N [60], and for pasture quality as the facilitative interaction can increase nutrient uptake and storage [16, 61, 62]. In addition to the facilitative benefit mixed cultivation presented for grasses, our data also revealed a clear growth benefit for legumes. Whilst not commonly discussed that grasses ‘facilitate’ legumes, grasses are demonstrated to prevent N leaching via their root structures [63] and to trigger up-regulation of BNF to benefit the legume [18, 58, 64]. In our study, grasses increased their shoot N and P concentrations in mixtures, and despite a growth increase in mixture, legume N and P concentrations remained constant. Robust, positive relationships between foliar N concentrations, RGR and yield have been reported in grasses [65], and the positive response of grass to N fertilization has been physiologically related to its fast RGR [66], which strongly determines a plant’s metabolic requirement for N and P [30]. In our study, increased RGR and foliar N concentrations in grasses following N addition with an augmented effect in mixture, suggested both a facilitative effect of legumes on grass growth [21] and the presence of an exploitative nutrient uptake strategy in grasses [66–68]. Our observed increases to shoot P concentrations (mg g-1) in grasses under N addition or in mixture, coupled with increased growth, provides support for the well-known theory of an increased P requirement in faster growing organisms [30, 69].

Legumes alter their growth in the presence of grasses but maintain consistent shoot nutrient concentrations

RGR in legumes appeared to differ between early-to-mid stage growth (DAP 35–50) that was significantly correlated with soil N availability in mixtures, and late stage growth (DAP 50–70) was not correlated with soil N. We suggest that during this early stage growth related more to mineral N than biologically fixed N, as legume nodules may not yet have properly developed [44, 45]. The mid-to-late legume growth stage (DAP 50–70), was associated with soil P and shoot N and P concentrations in both cultivations, but despite legume RGR slowing in mixture, there was no cultivation difference in final shoot nutrient concentrations. The legume RGR being more strongly affected by P availability in mixture towards the end of the experiment is likely owing to the prolonged interaction with grasses which may have up-regulated biological nitrogen fixation [27, 70]. A more conservative growth and nutrient uptake strategy in legumes that is mediated by community interactions and facilitates increased growth [20, 67] is positive for increasing overall forage quality and productivity in mixed pastures [16, 21, 62, 71].

The potential value of high-throughput phenotyping for mixed cultivations

High throughput phenotyping continues to develop as a promising technique to replace at least some of the traditional approaches to plant functional trait assessment [35, 36]. As our study is one of the first to examine mixed species, there was some expected variability in the characterization of projected shoot area measured against harvested biomass between species having different growth forms. In previous monoculture studies, correlations between projected shoot area obtained from RGB images and harvested plant biomass have generally been demonstrated as strong, with variations reported in relation to growth stage and plant height [72, 73] and plant functional type (grass vs legume) [74]. In the latter case, however, only canopy height was assessed, sensing was remotely obtained, and images were taken from a primarily top view. We believe that the combined use of top and side-view cameras may have reduced some of this error in our study, leading to relatively strong correlations (r2 ranging from 0.67 to 0.85) despite the interspecific differences. Nevertheless, the future capacity of HTP to inform agricultural management will continue to rely on robust and repeated calibrations in both controlled and field-based systems, and validation of acquired data in systems of interest.

Conclusion

Pastoral agriculture is a multi-billion-dollar industry that currently relies heavily on intensive fertilization for effective agricultural production. Novel sampling techniques such as image-based phenotyping have the potential to revolutionize our understanding of plant growth dynamics, thereby improving our ability to reduce reliance on mineral fertilizers. In addition, this dataset, calibrated under highly controlled conditions contributes to a more comprehensive understanding of plant-plant interactions in common agricultural systems, providing a platform for future testing in field-based systems.

Supporting information

(DOCX)

Click here for additional data file.

The authors wish to acknowledge the invaluable contributions of APPF technical staff, in particular Lidia Mischis, Fiona Groskreutz and Nicole Bond who worked tirelessly to ensure that this experiment was successful. We thank Dr Guntur Tanjung and George Sainsbury for rapid and accurate image analyses throughout the experiment. Additionally, we acknowledge the incredible assistance of Dr Krista Plett, Johanna Wong and Emi Stuart without whom the final harvest would have been far more arduous.

