A global dataset for assessing nitrogen-related plant traits using drone imagery in major field crop species

ABSTRACT Enhancing rapid phenotyping for key plant traits, such as biomass and nitrogen content, is critical for effectively monitoring crop growth and maximizing yield. Studies have

explored the relationship between vegetation indices (VIs) and plant traits using drone imagery. However, there is a gap in the literature regarding data availability, accessible datasets.

Based on this context, we conducted a systematic review to retrieve relevant data worldwide on the state of the art in drone-based plant trait assessment. The final dataset consists of 41

peer-reviewed papers with 11,189 observations for 11 major crop species distributed across 13 countries. It focuses on the association of plant traits with VIs at different

growth/phenological stages. This dataset provides foundational knowledge on the key VIs to focus for phenotyping key plant traits. In addition, future updates to this dataset may include new

open datasets. Our goal is to continually update this dataset, encourage collaboration and data inclusion, and thereby facilitate a more rapid advance of phenotyping for critical plant

traits to increase yield gains over time. SIMILAR CONTENT BEING VIEWED BY OTHERS NEAR-REAL-TIME MODIS-DERIVED VEGETATION INDEX DATA PRODUCTS AND ONLINE SERVICES FOR CONUS BASED ON NASA LANCE

Article Open access 04 August 2022 ASSESSING THE EFFECTIVENESS OF VEGETATION INDICES IN DETECTING FOREST DISTURBANCES IN THE SOUTHEAST AMAZON Article Open access 08 November 2024 MAPPING

CANOPY TRAITS OVER QUÉBEC USING AIRBORNE AND SPACEBORNE IMAGING SPECTROSCOPY Article Open access 11 October 2023 BACKGROUND & SUMMARY Agriculture is an important industry, serving as the

foundation of food security and of the global economy1. The complexity of biological systems is reflected in the spatial temporal variability of the soil and crop N status within a field2.

To ensure optimal use of outputs, fertilizers should be provided at the right time, place, with an adequate source and at the right rate, only when necessary3. Therefore, a variable nitrogen

(N) management strategy must be implemented to optimize fertilizer N rates, economic benefits, and maintaining or increasing both yield and quality4. N is a critical element for crop growth

and one of the most important nutrients in agriculture to improve crop yield and for protein formation5. Furthermore, the utilization of the right fertilizer N rate is crucial not only to

increase yields but to reduce the environmental footprint of this practice3,6. Traditional methods for detecting crop N status involve time-consuming field sampling and costly laboratory

analysis7. Monitoring crop N status efficiently and effectively remains an urgent problem to be solved8,9. In recent years, technological innovations based on the utilization of

multispectral and hyperspectral sensors mounted in different platforms help to provide critical imagery data for phenotyping and developing new tools for precision agriculture10. The

emergence of unmanned aerial vehicles (UAV, or commonly known as drone) has advanced remote sensing applications at fine scales. UAV have gained significant scientific and public interest,

due to their flexibility, easiness to use, and affordability11,12. The aerial platform and sensor cost with a rapid image availability make this equipment valuable for assessing critical

plant traits for advancing yield gains13,14. Characterization of key plant traits can vary depending on the crop and growth stage. Many efforts have been dedicated to identifying VIs that

best correlates to plant traits15,16,17,18. The most relevant linked to crop N status include leaf N content, leaf N concentration (LNC), plant N concentration (PNC), N nutrition index

(NNI), and N concentration (NC) for different plant fractions19,20,21,22,23,24. Several crops have been investigated using drone technology to assess plant traits, including but not limited

to wheat (_Triticum aestivum_ L.)15, corn (_Zea mays_ L.)16, rice (_Oryza sativa_ L.)17, and barley _(Hordeum vulgare_)18. Different crops may require species-specific VIs to better

characterize crop N status, as differences in leaf structure, canopy architecture, N allocation, and phenological stage should be taken into account when comparing across them4,25,26. In

addition, other factors such as soil exposure, crop residues, and N application levels can also affect the stability of an index27, provide restrictions to use a more universal index for

accurately estimate similar plant traits across crop species11. For example, NDVI has been found to be a reliable index for N estimation in corn but less effective for rice28,29,30,31.

