Edit on GitHub

Epicrop

Documentation for Epicrop.

Helper Functions

Quickly and easily model bacterial blight, brown spot, leaf blast, sheath blight and tungro. These functions provide the basic values necessary as published in Savary et al. (2012) to model these diseases with just daily weather data that spans the growing season of interest and growing season start date as values.

Bacterial Blight

A dynamic mechanistic simulation of bacterial blight disease of rice, causal agent Xanthomonas oryzae pv. oryzae. The model represents site size as 1 rice plant's leaf. Default values for this disease model are derived from Table 2 (Savary et al. 2012).

Epicrop.bacterialblight — Function
bacterialblight(wth::DataFrames.AbstractDataFrame, emergence::Dates.Date)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for rice bacterial blight Xanthomonas oryzae pv. oryzae.

Arguments

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days), defaults to 120 as in Savary et al. 2012. From Table 1 Savary et al. 2012.

Returns

A DataFrames.DataFrame with predictions for bacterial blight severity. Latitude and longitude are included for mapping purposes if they are present in the input weather data. See hlipmodel for a full description of the return values.

Notes

This is one of a family of helper functions. Also see brownspot, leafblast, sheathblight and tungro.

Examples

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true)

julia> w = DataFrame(data, vec(header))

julia> emergence = "2000-07-01"

julia> bacterialblight(w, emergence)
source

Brown Spot

A dynamic mechanistic simulation of rice brown spot, causal agent Cochliobolus miyabeanus. The model represents site size as 10 mm² of a rice plant's leaf. Default values for this disease model are derived from Table 2 (Savary et al. 2012).

Note The optimum temperature for brown spot as presented in Table 2 of Savary et al. 2012 has a typo. The optimal value should be 25 °C, not 20 °C as shown. The correct value, 25 °C, is used in this implementation.

Epicrop.brownspot — Function
brownspot(wth::DataFrames.AbstractDataFrame, emergence::Dates:Date)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for rice brown spot caused by Cochliobolus miyabeanus.

Arguments

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days), defaults to 120 as in Savary et al. 2012. From Table 1 Savary et al. 2012.

Returns

A DataFrames.DataFrame with predictions for brown spot severity. Latitude and longitude are included for mapping purposes if they are present in the input weather data. See hlipmodel for a full description of the return values.

This function is one of a family of helper functions. Also see bacterialblight, leafblast, sheathblight and tungro.

Examples

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true)

julia> w = DataFrame(data, vec(header))

julia> emergence = "2000-07-01"

julia> brownspot(w, emergence)
source

Leaf Blast

A dynamic mechanistic simulation of leaf blast disease of rice, causal agent Magnaporthe oryzae. The model represents site size as 45 mm² of a rice plant's leaf.

Note The optimum temperature for leaf blast as presented in Table 2 of Savary et al. 2012 has a typo. The optimal value should be 20 °C, not 25 °C as shown. The correct value, 20 °C, is used in this implementation.

Epicrop.leafblast — Function
leafblast(wth::DataFrames.AbstractDataFrame, emergence::Dates.Date)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for rice leaf blast caused by Magnaporthe oryzae.

Arguments

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days), defaults to 120 as in Savary et al. 2012. From Table 1 Savary et al. 2012.

Returns

A DataFrames.DataFrame with predictions for leaf blast severity. Latitude and longitude are included for mapping purposes if they are present in the input weather data. See hlipmodel for a full description of the return values.

Notes

This is one of a family of helper functions. Also see bacterialblight, brownspot, sheathblight and tungro.

Examples

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true)

julia> w = DataFrame(data, vec(header))

julia> emergence = "2000-07-01"

julia> leafblast(w, emergence)
source

Sheath Blight

A dynamic mechanistic simulation of sheath blight disease of rice, causal agent Rhizoctonia solani AG1-1A Kühn. The model represents site size as 1 rice plant's tiller.

Epicrop.sheathblight — Function
sheathblight(wth::DataFrames.AbstractDataFrame, emergence::Dates.Date)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for rice sheath blight caused by Rhizoctonia solani AG1-1A Kühn.

Arguments

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days), defaults to 120 as in Savary et al. 2012. From Table 1 Savary et al. 2012.

Returns

A DataFrames.DataFrame with predictions for sheath blight severity. Latitude and longitude are included for mapping purposes if they are present in the input weather data. See hlipmodel for a full description of the return values.

Notes

This is one of a family of helper functions. Also see bacterialblight, brownspot, leafblast and tungro.