ReferencesTilmanD, CassmanKG, MatsonPA, NaylorR, PolaskyS. Agricultural sustainability and intensive production practices. Nature. 2002;418(6898):671 10.1038/nature01014 12167873MacLeod M, Eory V, Gruère G, Lankoski J. Cost-Effectiveness of Greenhouse Gas Mitigation Measures for Agriculture: A LITERATURE REVIEW. Paris: Organisation for Economic Cooperation and Development (OECD), 2015 Aug 2015. Report No.WhiteC, DupontS, HautauM, HartmanD, FinneyD, BradleyB, et al Managing the trade off between nitrogen supply and retention with cover crop mixtures. Agriculture, Ecosystems & Environment. 2017;237:121.StevensR, LaughlinR. Measurement of nitrous oxide and di-nitrogen emissions from agricultural soils. Nutrient Cycling in Agroecosystems. 1998;52(2–3):131–9.Van KauwenberghSJ. World phosphate rock reserves and resources: IFDC Muscle Shoals; 2010.CordellD, DrangertJ-O, WhiteS. The story of phosphorus: Global food security and food for thought. Global Environmental Change. 2009;19(2):292–305. 10.1016/j.gloenvcha.2008.10.009.CassmanKG. Ecological intensification of cereal production systems: yield potential, soil quality, and precision agriculture. Proceedings of the National Academy of Sciences. 1999;96(11):5952–9.10.1073/pnas.96.11.5952PMC3421110339523NicholsJD, RosemeyerME, CarpenterFL, KettlerJ. Intercropping legume trees with native timber trees rapidly restores cover to eroded tropical pasture without fertilization. Forest Ecology and Management. 2001;152(1):195–209. 10.1016/S0378-1127(00)00603-4.SapkotaTB, MajumdarK, JatML, KumarA, BishnoiDK, McDonaldAJ, et al Precision nutrient management in conservation agriculture based wheat production of Northwest India: Profitability, nutrient use efficiency and environmental footprint. Field Crops Research. 2014;155:233–44. 10.1016/j.fcr.2013.09.001.McAllisterCH, BeattyPH, GoodAG. Engineering nitrogen use efficient crop plants: the current status. Plant Biotechnology Journal. 2012;10(9):1011–25. 10.1111/j.1467-7652.2012.00700.x 22607381DucheneO, VianJ-F, CeletteF. Intercropping with legume for agroecological cropping systems: Complementarity and facilitation processes and the importance of soil microorganisms. A review. Agriculture, Ecosystems & Environment. 2017;240:148–61. 10.1016/j.agee.2017.02.019.ReichPB. The world-wide ‘fast–slow’ plant economics spectrum: a traits manifesto. Journal of Ecology. 2014;102(2):275–301. 10.1111/1365-2745.12211WrightIJ, ReichPB, WestobyM, AckerlyDD, et al The worldwide leaf economics spectrum. Nature. 2004;428(6985):821–7. 10.1038/nature02403 15103368ChapinFS. The Mineral Nutrition of Wild Plants. Annual Review of Ecology and Systematics. 1980;11:233–60.Del PozoA, GarnierE, AronsonJ. Contrasted nitrogen utilization in annual C 3 grass and legume crops: Physiological explorations and ecological considerations. Acta Oecologica. 2000;21(1):79–89. 10.1016/S1146-609X(00)00113-2BroderickGA. Desirable characteristics of forage legumes for improving protein utilization in ruminants. Journal of Animal Science. 1995;73(9):2760–73. 10.2527/1995.7392760x %J Journal of Animal Science. 8582869PeoplesMB, BrockwellJ, HerridgeDF, RochesterIJ, AlvesBJR, UrquiagaS, et al The contributions of nitrogen-fixing crop legumes to the productivity of agricultural systems. Symbiosis. 2009;48(1–3):1–17.PeoplesMB, ChalkPM, UnkovichMJ, BoddeyRM. Can differences in 15N natural abundance be used to quantify the transfer of nitrogen from legumes to neighbouring non-legume plant species? Soil Biology and Biochemistry. 2015;87:97–109. 10.1016/j.soilbio.2015.04.010.MasonNW, OrwinK, LambieS, WoodwardSL, McCreadyT, MudgePJE, et al Leaf economics spectrum–productivity relationships in intensively grazed pastures depend on dominant species identity. 2016;6(10):3079–91.10.1002/ece3.1964PMC482184127092237TilmanD. Resource competition and community structure: Princeton university press; 1982.7162524NyfelerD, Huguenin-ElieO, SuterM, FrossardE, LüscherA. Grass–legume mixtures can yield more nitrogen than legume pure stands due to mutual stimulation of nitrogen uptake from symbiotic and non-symbiotic sources. Agriculture, Ecosystems & Environment. 2011;140(1):155–63. 10.1016/j.agee.2010.11.022.SchmidB, HectorA, SahaP, LoreauMJJoPE. Biodiversity effects and transgressive overyielding. 2008;1(2):95–102.GravelD, BellT, BarberaC, CombeM, PommierT, MouquetNJNc. Phylogenetic constraints on ecosystem functioning. 2012;3:1117.10.1038/ncomms212323047675HooperDU, DukesJSJEL. Overyielding among plant functional groups in a long‐term experiment. 2004;7(2):95–105.LiL, ZhangQ, HuangD. A review of imaging techniques for plant phenotyping. Sensors (Basel, Switzerland). 