Therefore, it is important to identify and evaluate the most effective VIs more directly targeting specific plant traits across major field crops. A critical challenge as technology is

evolving and the number of published studies on this topic grows exponentially with time is to keep up with the current progress and identify research knowledge gaps. Furthermore, as the

analysis is based on a single experiment with N treatments and small plots, further research is needed to translate current findings to real-world scenarios32. Most studies using UAV assess

nutrient content using a simple regression model, typically linear models14, and usually focused on a few plant traits. To date, the existing academic literature on the merger of studies

utilizing the same VIs and plant traits is limited. Therefore, developing a more organized and structured review can help identify promising VIs and plant traits while developing an open

dataset to assist future progress on this topic. It is acknowledged that variances in plant traits accuracies exist between studies33,34. These discrepancies often arise from differences

such as geographical location, types of drones and camera sensors, and the application of signal processing techniques (multivariate linear methods, (e.g., partial least squares regression,

stepwise multiple linear regression, and multiple linear regression), multivariate non-linear methods (e.g., random forest and support vector machine), and univariate methods (e.g. linear

regression). As most studies using UAV assess nutrient content using univariate methods14, typically linear or non-linear regressions, we focused our study on gathering information on

studies that used VIs to predict any N-related trait and/or yield. Following this rationale, a systematic review process focusing on retrieving datasets on the state of the art in

drone-based plant traits assessment was executed. Our global dataset focuses on major field crop species, 11 total, retrieved from studies published during the last two decades (2000 to

2023) in 13 countries. The final dataset contains 41 peer-reviewed scientific manuscripts focusing on the relationship between VIs and plant traits for characterizing crop N status and

identifying knowledge gaps to guide future research on drone-based plant traits assessments. METHODS A literature search was conducted, involving identification, screening, eligibility, and

inclusion of relevant records (Fig. 1a). The Scopus and Web of Science search engines were the main data sources. The keywords “multispectral airborne images” or “drone” or “UAV” or “UAS” or

“unmanned aerial vehicle” or “remotely piloted aircraft system”, AND “nitrogen” AND “yield” were included in the search criteria, restricting the duration from 2000 to March 2023 to

identify the most promising modern technologies. After retrieving all relevant records (number of studies, n = 372), a first screening process was performed to remove duplicates (n = 264)

for further processing. As a next step, an intensive analysis/full text reading was executed. Studies presenting the following criteria were excluded of the final dataset: 1) languages other

than English, 2) unavailability of full-text publication, 3) lack of focus on field experiments (other setting greenhouse, growth chamber, pots, etc.), 4) articles that did not use drones

to collect RGB, multi- or hyperspectral images, 5) used more than one crop/plant mixed (not focus on a single cropping specie), 6) study not focused on plant N/yield association, 7) lack of

observed N determinations (excluding indirect N measurements such as chlorophyll meters, handheld sensors), and lastly, 8) those studies only benchmarking UAV derived imagery data with

handheld sensors. In the next step, full-text screening was performed to exclude studies that did not report data on VIs and plant trait, removing 222 papers. An additional 23 papers were

identified and reviewed by examining citations from the remaining manuscripts, resulting in the inclusion of 22 additional studies. These studies were checked for duplicates. As a result, a

final database comprises 41 articles published between 2010 to 2023 (Fig. 1b). A total of 41 records were identified fulfilling the main criterion of crop N estimation using different crops

with RGB, multispectral, and hyperspectral data derived from the utilization of drones. The data retrieved from each paper included: i) geographic location of the experiments, ii) crop

species, iii) plant traits (N content/concentration, N nutrition index (NNI), N uptake, leaf/plant N accumulation, canopy N content, N rate, and biomass), iv) VIs/bands, v) applied N rates,

and vi) timing of UAV flights/phenological stages (further details presented in Table 1). For each article presented in Table 1, all available information on VIs and plant traits from

figures, tables, text, and supplementary material for figshare repository was extracted using the ‘juicr’ R package35. Also, the data were visually inspected to ensure the information was

associated with plant development stage. Among the 14 plant traits identified during data extraction, only two plant traits (i.e., relative yield and N uptake) could be combined, considering

crop type (i.e., wheat and cotton), VIs (i.e., NDVI and NDRE), and phenological stage3,15,33,34,36,37. DATA COLLECTION FOR META-ANALYSIS To explore the predictive abilities of drones in

estimating agricultural traits, we undertook a meta-analysis encompassing 41 selected studies. This meta-analysis aimed to evaluate the potential of UAVs in estimating yield and

nitrogen-related plant traits, with an approach that does not prioritize any VI. N-related traits (plant N density, plant N content, plant N concentration, plant N accumulation, NNI, N

uptake, leaf N density, leaf N content, leaf N concentration, leaf N accumulation, canopy N content) were merged into a single category labeled “nitrogen” for simplification. This

preprocessing step ensured consistency and clarity in trait categorization. For each trait of interest (nitrogen and yield), we created individual plots. Within each plot, we iterated over

growth stages and crops to calculate Fisher’s Z transformation effect sizes along with their 95% confidence intervals. This transformation converts the R² values into a metric that

approximates a normal distribution, thereby making it more suitable for our analytic model. Fisher’s Z transformation was computed using Eq. (1):

$$Z=\frac{1}{2}ln\left(\frac{1+r}{1-r}\right)$$ (1) where r represents the Pearson correlation coefficient, which was derived from the R² values provided in the dataset (supplementary