Examples

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true)

julia> w = DataFrame(data, vec(header))

julia> emergence = "2000-07-01"

julia> sheathblight(w, emergence)
source

Tungro

A dynamic mechanistic simulation of tungro disease of rice, causal agents Rice Tungro Spherical Virus and Rice Tungro Bacilliform Virus. The model represents site size as 1 rice plant. Default values for this disease model are derived from Table 2 (Savary et al. 2012).

Epicrop.tungro — Function
tungro(wth::DataFrames.AbstractDataFrame, emergence::Dates.Date)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for rice tungro disease caused by Rice Tungro Spherical and Rice Tungro Bacilliform viruses.

Arguments

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days), defaults to 120 as in Savary et al. 2012. From Table 1 Savary et al. 2012.

Returns

A DataFrames.DataFrame with predictions for tungro incidence. Latitude and longitude are included for mapping purposes if they are present in the input weather data. See hlipmodel for a full description of the return values.

Notes

This is one of a family of helper functions. Also see bacterialblight, brownspot, leafblast and sheathblight.

Examples

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true)

julia> w = DataFrame(data, vec(header))

julia> emergence = "2000-07-01"

julia> tungro(w, emergence)
source

Advanced Use

Epicrop.jl provides a basic function, hlipmodel, that can be used to predict unmanaged plant disease epidemics given the proper input. Predefined values for the EPIRICE model can be found in Savary et al. (2012) for the following diseases of rice: bacterial blight, brown spot, leaf blast, sheath blight and tungro and are included as helper functions, bacterialblight, brownspot, leafblast, sheathblight, and tungro, that simplify using the model's interface. Given other parameters, the model framework is capable of modelling other diseases using the methods as described by Savary et al. (2012).

Epicrop.hlipmodel — Function
hlipmodel(;
wth::DataFrames.AbstractDataFrame,
rhlim::Int64,
rainlim::Int64,
H0::Int64,
I0::Int64,
RcA::Matrix{Float64},
RcT::Matrix{Float64},
RcOpt::Float64,
p::Int64,
i::Int64,
Sx::Int64,
a::Float64,
RRS::Float64,
RRG::Float64,
)

Run a healthy-latent-infectious-postinfectious (HLIP) model using weather data and optimal curve values for respective crop diseases.

Keywords

  • wth::DataFrames.AbstractDataFrame: weather data on a daily time-step containing data with the following values.
    • `YYYYMMDD::Union{AbstractString, Dates.Date}: the date in ISO8601 format, i.e., YYYY-MM-DD
    • DOY::Int64: the consecutive day of year, commonly called "Julian date"
    • TEMP::AbstractFloat: the mean daily temperature (°C)
    • RHUM::AbstractFloat: the mean daily relative humidity (%)
    • RAIN::AbstractFloat: the mean daily rainfall (mm)
  • emergence::Union{AbstractString, Dates.Date}: the expected date of plant emergence. From Table 1 Savary et al. 2012.
  • onset::Int64: the expected number of days until the onset of disease after emergence date. From Table 1 Savary et al. 2012.
  • duration::Int64: the simulation duration (growing season length in days). From Table 1 Savary et al. 2012.
  • rhlim::Int64: the threshold to decide whether leaves are wet or not (usually 90%). From Table 1 Savary et al. 2012.
  • rainlim::Int64: the threshold to decide whether leaves are wet or not. From Table 1 Savary et al. 2012.
  • H0::Int64: the initial number of plant's healthy sites. From Table 1 Savary et al. 2012.
  • I0::Int64: the initial number of infective sites. From Table 1 Savary et al. 2012.
  • RcA::Matrix{Float64}: a crop age modifier for Rc (the basic infection rate corrected for removals). From Table 1 Savary et al. 2012.
  • RcT::Matrix{Float64}: a temperature modifier for Rc (the basic infection rate corrected for removals). From Table 1 Savary et al. 2012.
  • RcOpt::Float64: the potential basic infection rate corrected for removals. From Table 1 Savary et al. 2012.
  • p::Int64: the duration of latent period. From Table 1 Savary et al. 2012.
  • Sx::Int64: the maximum number of sites. From Table 1 Savary et al. 2012.
  • a::Float64: an aggregation coefficient. From Table 1 Savary et al. 2012.
  • RRS::Float64: the relative rate of physiological senescence. From Table 1 Savary et al. 2012.
  • RRG::Float64: the relative rate of growth. From Table 1 Savary et al. 2012.