2014;14(11):20078–111. 10.3390/s141120078 .25347588PMC4279472CavieresLA, BrookerRW, ButterfieldBJ, CookBJ, KikvidzeZ, LortieCJ, et al Facilitative plant interactions and climate simultaneously drive alpine plant diversity. 2014;17(2):193–202.10.1111/ele.1221724238015Nyfeler D, Huguenin-Elie O, Frossard E, Lüscher A, editors. Regulation of symbiotic nitrogen fixation in grass-clover mixtures. Sustainable grassland productivity: Proceedings of the 21st General Meeting of the European Grassland Federation, Badajoz, Spain, 3–6 April, 2006; 2006: Sociedad Española para el Estudio de los Pastos (SEEP).NyfelerD, Huguenin-ElieO, SuterM, FrossardE, ConnollyJ, LüscherA. Strong mixture effects among four species in fertilized agricultural grassland led to persistent and consistent transgressive overyielding. 2009;46(3):683–91. 10.1111/j.1365-2664.2009.01653.xHuntR, CaustonD, ShipleyB, AskewAJAob. A modern tool for classical plant growth analysis. 2002;90(4):485–8.10.1093/aob/mcf214PMC424038012324272ÅgrenGI, WetterstedtJÅM, BillbergerMFK. Nutrient limitation on terrestrial plant growth–modeling the interaction between nitrogen and phosphorus. New Phytologist. 2012;194(4):953–60. 10.1111/j.1469-8137.2012.04116.x 22458659Marsden J, Martin G, Parham D, Smith TR, Johnston B. Returns on Australian agricultural research. The joint Industries Assistance Commission-CSIRO benefit-cost study of the CSIRO Division of Entomology. Returns on Australian agricultural research The joint Industries Assistance Commission-CSIRO benefit-cost study of the CSIRO Division of Entomology. 1980.EvensonRE. Economic impacts of agricultural research and extension. Handbook of agricultural economics. 2001;1:573–628.GraceJB, KeoughJ, GuntenspergenGRJO. Size bias in traditional analyses of substitutive competition experiments. 1992;90(3):429–34.10.1007/BF0031770228313532HoffmannWA, PoorterH. Avoiding bias in calculations of relative growth rate. Annals of botany. 2002;90(1):37–42. 10.1093/aob/mcf140 .12125771PMC4233846ArausJL, CairnsJE. Field high-throughput phenotyping: the new crop breeding frontier. Trends in plant science. 2014;19(1):52–61. 10.1016/j.tplants.2013.09.008 24139902FahlgrenN, GehanMA, BaxterI. Lights, camera, action: high-throughput plant phenotyping is ready for a close-up. Current opinion in plant biology. 2015;24:93–9. 10.1016/j.pbi.2015.02.006 25733069BergerB, de RegtB, TesterM. High-throughput phenotyping of plant shoots High-throughput phenotyping in plants: Springer; 2012 p. 9–20.10.1007/978-1-61779-995-2_222893282Facility APP. Plant phenotyping: Australian Plant Phenomics Facility [cited 2019 4 April]. www.plantphenomics.com.SchlemmerMR, FrancisDD, ShanahanJ, SchepersJS. Remotely measuring chlorophyll content in corn leaves with differing nitrogen levels and relative water content. Agronomy journal. 2005;97(1):106–12.De VylderJ, VandenbusscheF, HuY, PhilipsW, Van Der StraetenD. Rosette tracker: an open source image analysis tool for automatic quantification of genotype effects. Plant physiology. 2012;160(3):1149–59. 10.1104/pp.112.202762 22942389PMC3490612WhitakerD, WilliamsER, JohnJA. CycDesigN: a package for the computer generation of experimental designs. 2.0 ed Canberra, Australia: CSIRO; 2002.Brien CJ. dae: functions useful in the design and ANOVA of experiments. R package version 3.0–32 2019 [June 17, 2019]. https://cran.at.r-project.org/package=dae.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2019.SchombergH, WeaverRJAj. Nodulation, nitrogen fixation, and early growth of arrowleaf clover in response to root temperature and starter nitrogen. 1992;84(6):1046–50.EversGW. Seedling Growth and Nodulation of Arrowleaf, Crimson, and Subterranean Clovers1. 1982;74(4):629–32. 10.2134/agronj1982.00021962007400040010xOlsenSR. Estimation of available phosphorus in soils by extraction with sodium bicarbonate: United States Department Of Agriculture; Washington; 1954.CrousKY, ÓsvaldssonA, EllsworthDS. Is phosphorus limiting in a mature Eucalyptus woodland? Phosphorus fertilisation stimulates stem growth. Plant and Soil. 2015;391(1):293–305. 10.1007/s11104-015-2426-4BrienC, JewellN, GarnettT, Watts-WilliamsSJ, BergerB. Smoothing and extraction of traits in the growth analysis of noninvasive phenotypic data. Plant Methods. 2020;16:36 Epub 10th March 2020. 10.1186/s13007-020-00577-6. 32180825PMC7065360Brien CJ. growthPheno: plotting, smoothing and growth trait extraction for longitudinal data. R package version 1.0–22 2020 [July 2, 2020]. https://cran.at.r-project.org/package=growthPheno.Al-TamimiN, BrienC, OakeyH, BergerB, SaadeS, HoYS, et al Salinity tolerance loci revealed in rice using high-throughput non-invasive phenotyping. Nature Communications. 