material for figshare repository). The mean Fisher’s Z value and standard error were calculated for each group, and the error bars were plotted accordingly. Analysis was conducted separately

for each growth stage and crop, facilitating comparative evaluations. When assessing the accuracy of plant trait estimations, we prioritize R² as our main metric due to its broad

acceptance, straightforward interpretation, and most used metric compared to others. Two types of regression analyses were performed to explore moderator effects: crop moderator analysis and

growth stage moderator analysis. Ordinary Least Squares regression models were fitted to assess the influence of crop type and growth stage on R² values. One-hot encoding was applied to

categorical crop variables and growth stages, with coefficients, standard errors, and p-values extracted to quantify the impact of individual crops and the role of different growth stages in

trait prediction. ANOVA was conducted to evaluate the significance of moderator effects, both for crop type and growth stage, on trait variability. F-values were computed for the entire

sets of crops and growth stages, providing insights into the overall impact of these moderators on model fit. Data preprocessing, analysis, and visualization were performed using the Python

programming language, leveraging libraries such as “Pandas”, “NumPy”, “Matplotlib”, and “Statsmodels”. These tools facilitated efficient data manipulation, statistical modeling, and

graphical representation of results. To standardize the data monitoring period across all studies, we converted the reported growth stages to the BBCH scale38, a very known scale for

phenological staging. We categorized the growth stages as follows: early (BBCH 0–30), mid (BBCH 31–60), and late stage (BBCH 61–90). We also considered the entire growth period – all (BBCH

0–90) as a separate category. These categorizations were employed to assess the impact of different growth stages on the accuracy of N-related traits and yield prediction in major crops.

DATA RECORDS The data are accessible on the figshare repository39, available at https://doi.org/10.6084/m9.figshare.22938797, and includes the following files: * 1. “Dataset.xlsx” includes

the data. It contains three tabs: “UAV_dataset”, “Sensor and processing info”, and “Quantitatively analysis”. * 2. “Summary of the dataset.docx”, includes a summary of the dataset excel file

(UAV_dataset tab), defining each column, data extracted from the studies, the units for each variable when pertinent, and a definition for each variable. * 3. “Figure2_N_Uptake.r”, includes

the code to reproduce Fig. 2. * 4. “Figure3_effect_size.ipynb”, includes the code to build Fig. 3. * 5. “Figure4_RelativeYield_NDVI.r”, includes the code to run Fig. 4. The “Dataset.xlsx”

file (UAV_dataset tab) contains all the information collected on this systematic analysis. The “Summary of the dataset.docx” presents a description for each column of the “UAV_dataset” tab

with the information separated into three categories: Category I, general specification of the study, containing information for author and publication year, and paper identification for

each study included in the dataset. Category II, experiment information, describing species, VI used, VI value, coefficient of determination (R²), root mean square error (RMSE), and

phenological stage sampling moment or dates. Category III, key for the dataset related to plant traits used. All plant traits information is reported with their units, as expressed in the

data collected from those respective studies. This category shows the amount of N rate applied, plant/leaf N concentration/content, N nutrition index, yield, relative yield, N uptake,

leaf/plant N density, leaf/plant N accumulation, canopy N content, and the aboveground biomass values. The “Dataset.xlsx” file (“Sensor and processing info” tab) describes topics related to

sensor and processing information, including sensor band, spatial resolution, UAV flight height, plot size, VIs procedure, calibration of the sensor, weather/field condition, soil texture,

latitude/longitude, and year of experiment for each of the 41 selected studies. The “Dataset.xlsx” file (“Quantitatively analysis” tab) describes the study number, plant traits, R² metric,

sample size for each trait (N), growth stage (BBCH scale), and field crop extracted. This tab was used for meta-analysis process. Table 1 describes the main topics of the 41 selected

studies, including species, country for the study location, author, and year of publication, phenological stage sampling moment, plant traits and VIs utilized for each study, and relevant

keyword for the study. Table 2 describes the regression models with one moderator (crop or growth stage) for nitrogen and yield plant traits. This table investigates the predictive

capabilities of drones in estimating agricultural traits without focusing on specific VI. TECHNICAL VALIDATION To demonstrate the value of the dataset, the relationship between VIs and plant

traits was investigated. After constructing the dataset, we checked for potential outliers and carefully summarized the information to analyze the interaction of multiple studies with the

goal of merging them. Data of N uptake and two VIs (NDRE and NDVI) are presented in Fig. 2 for three studies across stages and crops (cotton and wheat). These three studies were conducted in

Switzerland, Australia, and the United States for wheat crop at three stages (tillering, stem elongation, and heading) and then for cotton crop at three stages (first flower, first cracked

boll, and maturity). The N uptake (kg N ha−1) was calculated by multiplying the dry matter biomass (kg ha−1) with the corresponding N concentration (%) of the plant sample40. Although these

two crops are managed differently with respect to crop management such as time of nutrient application, the overall trend was similar between VIs and the key plant N trait identified for

these crop species. In contrast to NDRE, NDVI saturated shortly after stem elongation before decreasing rapidly during the senescence phase1 when studies were combined (saturation point when