Output

A DataFrames.AbstractDataFrame with the model's output with the following fields and values.

  • simday::Int64: the zero indexed day of simulation that was run.
  • dates::Dates.Date: the date of the simulation.
  • sites::Float64: the otal number of sites present on day "x".
  • latent::Float64: the number of latent sites present on day "x".
  • infectious::Float64:: the number of infectious sites present on day "x".
  • removed::Float64: the number of removed sites present on day "x".
  • senesced::Float64: the number of senesced sites present on day "x".
  • rateinf::Float64: the rate of infection.
  • rtransfer::Float64: the rate of transfer from latent to infectious sites.
  • rgrowth::Float64: the rate of growth of healthy sites.
  • rsenesced::Float64: the rate of senescence of healthy sites.
  • diseased::Float64: the number of diseased (latent + infectious + removed) sites.
  • intensity::Float64: the number of diseased sites as a proportion of total sites.
  • audpc::Float64: the area under the disease progress curve for the whole of simulated season.
  • lat::Float64: the latitude value as provided by wth object.
  • lon::Float64: the longitude value as provided by wth object.

Examples

Using data downloaded and saved from NASA POWER for Los Baños, Calabarzon, PH.

julia> using Epicrop

julia> using DelimitedFiles

julia> using DataFrames

julia> data, header = readdlm(joinpath(
                                dirname(pathof(Epicrop)),
                                    "..", "docs", "src", "assets",
                                        "POWER_data_LB_PHI_2000_ws.csv"),
                                ',', header=true);

julia> w = DataFrame(data, vec(header));

julia> emergence = "2000-07-01";

julia> RcA = [0 0.35;
            20 0.35;
            40 0.35;
            60 0.47;
            80 0.59;
            100 0.71;
            120 1.0];

julia> RcT = [15 0.0;
            20 0.06;
            25 1.0;
            30 0.85;
            35 0.16;
            40 0.0];

julia> bs = hlipmodel(;
        wth=w,
        emergence=emergence,
        onset=20,
        duration=120,
        rhlim=90,
        rainlim=5,
        H0=600,
        I0=1,
        RcA=RcA,
        RcT=RcT,
        RcOpt=0.61,
        p=6,
        i=19,
        Sx=100000,
        a=1.0,
        RRS=0.01,
        RRG=0.1)

        120×16 DataFrame
        Row │ simday  dates       sites      latent    infectious  removed   senesced ⋯
            │ Int64   Date        Float64    Float64   Float64     Float64   Float64  ⋯
       ─────┼──────────────────────────────────────────────────────────────────────────
          1 │      1  2000-07-01    600.0       0.0         0.0      0.0         0.0  ⋯
          2 │      2  2000-07-02    653.64      0.0         0.0      0.0         6.0
          3 │      3  2000-07-03    712.04      0.0         0.0      0.0        12.53
          4 │      4  2000-07-04    775.617     0.0         0.0      0.0        19.65
          5 │      5  2000-07-05    844.821     0.0         0.0      0.0        27.41 ⋯
          6 │      6  2000-07-06    920.141     0.0         0.0      0.0        35.86
          7 │      7  2000-07-07   1002.11      0.0         0.0      0.0        45.06
          8 │      8  2000-07-08   1091.29      0.0         0.0      0.0        55.08
            │   ⋮         ⋮           ⋮         ⋮          ⋮          ⋮          ⋮    ⋱
        114 │    114  2000-10-22  87883.5    1038.79      454.187   70.2366  49807.5  ⋯
        115 │    115  2000-10-23  87923.0     941.617     542.256   79.3388  50695.5
        116 │    116  2000-10-24  87958.1     825.098     648.67    89.4445  51584.8
        117 │    117  2000-10-25  87988.2     686.497     775.218  101.498   52476.4
        118 │    118  2000-10-26  87594.1     959.018     936.18   101.498   53356.3  ⋯
        119 │    119  2000-10-27  87095.0    1332.85     1097.35   101.498   54232.2
        120 │    120  2000-10-28  86500.7    1797.92     1259.08   101.498   55103.2
                                                        10 columns and 105 rows omitted
source

References

Savary, S., Nelson, A., Willocquet, L., Pangga, I., and Aunario, J. Modeling and mapping potential epidemics of rice diseases globally. Crop Protection, Volume 34, 2012, Pages 6-17, ISSN 0261-2194 DOI: 10.1016/j.cropro.2011.11.009.