2016;7:11 10.1038/ncomms13342 27853175PMC5118543Butler D, Cullis B, Gilmour A, Gogel B. asreml: asreml () fits the linear mixed model. R package version 3.0. VSNi; 2009.Brien CJ. asremlPlus: augments ASReml-R in fitting mixed models and packages generally in exploring prediction differences. R package version 4.1–28 2019 [June 17, 2019]. https://cran.at.r-project.org/package=asremlPlus.Hooper DUJE. The role of complementarity and competition in ecosystem responses to variation in plant diversity. 1998;79(2):704–19.LenthRV. Least-Squares Means: The R Package lsmeans. Journal of Statistical Software. 2016;69(1):33 Epub 2016-01-29. 10.18637/jss.v069.i01FranklinSB, GibsonDJ, RobertsonPA, PohlmannJT, FralishJS. Parallel Analysis: a method for determining significant principal components. 1995;6(1):99–106. 10.2307/3236261BryantFB, YarnoldPR. Principal-components analysis and exploratory and confirmatory factor analysis. 1995.MendozaR, GarcíaI, DeplamaD, LópezCF. Competition and growth of a grass–legume mixture fertilised with nitrogen and phosphorus: effect on nutrient acquisition, root morphology and symbiosis with soil microorganisms. Crop and Pasture Science. 2016;67(6):629–40.LedgardSF, SteeleKWSoilJP. Biological nitrogen fixation in mixed legume/grass pastures. 1992;141(1):137–53. 10.1007/bf00011314MatzekV, VitousekPMJEL. N: P stoichiometry and protein: RNA ratios in vascular plants: an evaluation of the growth‐rate hypothesis. 2009;12(8):765–71.10.1111/j.1461-0248.2009.01310.x19392715Whitehead DCJBCBoP, Crops F. The role of nitrogen in grassland productivity. A review of information from temperate regions. 1970;(48).AgnusdeiMG, AssueroSG, LattanziFA, MarinoMA. Critical N concentration can vary with growth conditions in forage grasses: implications for plant N status assessment and N deficiency diagnosis. Nutrient Cycling in Agroecosystems. 2010;88(2):215–30.BijelićZ, TomićZ, Ružic-MuslićD, KrnjajaV, MandićV, VučkovićS, et al Forage quality and energy content of perennial legume-grass mixtures at three level of N fertilization. Biotechnology in Animal Husbandry. 2014;30(3):539–47. 10.2298/BAH1403539BRanellsNN, WaggerMGJAJ. Grass-legume bicultures as winter annual cover crops. 1997;89(4):659–65.VitousekPM, MengeDNL, ReedSC, ClevelandCC. Biological nitrogen fixation: rates, patterns and ecological controls in terrestrial ecosystems. Philos Trans R Soc Lond B Biol Sci. 2013;368(1621):20130119-. 10.1098/rstb.2013.0119 23713117PMC3682739RobinsonD, RorisonIH. Plasticity in Grass Species in Relation to Nitrogen Supply. Functional Ecology. 1988;2(2):249–57. 10.2307/2389701GrimeJ, HuntR. Relative growth-rate: its range and adaptive significance in a local flora. The Journal of Ecology. 1975:393–422.DiazS, HodgsonJG, ThompsonK, CabidoM, CornelissenJHC, JaliliA, et al The plant traits that drive ecosystems: Evidence from three continents. 2004;15(3):295–304. 10.1111/j.1654-1103.2004.tb02266.xGrimeJP. The C-S-R model of primary plant strategies—origins, implications and tests In: GottliebLD, JainSK, editors. Plant Evolutionary Biology. Dordrecht: Springer Netherlands; 1988 p. 371–93.MainTM, DobberfuhlDR, ElserJJ. N: P stoichiometry and ontogeny of crustacean zooplankton: A test of the growth rate hypothesis. Limnology and Oceanography. 1997;42(6):1474–8. 10.4319/lo.1997.42.6.1474SchulzeJ. How are nitrogen fixation rates regulated in legumes? 2004;167(2):125–37. 10.1002/jpln.200320358SuterM, ConnollyJ, FinnJA, LogesR, KirwanL, SebastiàMT, et al Nitrogen yield advantage from grass–legume mixtures is robust over a wide range of legume proportions and environmental conditions. Global Change Biology. 2015;21(6):2424–38. 10.1111/gcb.12880 25626994BendigJ, BoltenA, BennertzS, BroscheitJ, EichfussS, BarethGJRs. Estimating biomass of barley using crop surface models (CSMs) derived from UAV-based RGB imaging. 2014;6(11):10395–412.GeY, BaiG, StoergerV, SchnableJC. Temporal dynamics of maize plant growth, water use, and leaf water content using automated high throughput RGB and hyperspectral imaging. Computers and Electronics in Agriculture. 2016;127:625–32. 10.1016/j.compag.2016.07.028.GrünerE, AstorT, WachendorfM. Biomass Prediction of Heterogeneous Temperate Grasslands Using an SfM Approach Based on UAV Imaging. 2019;9(2):54 10.3390/agronomy902005410.1371/journal.pone.0239673.r001Decision Letter 0PapaRobertoAcademic Editor© 2020 Roberto Papa2020Roberto PapaThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Submission Version0