NDVI ≥ 0.5). The NDVI seems to be a viable N status indicator for a first N application, when the crop leaf canopy has not closed yet. The NDRE progress was linear until the stage of spike

emergence, which takes place after the fertilizer application in winter wheat. Thus, it is plausible that NDRE is a better VI for the creation of fertilizer prescription maps and N uptake

assessment than NDVI. Moreover, the correlation value for NDRE (R2 = 0.80 and 0.67 for wheat and cotton, respectively) confirmed the assumption that NDRE can be used to monitor the N status

of the wheat and cotton crops. Lastly, these studies provide initial evidence of the potential superiority of red-edge-NIR based spectral indices over NDVI (R2 between 0.11–0.65 across

crops)40,41. In addition, this dataset helps to demonstrate the need of expanding the exploration of other spectral bands to target specific plant N traits more directly. Data collected can

also be used to predict N-related traits and yield for the given field crop and growth stage, merging all the VIs used in each paper. For N-related traits, the F-value for the entire crop

set and growth stage was 4.91 and 16.79, respectively (Table 2). The F-values for the entire growth stage set are notably higher than those for the entire crop set, suggesting that growth

stage may have a stronger influence on these traits compared to crop type. However, for yield, both the F-values for entire crop sets and entire growth stage sets are relatively low,

indicating that neither crop type nor growth stage may have a significant impact on yield in this dataset. The coefficients and associated p-values indicate that different crops have

significant effects on N content and yield. For instance, crops like cotton, grass, oilseed rape, rice, soybean, and winter wheat show significant effects on N-related traits, with p-values

indicating a positive relationship between these crops and higher R² values for N-related vegetation indices. Similarly, barley, corn, rice, and winter wheat show significant positive

coefficients (p < 0.05), indicating their positive influence on R² values for yield-related vegetation indices, highlighting their suitability for accurate yield prediction using remote

sensing data. Notably, the mid and late growth stages further enhanced the model predictive capability for the N trait estimation, indicating their positive influence on R² values for

N-related vegetation indices derived during these stages. Compared with all growth stage (Estimate = 0.344, SE = 0.047, p < 0.001), both mid stage (Estimate = 0.568, SE = 0.014, p < 

0.001) and late stage (Estimate = 0.585, SE = 0.030, p < 0.001) showed a more significant influence, reflecting the relevance of aiming for a specific growth stage to estimate N during

the plant development (Table 2). While our dataset lacked early growth stage data based on the BBCH scale, the absence of this information underscores the importance of early-stage data.

This indicates that the initial assimilation of N strongly influences subsequent plant productivity2,37,42. In the context of yield, it was unexpected to observe that the early stage showed

a high estimate accuracy (Estimate = 0.754, SE = 0.064, p < 0.001), followed by the late stage (Estimate = 0.717, SE = 0.166, p < 0.001) with a higher standard error. The higher

estimation accuracy for yield at the early growth stages may be attributed to the distinct spectral signatures captured by VIs where unique phenological signatures are assessed, indicating

rapid vegetative growth or early stressors, leading to higher accuracy in yield estimation. Results also reveal the uncertainty (reflected as the length of the 95% credibility interval) is

higher for the late growth stage yield prediction (Fig. 3). The level of uncertainty depends on the number of observations within a study and on the total number of studies for a growth

stage. We found only three studies that used the late stage to estimate yield2,15,32. When the number of data is small, the determination of yield can produce estimates with large

uncertainty (wide credibility intervals). USAGE NOTES This dataset can also be used in studies to diagnose N status and various plant traits in different crop species using UAV imagery. For

example, recent studies have used NDVI as a yield predictor for wheat3,15,37,43. However, when combining data from these studies conducted across different environments, it is not possible

to gain insights about the relationship between relative yield and NDVI (Fig. 4). It is noteworthy to understand that different growth stages will present varying conditions, which indicates

the need to properly report crop phenology (growth stage) and environmental conditions (rainfed vs. irrigated) when correlating yield with any VI44. For instance, some studies have

highlighted the importance of obtaining an estimation of crop biomass in reducing variability/noise when exploring crop N status45. This approach could lead to more reliable models and the

development of more universal N management guidelines. Despite considerable progress, there are still many relevant research knowledge gaps in drone-based crop research. Many studies do not

use the same VIs to analyze a specific plant trait and/or phenological stage. In addition, numerous studies provided metrics (R² and/or RMSE) of the relationship between plant trait and VI,

but often omitted the corresponding data, restricting the future use of those studies. Additionally, the dataset could be expanded to include other plant traits such as other nutrient