25 Aug 2020

PONE-D-20-20899

High-throughput, image-based phenotyping reveals nutrient-dependent growth facilitation in a grass – legume mixture

PLOS ONE

Dear Dr. Ball,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 09 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Roberto Papa, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

'This project was supported in part by a Postgraduate Internship Award by the Australian Plant Phenomics Facility (APPF) awarded to Kirsten Ball. The APPF is funded by the Australian Government under the National Collaborative Research Infrastructure Strategy (NCRIS).'

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

'This project was supported in part by a Postgraduate Internship Award by the Australian Plant Phenomics Facility (APPF) awarded to KB.

https://www.plantphenomics.org.au/

Staff at the APPF (BB, CB, NJ) assisted in experimental design, statistical analyses and manuscript development.'

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

4. We note that you have included the phrase “data not shown” in your manuscript. Unfortunately, this does not meet our data sharing requirements. PLOS does not permit references to inaccessible data. We require that authors provide all relevant data within the paper, Supporting Information files, or in an acceptable, public repository. Please add a citation to support this phrase or upload the data that corresponds with these findings to a stable repository (such as Figshare or Dryad) and provide and URLs, DOIs, or accession numbers that may be used to access these data. Or, if the data are not a core part of the research being presented in your study, we ask that you remove the phrase that refers to these data.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: N/A

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: It was not possible to find a URL/accession or a number/DOIs where it was possible to download the data, unlike what was declared.

I would like to bring to the attention of the author some passages: row (64-65-66), reference is made to the types of interaction between plants suggesting that there was a sequence between them, which is not the case.

In line (102), reference is made to the objectives of the study using a past tense "were", usually a tense is used at present.

In row (182), reference is made to the addition of 99 mg N and 11 of P with a ratio of (6: 1), this ratio is incorrect.

In line (241, 242), reference is made to the methodology only through the numerical label of the bibliography, in my opinion I do not find it properly correct.

Furthermore, I would like to bring to the attention of the author the fact that in some cases the double quotation in the same sentence is separated by a comma (line 87), while in other cases it is separated by a hyphen (line 70).

Overall in my opinion it is well done and the concepts are well expressed and connected.

Reviewer #2: Overall, the manuscript is well done and highlights important issues, which could be used in further studies on mixtures and are applicable also on other species. Moreover, it underlines hight-throughtput phenotyping potential for studying crop growth patterns and facilitation in different conditions (i.e. nutrient availability) and its needs for careful calibration and validation. It only needs a minor revision in terms of sintax and lexical form. Result discussion is very well done and conclusions are strongly supported by data.

I put some advices, comments and corrections in the following lines.

17: delete “To our knowledge”

25-28: focus attention on results achieved in the mixture, rather than indicate species favourite nutritional conditions, that are largely known

53-56: please, better define this point, it’s not so clear

60: biological nitrogen fixation or biological N2 fixation

84: delete “Regardless of the measurement”

88: delete comma after “rapid”

98: delete “Regardless of the technique”

111 and 116: biological nitrogen fixation or biological N2 fixation

123: I suggest “Experimental design and plant material”; this section is not so clear; I put some suggestion in the following lines.

124-126: introduce better the experimental site, indicating GPS coordinates, specify that the experiment was conducted in a greenhouse using Lemnatec Imaging System and move the info about the images in section 2.6

127-130: it could be better to modify this sentence. I propose “The experiment investigated the effects of four fertiliser treatments derived by factorial combination of two levels of nitrogen (LN and HN) and two levels of phosphorus (LP and HP) on one grass (Phalaris aquatica) and one legume (Trifolium vesiculosum) pasture species grown in monoculture and mixture.”

130-131: it could be better to modify this sentence. I propose “The twelve treatments were arranged in a latinized, resolved incomplete-block design with ten replicates, for a total of 120 pots.”

139-143: I suggest to reorganize this part, adding in brackets the nutrient combinations (LNLP, HNHP, HNLP and LNHP) and deleting the reference to the experimental design, which you will move in the main text. I suggest: “Within each replicate (indicated by the solid red box) the four nutrient combinations (LNLP, HNHP, HNLP and LNHP) were assigned to 4 main units arranged in a grid according to the experimental design. The three species combinations [grass only (Gra), legume only (Leg) and mixture (Gra Leg)] were randomized to the three consecutive pots within each main unit. Colored blocks indicate 1 pot containing two halves.”

145: I suggest to rename this part in “Growing conditions”

146: please add a sentence to specify growth condition in general, in order to better introduce the experimental conditions and parameters description.

149-151: please, improve the sentence

174: I suggest “Fertilisation treatments”

178-179: please, adjust the sentence, including punctuation and brackets

182: ratio 9:1

182-183: adjust punctuation and brackets

187: indicate in which table number

361-362: add punctuation and adjust brackets

367: you can add the reference on equation 1

375, 387, 391, 403: correct the figure number

376, 380: I guess you should better introduce the use of ratios in the main text, maybe in the section 2.10

376-382: please, improve this part

393-400: caption is not so clear; you could improve it moving some informations in the main text and reorganizing the text in order to promote immediate understanding

422-424: explain in the caption text what Gra (Leg) and Leg (Gra) mean

443, 454: add a bracket after “shoot N”

456: in all this section, please add tables and figures references.

464: delete “in mixture”

479: add comma after “grasses”

528: “conclusion” should have a separate paragraph

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Papalini Simone

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

10.1371/journal.pone.0239673.r002Author response to Decision Letter 0Submission Version1

9 Sep 2020

Dear Dr Papa,

We enclose herewith our revised manuscript entitled ‘High-throughput, image-based phenotyping reveals nutrient-dependent growth facilitation in a grass – legume mixture’ for consideration for publication in the “Plant Phenomics and Precision Agriculture” special edition.

We have revised the manuscript accounting for PLOS ONE’s style requirements as requested and have provided repository information for the data used in the manuscript. At present this is a conditional DOI (10.25909/12895121) which will be linked to the article upon publication. Presently, the data can be accessed via this private link: https://figshare.com/s/99e05c190be6cb416164

The updated financial disclosure which has been removed from the acknowledgements section is as follows: “The Australian Plant Phenomics Facility received grant funding from the Australian Government through the National Collaborative Research Infrastructure Strategy (NCRIS). KB received a Postgraduate Internship Award from the Australian Plant Phenomics Facility towards the completion of this project”.

We wish to sincerely thank the reviewers for their valuable assistance in refining the manuscript for publication and provide responses to their comments herein. We hope that you find these responses satisfactory, and we look forward to positive news of our article’s acceptance.

Yours sincerely,

Kirsten Ball (on behalf of all co-authors)

 

Reviewer #1:

It was not possible to find a URL/accession or a number/DOIs where it was possible to download the data, unlike what was declared.