deficiencies (e.g., potassium), drought status, and pest and disease detection. Drone-based imagery data can help detect changes in crop N status early in the season, permitting to adjust

via interventions. Improving the ability to more precisely and dynamically correct crop N deficiencies will help farmers focus on a more sustainable approach to monitor large areas in a

short period of time, improving farming profitability and reducing the environmental footprint. CODE AVAILABILITY Scripts using R and python programming languages are provided to produce

figures. Additional code and related files are available at figshare repository39. REFERENCES * Yu, J., Wang, J. & Leblon, B. Evaluation of soil properties, topographic metrics, plant

height, and unmanned aerial vehicle multispectral imagery using machine learning methods to estimate canopy nitrogen weight in corn. _Remote Sens_. 13, (2021). * Zhang, J. _et al_. In-season

variable rate nitrogen recommendation for wheat precision production supported by fixed-wing UAV imagery. _Precis. Agric._ 23, 830–853 (2022). Article  MathSciNet  CAS  Google Scholar  *

Benincasa, P. _et al_. Reliability of ndvi derived by high resolution satellite and uav compared to in-field methods for the evaluation of early crop n status and grain yield in Wheat. _Exp.

Agric._ 54, 604–622 (2017). Article  Google Scholar  * Han, S. _et al_. Monitoring Key Wheat Growth Variables by Integrating Phenology and UAV Multispectral Imagery Data into Random Forest

Model. _Remote Sens_. 14 (2022). * Javed, T. _et al_. Recent Advances in Agronomic and Physio-Molecular Approaches for Improving Nitrogen Use Efficiency in Crop Plants. _Front. Plant Sci._

13, 1–21 (2022). Article  Google Scholar  * Arroyo, J. A. _et al_. Assessing nitrogen nutrition in corn crops with airborne multispectral sensors. _Lect. Notes Comput. Sci. (including

Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics)_ 10351 LNCS, 259–267 (2017). Google Scholar  * Zheng, H. _et al_. Combining unmanned aerial vehicle (UAV)-based multispectral

imagery and ground-based hyperspectral data for plant nitrogen concentration estimation in rice. _Front. Plant Sci._ 9, 1–13 (2018). Article  ADS  Google Scholar  * Li, M. _et al_. Retrieval

of Nitrogen Content in Apple Canopy Based on Unmanned Aerial Vehicle Hyperspectral Images Using a Modified Correlation Coefficient Method. _Sustain_. 14 (2022). * Wang, L. _et al_.

Estimation of paddy rice nitrogen content and accumulation both at leaf and plant levels from uav hyperspectral imagery. _Remote Sens._ 13, 1–21 (2021). Google Scholar  * Näsi, R. _et al_.

Estimating biomass and nitrogen amount of barley and grass using UAV and aircraft based spectral and photogrammetric 3D features. _Remote Sens._ 10, 1–32 (2018). Article  Google Scholar  *

Chen, P. & Wang, F. Effect of crop spectra purification on plant nitrogen concentration estimations performed using high-spatial-resolution images obtained with unmanned aerial vehicles.

_F. Crop. Res._ 288, 108708 (2022). Article  Google Scholar  * Lebourgeois, V., Bégué, A., Labbé, S., Houlès, M. & Martiné, J. F. A light-weight multi-spectral aerial imaging system for

nitrogen crop monitoring. _Precis. Agric._ 13, 525–541 (2012). Article  Google Scholar  * Xiang, H. & Tian, L. An automated stand-alone in-field remote sensing system (SIRSS) for

in-season crop monitoring. _Comput. Electron. Agric._ 78, 1–8 (2011). Article  Google Scholar  * Barbedo, J. G. A. A review on the use of unmanned aerial vehicles and imaging sensors for

monitoring and assessing plant stresses. _Drones_ 3, 1–27 (2019). Article  Google Scholar  * Almeida-Ñauñay, A. F. _et al_. Optimization of soil background removal to improve the prediction

of wheat traits with UAV imagery. _Comput. Electron. Agric_. 205 (2023). * Cai, Y. _et al_. Detecting In-Season Crop Nitrogen Stress of Corn for Field Trials Using UAV-and CubeSat-Based

Multispectral Sensing. _IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens._ 12, 5153–5166 (2019). Article  ADS  Google Scholar  * Li, J., Zhang, F., Qian, X., Zhu, Y. & Shen, G.

Quantification of rice canopy nitrogen balance index with digital imagery from unmanned aerial vehicle. _Remote Sens. Lett._ 6, 183–189 (2015). Article  Google Scholar  * Kefauver, S. C. _et

al_. Comparative UAV and field phenotyping to assess yield and nitrogen use efficiency in hybrid and conventional barley. _Front. Plant Sci._ 8, 1–15 (2017). Article  Google Scholar  * Kou,

J. _et al_. Predicting Leaf Nitrogen Content in Cotton with UAV RGB Images. _Sustain_. 14 (2022). * Oliveira, R. A. _et al_. Machine learning estimators for the quantity and quality of

grass swards used for silage production using drone-based imaging spectrometry and photogrammetry. _Remote Sens. Environ._ 246, 111830 (2020). Article  Google Scholar  * Heinemann, P. &

Schmidhalter, U. Spectral assessments of N-related maize traits: Evaluating and defining agronomic relevant detection limits. _F. Crop. Res._ 289, 108710 (2022). Article  Google Scholar  *

Qiu, Z. _et al_. Estimation of nitrogen nutrition index in rice from UAV RGB images coupled with machine learning algorithms. _Comput. Electron. Agric._ 189, 106421 (2021). Article  Google

Scholar  * Gabriel, J. L. _et al_. Airborne and ground level sensors for monitoring nitrogen status in a maize crop. _Biosyst. Eng._ 160, 124–133 (2017). Article  Google Scholar  * Li, X.

_et al_. Sugarcane Nitrogen Concentration and Irrigation Level Prediction Based on UAV Multispectral Imagery. _Sensors_ 22 (2022). * Richardson, A. D., Duigan, S. P. & Berlyn, G. P. An

evaluation of noninvasive methods to estimate foliar chlorophyll content. _New Phytol._ 153, 185–194 (2002). Article  CAS  Google Scholar  * Gitelson, A. A., Gritz, Y. & Merzlyak, M. N.

Relationships between leaf chlorophyll content and spectral reflectance and algorithms for non-destructive chlorophyll assessment in higher plant leaves. _J. Plant Physiol._ 160, 271–282

(2003). Article  CAS  PubMed  Google Scholar  * Yin, C. _et al_. Hyperspectral UAV Images at Different Altitudes for Monitoring the Leaf Nitrogen Content in Cotton Crops. _Remote Sens._ 14,

1–19 (2022). Article  Google Scholar  * Lee, H., Wang, J. & Leblon, B. Using linear regression, random forests, and support vector machine with unmanned aerial vehicle multispectral

images to predict canopy nitrogen weight in corn. _Remote Sens_. 12 (2020). * Lee, H., Wang, J. & Leblon, B. Intra-Field Canopy Nitrogen Retrieval from Unmanned Aerial Vehicle Imagery

for Wheat and Corn Fields. _Can. J. Remote Sens._ 0, 454–472 (2020). Article  Google Scholar  * Xu, X. _et al_. Estimating leaf nitrogen content in corn based on information fusion of

multiple-sensor imagery from uav. _Remote Sens._ 13, 1–17 (2021). Article  ADS  Google Scholar  * Li, G. S. _et al_. Prediction of plant nutrition state of rice under water-saving

cultivation and panicle fertilization application decision making. _Agronomy_ 11 (2021). * Geipel, J., Link, J., Wirwahn, J. A. & Claupein, W. A programmable aerial multispectral camera

system for in-season crop biomass and nitrogen content estimation. _Agric._ 6, 1–19 (2016). Google Scholar  * Argento, F. _et al_. Site-specific nitrogen management in winter wheat supported

by low-altitude remote sensing and soil data. _Precis. Agric._ 22, 364–386 (2020). Article  Google Scholar  * Fu, Y. _et al_. Winter wheat nitrogen status estimation using uav-based rgb

imagery and gaussian processes regression. _Remote Sens._ 12, 1–27 (2020). Article  Google Scholar  * Lajeunesse, M. J. _Automated, semi-automated, and manual extraction of numerical data

from scientific images, plot, charts, and figures. R package version 0.1_ https://cran.r-project.org/package=juicr (2021). * Walsh, O. S. _et al_. Assessment of UAV Based Vegetation Indices

for Nitrogen Concentration Estimation in Spring Wheat. _Adv. Remote Sens._ 07, 71–90 (2018). Article  Google Scholar  * Walsh, O. S. _et al_. Wheat yield and protein estimation with

handheld- and UAV-based reflectance measurements. _Agrosystems, Geosci. Environ._ 5, 1–14 (2022). Article  Google Scholar  * Finn, G. A., Straszewski, A. E. & Peterson, V. A general

growth stage key for describing trees and woody plants. _Ann. Appl. Biol._ 151, 127–131 (2007). Article  Google Scholar  * Castilho, D., Tedesco, D., Hernandez, C., Madari, B. E. &

Ciampitti, I. A global dataset for assessing nitrogen-related plant traits using drone imagery in major field crop species. _figshare_ https://doi.org/10.6084/m9.figshare.22938797.v4 (2023).