RESPONSE: Thank you for this comment, the associated data for the manuscript has now been placed in a repository and the provisional DOI is 10.25909/12895121, which once published will be linked to the article. Until the article is published, the data can be viewed via this private link: https://figshare.com/s/99e05c190be6cb416164

I would like to bring to the attention of the author some passages: row (64-65-66), reference is made to the types of interaction between plants suggesting that there was a sequence between them, which is not the case.

RESPONSE: Thank you for this suggestion. This syntax has now been changed to “Interspecific interactions between plants can be negative, in the case of competition for resources [22], neutral, where complementarity ensures that species do not compete for the same resources [11], and positive, where facilitation leads to higher performance of a species when grown in mixture than in monoculture [23]” removing the suggestion of sequence.

In line (102), reference is made to the objectives of the study using a past tense "were", usually a tense is used at present.

RESPONSE: Thank you. The word “were” was used not to denote tense but because there were multiple aims of the study, hence “the aims were….” We have now changed this to “This study used” to clarify.

In row (182), reference is made to the addition of 99 mg N and 11 of P with a ratio of (6: 1), this ratio is incorrect.

RESPONSE: Thank for pointing out this typographic error. It has now been rectified.

In line (241, 242), reference is made to the methodology only through the numerical label of the bibliography, in my opinion I do not find it properly correct.

RESPONSE: The method employed has now been specified in the initial sentence and the following description modified to emphasize the steps involved in the method. These are now lines 248-250.

Furthermore, I would like to bring to the attention of the author the fact that in some cases the double quotation in the same sentence is separated by a comma (line 87), while in other cases it is separated by a hyphen (line 70).

RESPONSE: Where there are only 2 references in the citation a comma is used, in the case where there are more than 2, a hyphen is used.

Overall in my opinion it is well done and the concepts are well expressed and connected.

Reviewer #2:

Overall, the manuscript is well done and highlights important issues, which could be used in further studies on mixtures and are applicable also on other species. Moreover, it underlines hight-throughtput phenotyping potential for studying crop growth patterns and facilitation in different conditions (i.e. nutrient availability) and its needs for careful calibration and validation. It only needs a minor revision in terms of sintax and lexical form. Result discussion is very well done and conclusions are strongly supported by data.

I put some advices, comments and corrections in the following lines.

17: delete “To our knowledge”

RESPONSE: Completed.

25-28: focus attention on results achieved in the mixture, rather than indicate species favourite nutritional conditions, that are largely known

RESPONSE: Thank you for this comment. We have removed the single-species favored nutrient conditions from these lines and included only the interactions that occurred in mixtures.

53-56: please, better define this point, it’s not so clear

RESPONSE: This has been altered to make the point clearer “Plant growth strategies involve trade-offs between increasing both productivity and resource acquisition (interpreted herein as an acquisitive growth strategy), and reducing productivity and conserving resources (interpreted herein as a conservative growth strategy) [14].”

60: biological nitrogen fixation or biological N2 fixation

RESPONSE: Completed

84: delete “Regardless of the measurement”

RESPONSE: Completed.

88: delete comma after “rapid”

RESPONSE: Completed

98: delete “Regardless of the technique”

RESPONSE: Completed

111 and 116: biological nitrogen fixation or biological N2 fixation

RESPONSE: Completed

123: I suggest “Experimental design and plant material”; this section is not so clear; I put some suggestion in the following lines.

RESPONSE: Thank you. We have now changed this to “Experimental design and species”

124-126: introduce better the experimental site, indicating GPS coordinates, specify that the experiment was conducted in a greenhouse using Lemnatec Imaging System and move the info about the images in section 2.6

RESPONSE: We have revised this sentence accordingly: “Our experiment was conducted between the 24th July 2018 and 5th October 2018 at the Australian Plant Phenomics Facility at the University of Adelaide (-34.971298, 138.639627) in a Lemnatec Imaging System”

127-130: it could be better to modify this sentence. I propose “The experiment investigated the effects of four fertiliser treatments derived by factorial combination of two levels of nitrogen (LN and HN) and two levels of phosphorus (LP and HP) on one grass (Phalaris aquatica) and one legume (Trifolium vesiculosum) pasture species grown in monoculture and mixture.”

RESPONSE: We have altered the sentence according to the reviewer’s suggestion.

130-131: it could be better to modify this sentence. I propose “The twelve treatments were arranged in a latinized, resolved incomplete-block design with ten replicates, for a total of 120 pots.”

RESPONSE: We have altered the sentence according to the reviewers suggestion.