* Argento, F. _et al_. Site-specific nitrogen management in winter wheat supported by low-altitude remote sensing and soil data. _Precis. Agric._ 22, 364–386 (2021). Article  CAS  Google

Scholar  * Ballester, C., Hornbuckle, J., Brinkhoff, J., Smith, J. & Quayle, W. Assessment of in-season cotton nitrogen status and lint yield prediction from unmanned aerial system

imagery. _Remote Sens._ 9, 1–18 (2017). Article  Google Scholar  * Pipatsitee, P., Tisarum, R. & Taota, K. Effectiveness of vegetation indices and UAV-multispectral imageries in

assessing the response of hybrid maize (Zea mays L.) to water deficit stress under field environment. _Env. Monit Assess_ 195 (2023). * Walsh, O. S. _et al_. Wheat Yield and Protein

Estimation with Handheld and Unmanned Aerial Vehicle-Mounted Sensors. _Agronomy_ 13, 1–14 (2023). Article  Google Scholar  * Zhao, Z. _et al_. A reappraisal of the critical nitrogen

concentration of wheat and its implications on crop modeling. _F. Crop. Res._ 164, 65–73 (2014). Article  Google Scholar  * Ratjen, A. M., Lemaire, G., Kage, H., Plénet, D. & Justes, E.

Key variables for simulating leaf area and N status: Biomass based relations versus phenology driven approaches. _Eur. J. Agron._ 100, 110–117 (2018). Article  CAS  Google Scholar  *

Caturegli, L. _et al_. Normalized Difference Vegetation Index versus Dark Green Colour Index to estimate nitrogen status on bermudagrass hybrid and tall fescue. _Int. J. Remote Sens._ 41,

455–470 (2019). Article  Google Scholar  * Chen, Z. _et al_. In-season diagnosis of winter wheat nitrogen status in smallholder farmer fields across a village using unmanned aerial

vehicle-based remote sensing. _Agronomy_ 9 (2019). * Junior, A. S. D. A. _et al_. Remote detection of water and nutritional status of soybeans using uav-based images. _Eng. Agric._ 4430,

9–23 (2022). Google Scholar  * Fassa, V., Pricca, N., Cabassi, G., Bechini, L. & Corti, M. Site-specific nitrogen recommendations’ empirical algorithm for maize crop based on the fusion

of soil and vegetation maps. _Comput. Electron. Agric_. 203 (2022). * Fu, Z. _et al_. Combining UAV multispectral imagery and ecological factors to estimate leaf nitrogen and grain protein

content of wheat. _Eur. J. Agron._ 132, 126405 (2022). Article  CAS  Google Scholar  * Heinemann, P., Haug, S. & Schmidhalter, U. Evaluating and defining agronomically relevant detection

limits for spectral reflectance-based assessment of N uptake in wheat. _Eur. J. Agron._ 140, 126609 (2022). Article  CAS  Google Scholar  * Jiang, J. _et al_. Combining fixed-wing UAV

multispectral imagery and machine learning to diagnose winter wheat nitrogen status at the farm scale. _Eur. J. Agron._ 138, 126537 (2022). Article  CAS  Google Scholar  * Li, F. _et al_.

Evaluating hyperspectral vegetation indices for estimating nitrogen concentration of winter wheat at different growth stages. _Precis. Agric._ 11, 335–357 (2010). Article  Google Scholar  *

Li, J., Shi, Y., Veeranampalayam-Sivakumar, A. N. & Schachtman, D. P. Elucidating sorghum biomass, nitrogen and chlorophyll contents with spectral and morphological traits derived from

unmanned aircraft system. _Front. Plant Sci._ 9, 1–12 (2018). Google Scholar  * Liu, S. _et al_. Diagnosis of nitrogen status in winter oilseed rape (Brassica napus L.) using _in-situ_

hyperspectral data and unmanned aerial vehicle (UAV) multispectral images. _Comput. Electron. Agric._ 151, 185–195 (2018). Article  ADS  Google Scholar  * Luo, S. _et al_. Remotely Sensed

Prediction of Rice Yield at Different Growth Durations Using UAV Multispectral Imagery. _Agriculture_ 12, 1447 (2022). Article  CAS  Google Scholar  * Maresma, Á., Ariza, M., Martínez, E.,

Lloveras, J. & Martínez-Casasnovas, J. A. Analysis of vegetation indices to determine nitrogen application and yield prediction in maize (zea mays l.) from a standard uav service.

_Remote Sens_. 8 (2016). * Namoi, N. _et al_. Aerial Imagery Can Detect Nitrogen Fertilizer Effects on Biomass and Stand Health of Miscanthus × giganteus. _Remote Sens_. 14 (2022). * Osco,

L. P. _et al_. Leaf nitrogen concentration and plant height prediction for maize using UAV-based multispectral imagery and machine learning techniques. _Remote Sens._ 12, 1–17 (2020).