139-143: I suggest to reorganize this part, adding in brackets the nutrient combinations (LNLP, HNHP, HNLP and LNHP) and deleting the reference to the experimental design, which you will move in the main text. I suggest: “Within each replicate (indicated by the solid red box) the four nutrient combinations (LNLP, HNHP, HNLP and LNHP) were assigned to 4 main units arranged in a grid according to the experimental design. The three species combinations [grass only (Gra), legume only (Leg) and mixture (Gra Leg)] were randomized to the three consecutive pots within each main unit. Colored blocks indicate 1 pot containing two halves.”

RESPONSE: We have altered the sentence according to the reviewer’s suggestion.

145: I suggest to rename this part in “Growing conditions”

RESPONSE: We have altered the header according to the reviewers suggestion.

146: please add a sentence to specify growth condition in general, in order to better introduce the experimental conditions and parameters description.

RESPONSE: Thank you or this suggestion however it is the authors belief that the methods section is already quite lengthy, and much attention has already been paid to demonstrating growth conditions using both text and figures. We have therefore not included more explanation as suggested.

149-151: please, improve the sentence

RESPONSE: We have now improved the sentence accordingly:” The plants were physically separated aboveground by a white plastic divider, set 5 cm into the soil and oriented north–south on the conveyor system in the greenhouse for consistency with respect to imaging orientation and solar exposure. The divider allowed intermingling of roots belowground (Fig 2).

174: I suggest “Fertilisation treatments”

RESPONSE: We have changed the header per the reviewer’s suggestion

178-179: please, adjust the sentence, including punctuation and brackets

RESPONSE: Thank you. We have now separated these sentences for the readers ease.

182: ratio 9:1

RESPONSE: Altered. Thank you.

182-183: adjust punctuation and brackets

RESPONSE: These alterations have been made, and brackets added around the units of measurement.

187: indicate in which table number

RESPONSE: Thank you for this suggestion however this information is not in tabular form but is present and clearly headed in the supplement.

361-362: add punctuation and adjust brackets

RESPONSE: These alterations have been made

367: you can add the reference on equation 1

RESPONSE: Thank you. This has been added.

375, 387, 391, 403: correct the figure number

RESPONSE: These have been corrected.

376, 380: I guess you should better introduce the use of ratios in the main text, maybe in the section 2.10

RESPONSE: We have now included this part in the section “Calculating overyielding from sPSA DAP 70 values”: We have altered the text accordingly: “Ymix is the observed yield of species i in mixture and Ymon is the observed yield of species i in monoculture expressed as a ratio (Ygra(Mix):Ygra and Yleg(Mix):Yleg).

376-382: please, improve this part

RESPONSE: We have broken down the sections and improved the structure accordingly: “At the whole pot level, productivity in mixtures increased in all treatments. On average, the LNHP treatment increased by 27%, followed by LNLP which increased 21%, then HNLP which increased 20% and finally HNHP which was 10% more productive in mixture”

393-400: caption is not so clear; you could improve it moving some informations in the main text and reorganizing the text in order to promote immediate understanding

RESPONSE: Thank you for this suggestion. We have now moved the larger proportion of the figure caption up into the section “Calculations of overyielding” and made the caption text clearer.

422-424: explain in the caption text what Gra (Leg) and Leg (Gra) mean

RESPONSE: This has now been added to the figure caption.

443, 454: add a bracket after “shoot N”

RESPONSE: Added. Thank you.

456: in all this section, please add tables and figures references.

RESPONSE: Thank you for this valuable suggestion, we have now added all table and figure references.

464: delete “in mixture”

RESPONSE:

479: add comma after “grasses”

RESPONSE: Altered.

528: “conclusion” should have a separate paragraph

RESPONSE: Thank you. This section is its own separated paragraph with header “Conclusion”

Submitted filename: PlosOne_Ball_Response to reviewers.docx

Click here for additional data file.

10.1371/journal.pone.0239673.r003Decision Letter 1PapaRobertoAcademic Editor© 2020 Roberto Papa2020Roberto PapaThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Submission Version1

11 Sep 2020

High-throughput, image-based phenotyping reveals nutrient-dependent growth facilitation in a grass – legume mixture

PONE-D-20-20899R1

Dear Dr. Ball,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Roberto Papa, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

10.1371/journal.pone.0239673.r004Acceptance letterPapaRobertoAcademic Editor© 2020 Roberto Papa2020Roberto PapaThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

28 Sep 2020

PONE-D-20-20899R1

High-throughput, image-based phenotyping reveals nutrient-dependent growth facilitation in a grass - legume mixture

Dear Dr. Ball:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Roberto Papa

Academic Editor

PLOS ONE