Article  Google Scholar  * Rodene, E. _et al_. A UAV‐based high‐throughput phenotyping approach to assess time‐series nitrogen.pdf. _Plant Phenome J_. 12, https://doi.org/10.1002/ppj2.20030

(2021). * Vergara-Díaz, O. _et al_. A novel remote sensing approach for prediction of maize yield under different conditions of nitrogen fertilization. _Front. Plant Sci._ 7, 1–13 (2016).

Article  Google Scholar  * Viljanen, N. _et al_. A novel machine learning method for estimating biomass of grass swards using a photogrammetric canopy height model, images and vegetation

indices captured by a drone. _Agric_. 8 (2018). * Wang, H., Mortensen, A. K., Mao, P., Boelt, B. & Gislum, R. Estimating the nitrogen nutrition index in grass seed crops using a

UAV-mounted multispectral camera. _Int. J. Remote Sens._ 40, 2467–2482 (2018). Article  Google Scholar  * Wang, W. _et al_. AAVI: A Novel Approach to Estimating Leaf Nitrogen Concentration

in Rice from Unmanned Aerial Vehicle Multispectral Imagery at Early and Middle Growth Stages. _IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens._ 14, 6716–6728 (2021). Article  ADS  CAS 

Google Scholar  * Wang, W. _et al_. An assessment of background removal approaches for improved estimation of rice leaf nitrogen concentration with unmanned aerial vehicle multispectral

imagery at various observation times. _F. Crop. Res._ 283, 108543 (2022). Article  Google Scholar  * Xu, Y. _et al_. Sustainability trait modeling of field-grown switchgrass (Panicum

virgatum) using uav-based imagery. _Plants_ 10, 1–22 (2021). Article  Google Scholar  * Xu, L. _et al_. An improved approach to estimate ratoon rice aboveground biomass by integrating

UAV-based spectral, textural and structural features. _Precis. Agric._ 23, 1276–1301 (2022). Article  Google Scholar  * Yang, M. _et al_. Assessment of Water and Nitrogen Use Efficiencies

Through UAV-Based Multispectral Phenotyping in Winter Wheat. _Front. Plant Sci._ 11, 1–16 (2020). Article  CAS  Google Scholar  * Zheng, H. _et al_. Evaluation of RGB, color-infrared and

multispectral images acquired from unmanned aerial systems for the estimation of nitrogen accumulation in rice. _Remote Sens_. 10 (2018). * Zheng, H. _et al_. Enhancing the nitrogen signals

of rice canopies across critical growth stages through the integration of textural and spectral information from unmanned aerial vehicle (UAV) multispectral imagery. _Remote Sens_. 12

(2020). Download references ACKNOWLEDGEMENTS The research to create this dataset was funded by the Coordination for the Improvement of Higher Education Personnel - Brazil (CAPES) - Finance

Code 001, sandwich program. B.E.M. is fellow (PQ2) of the National Council for Scientific and Technological Development of Brazil (CNPq Grant No. 307807/2022-3). This is a contribution no.

23-XYZ for the Kansas Agricultural Experiment Station. AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Graduate Program in Agronomy, Federal University of Goiás, Goiânia, Goiás, Brazil Diogo

Castilho & Beata Emoke Madari * Brazilian Agricultural Research Corporation (Embrapa Rice and Beans), Santo Antônio de Goiás, Goiás, Brazil Diogo Castilho & Beata Emoke Madari *

Department of Agronomy, Kansas State University, 1712 Claflin Rd., Manhattan, KS, 66506, USA Danilo Tedesco, Carlos Hernandez & Ignacio Ciampitti Authors * Diogo Castilho View author

publications You can also search for this author inPubMed Google Scholar * Danilo Tedesco View author publications You can also search for this author inPubMed Google Scholar * Carlos

Hernandez View author publications You can also search for this author inPubMed Google Scholar * Beata Emoke Madari View author publications You can also search for this author inPubMed 

Google Scholar * Ignacio Ciampitti View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS The manuscript design was a collaborative effort with

contributions from all authors. Diogo Castilho and Danilo Tedesco worked on the data collection process and were guided by Ignacio Ciampitti and Carlos Hernandez. Diogo Castilho wrote the

original draft of the manuscript, and all authors reviewed the final version of the manuscript. CORRESPONDING AUTHORS Correspondence to Diogo Castilho or Ignacio Ciampitti. ETHICS

DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in

published maps and institutional affiliations. RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,

sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative

Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated

otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds

the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and

permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Castilho, D., Tedesco, D., Hernandez, C. _et al._ A global dataset for assessing nitrogen-related plant traits using drone imagery in major

field crop species. _Sci Data_ 11, 585 (2024). https://doi.org/10.1038/s41597-024-03357-2 Download citation * Received: 15 June 2023 * Accepted: 08 May 2024 * Published: 05 June 2024 * DOI:

https://doi.org/10.1038/s41597-024-03357-2 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable link is not

currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative