This readme.txt file was generated on <20190508> by <Wanda Marsolek and Beth Fallon>


-------------------
GENERAL INFORMATION
-------------------

1. Title of Dataset 
Leaf and canopy spectra, symptom progression, and physiological data from experimental detection of oak wilt in oak seedlings

2. Author Information

Beth Fallon
eafallon@gmail.com

  Principal Investigator Contact Information
        Name: Jeannine Cavender-Bares
           Institution: University of Minnesota
           Address: 140 Gortner Labs, 1479 Gortner Ave, St Paul MN 55108
           Email: cavender@umn.edu

  Associate or Co-investigator Contact Information
        Name: Rebecca Montgomery
           Institution: University of Minnesota

  Associate or Co-investigator Contact Information
           Name: Jennifer Juzwik
           Institution: University of Minnesota

3. Date of data collection:
20170713 - 20181025

4. Geographic location of data collection: 
Research was conducted in USDA Forest Service greenhouse facilities on the University of Minnesota campus which occupies Dakota lands. 

5. Information about funding sources that supported the collection of the data and acknowledgements:
These data were collected under funding from a University of Minnesota Grand Challenges Research Grant (Protection of biodiversity and ecosystems services through early detection of tree disease using hyperspectral remote sensing, Co-PIs J. Cavender-Bares, J. Juzwik, and R. Montgomery) and a Minnesota Invasive Terrestrial Plants and Pests Center (Legislative-Citizen Commission on Minnesota Resources) Grant (Accurate detection and integrated treatment of oak wilt, PI J. Cavender-Bares). 

--------------------------
SHARING/ACCESS INFORMATION
-------------------------- 

1. Licenses/restrictions placed on the data:
Creative Commons, attribution

2. Links to publications that cite or use the data: NA

3. Links to other publicly accessible locations of the data: NA

4. Links/relationships to ancillary data sets:
1nm resolution canopy and leaf spectral reflectance: Available on-line [http://ecosis.org] from the Ecological Spectral Information System (EcoSIS)

5. Was data derived from another source? No

6. Recommended citation for the data:
Fallon, Beth; Yang, Anna; Nguyen, Cathleen; Armour, Isabella; Juzwik, Jennifer; Montgomery, Rebecca A.; Cavender-Bares, Jeannine. (2019). Leaf and canopy spectra, symptom progression, and physiological data from experimental detection of oak wilt in oak seedlings. Retrieved from the Data Repository for the University of Minnesota, http://hdl.handle.net/11299/202591.

---------------------
DATA & FILE OVERVIEW
---------------------

1. File List
   A. Filename: oakWiltSeedl_README.txt        
      Short description: overall data/project description with data file descriptions and variable definitions  

   B. Filename: oakWiltSeedl_individuals.csv       
      Short description: a reference set of species and treatments, with greenhouse locations. Individuals were randomly allotted to treatments and then allotted to benches in a stratified random way to have approximately equal representation of each species and treatment combination.

   C. Filename: oakWiltSeedl_inoculation_dates.csv        
      Short description: experimental start dates and dates plants were inoculated, used to calculate exp.day in rest of files
   
   D. Filename: oakWiltSeedl_leaf_waterpotential.csv        
      Short description: pre-dawn and mid-day water potentials of (green) leaves in experiment
   
   E. Filename: oakWiltSeedl_gas_exchange.csv        
      Short description: leaf gas exchange measurements on Li-COR 6400-XT 

   F. Filename: oakWiltSeedl_symptom_prevalence.csv        
      Short description: symptom tracking on BOB, OW, and D (2018 only for D) seedlings. On BOB plants, only inoculated leaves were tracked (all leaves in 2017, 5/plant in 2018)

   G. Filename: oakWiltSeedl_spectra_2017_leaf_10nm.csv          
      Short description: 2017 leaf level spectra, aggregated (mean value) from 4 measurements of leaves to each individual by date. Resampled to every 10nm

   H. Filename: oakWiltSeedl_spectra_2018_leaf_10nm.csv       
      Short description: 2018 leaf level spectra, aggregated (mean value) from up to 3 measurements of each leaf type on each individual to each individual by date by leaf type combination. Resampled to every 10nm

   I. Filename: oakWiltSeedl_spectra_2018_leaf_symptomatic_10nm.csv        
      Short description: 2018 leaf level spectra from symptomatic leaf measurements only, aggregated (mean value) from all symptomatic leaf measurements on an individual plant to each individual by date 

   J. Filename: oakWiltSeedl_spectra_2018_wholeplant_10nm.csv        
      Short description: 2018 whole plant spectra from each of up to four views of the plant canopy. Resampled to every 10nm and only including data up to 1890nm because of increased variance in light source after that point

   K. Filename: oakWiltSeedl_thermal_leaf.csv       
      Short description: mean values of leaf adaxial surface temperature and variation from whole leaf thermal images 

2. Relationship between files:        

README files are overall data description and describe variables in each dataset. _individuals.csv contains a full list of all oak seedlings with unique identifiers and _inoculation_dates.csv contains the start dates of the experiment. All other files contain experimental data.

--------------------------
METHODOLOGICAL INFORMATION
--------------------------


1. Description of methods used for collection/generation of data: 
These data were collected as part of an experimental effort to accurately detect oak wilt infections in oak seedlings using remote sensing tools and to differentiate that disease stress from other mechanisms of tree decline. Oak wilt disease causes rapid mortality in oaks in the central and eastern United States. Management of the disease requires early diagnosis and tree removal to prevent fungal spread. Hyperspectral tools provide a potential method of early remote diagnosis, but accurately differentiating oak wilt from other agents of oak decline is integral to effective management. We conducted experiments (2017 and 2018) on two year old seedlings of Quercus ellipsoidalis and Q. macrocarpa in which treatments were 1) maintained as healthy individuals, 2) subjected to chronic drought, or inoculated 3) stems with oak wilt fungus (Bretziella fagacearum, a fungal vascular wilt) or 4) leaves with bur oak blight fungus (Tubakia iowensis, a fungal leaf pathogen).  We measured leaf and whole plant hyperspectral reflectance (350 to 2400nm, Spectra Vista HR 1024i spectroradiometer (Spectra Vista Corporation, New York, USA)), gas exchange (LI-6440XT with a leaf chamber fluorometer attachment (LI-COR Environmental, Nebraska, USA)), and tracked symptom development in repeated measures of seedlings over the course of each experiment. In 2018, we explicitly measured spectral reflectance and gas exchange on both symptomatic and green leaves, as available and we also measured collected thermal images of leaves twice during the experiment (2018 only). 

2. Methods for processing the data: 
Spectra were resampled to every 10nm and sensor jumps were corrected.

3. Instrument- or software-specific information needed to interpret the data:
Spectra acquisition and processing details:
Leaf spectra of oak seedlings, healthy or treated with oak wilt, bur oak blight, drought
Foreoptic Type	Leaf Clip
Instrument Manufacturer	Spectra Vista Corporation
Instrument Model	HR-1024i
Instrument Serial Number	HI: 1152050
Light Source	Lamp
Measurement Quantity	Reflectance
Measurement Units	%
Processing Averaged	Yes
Processing Information Details	resampled to 10 nm spectral resolution, jumps between sensors corrected
Processing Interpolated	Yes
Processing Resampled	Yes
Target Status	Fresh
Target Type	Leaf
Theme	Ecology
Wavelength Units	nm

Seedling canopy spectra of oak seedlings, healthy or treated with oak wilt, bur oak blight
Foreoptic Type	fiber optic 25 deg
Instrument Manufacturer	Spectra Vista Corporation
Instrument Model	HR-1024i
Instrument Serial Number	HI: 1152050
Light Source	72W halogen lamp
Measurement Quantity	Reflectance
Measurement Units	%
Processing Averaged	No
Processing Information Details	resampled to 10 nm spectral resolution, jumps between sensors corrected
Processing Interpolated	Yes
Processing Resampled	Yes
Target Status	Fresh
Target Type	Leaf
Theme	Ecology
Wavelength Units	nm

4. Standards and calibration information, if appropriate:
Spectral measurements referenced to zenith or spectralon 99% reflectance white standards

5. Environmental/experimental conditions: NA

6. Describe any quality-assurance procedures performed on the data: NA

7. People involved with sample collection, processing, analysis and/or submission:
The project was conceived of by J. Cavender-Bares, J. Juzwik, and R. Montgomery, set up by C. Nguyen, A. Yang, and B. Fallon, data were collected by B. Fallon, A. Yang, I. Armour, and C. Nguyen, and analyses were performed by B. Fallon. B. Fallon and J. Cavender-Bares wrote the manuscript. Lillian Legg, Mason Martinez, Jake Botkin, Shan Kothari, Jose Eduardo Meireles, Anna Schweiger, Marisol Cruz Aguilar, Kali Hall, Alex Feltmeyer, and Mark Moeller helped collect data or inoculate seedlings. We are very thankful for the assistance of Paul Castillo (USDA Forest Service), who managed the greenhouse facilities, and Melanie Moore, who assisted with greenhouse management in a pinch, and Geoffrey Harms who constructed an apparatus for whole plant spectral measurements and creatively repaired malfunctioning equipment several times. Jose Eduardo Meireles, Anna Schweiger, and Tom Corl provided valuable advice on data collection, and especially on the methods for measuring whole canopy spectra.

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_individuals: 
a reference set of species and treatments, with greenhouse locations. Individuals were randomly allotted to treatments and then allotted to benches in a stratified random way to have approximately equal representation of each species and treatment combination.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Number of variables = 10
Number of cases/rows = 358
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
bench = bench location within greenhouse (ordered from smallest to largest, south to north in greenhouse)
row = row within bench
col = column position within bench
wholeplant = whether the plants were used for whole plant measurements (WP) or not (NA)
removed and removed.notes = if plants were removed from the experiment after inoculations
year = year data collected

-------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_inoculation_dates
dates used to calculated exp.day
-------------------------------------------------------------
Number of variables = 3
Number of cases/rows = 2
year = year of data collection
inoc.date = date of inoculation with OW or BOB. Drought was imposed same day (2018) or week prior (2017)
inoc.yday = julian day of year of inoculation date

----------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_leaf_waterpotential 
pre-dawn and mid-day water potentials of (green) leaves in experiment
----------------------------------------------------------------------
Number of variables = 14
Number of cases/rows: = 463
id = unique individual plant identification field
year =  year of data collection
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
date = measurement date
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
PD.Leaf = pre-dawn water potential, MPa
MD.Leaf = mid-day (12-13:30) water potential, same day as pre-dawn, MPa
WP.diff = difference between MD and PD water potential (MD-PD), MPa
ctrl.meanPD = average value of control trt water pre-dawn water potential for that species on that day, MPa
ctrl.meanMD = average value of control trt water mid-day water potential for that species on that day, MPa
PD.diffCtrl = difference between PD.Leaf and ctrl.meanPD for that individual (PD.Leaf - ctrl.meanPD), MPa
MD.diffCtrl = difference between MD.Leaf and ctrl.meanMD for that individual (MD.Leaf - ctrl.meanMD), MPa

---------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_gas_exchange 
gas exchange measurements on Li-COR 6400-XT measurements 
---------------------------------------------------------
Number of variables = 26
Number of cases/rows = 1152
date = measurement date
HHMMSS = measurement time
year = year of data collection
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
leaftype = GREEN (not obviously symptomatic leaf), SYM (advanced symptomatic tissue), S.T (area of leaf tissue between symptomatic and green tissue on the same leaf), and S.G (green tissue on a symptomatic leaf)
Photo = instantaneous photosynthetic rate, micromol (CO2) m^-2 s^-1
Cond = instantaneous stomatal conductance, mol (H20) m^-2 s^-1
Trmmol = transpiration rate, mmol (H20) m^-2 s^-1 
PhiPS2 = photosynthetic yield from light-adapted chlorophyll fluorescence
Ci = leaf intercellular CO2 concentration, micromol mol^-1
VpdL = leaf vapor pressure deficit, kPa
Tair = air temperature in chamber, deg C
Tleaf = leaf temperature, deg C
TBlk = Li-COR internal temperature, deg C
CO2R = concentration of reference CO2,ppm
RH_R = relative humidity reference, %
RH_S = relative humidity sample, %
Flow = air flow rate, micromol s^-1
ctrl.meanPhoto = average value of control Photo for that species in that experimental week, micromol (CO2) m^-2 s^-1
ctrl.meanCond = average value of control Cond for that species in that experimental week, mol (H20) m^-2 s^-1
Photo.diffCtrl = difference between Photo and ctrl.meanPhoto for that individual (Photo - ctrl.meanPhoto), micromol (CO2) m^-2 s^-1
Cond.diffCtrl = difference between Cond and ctrl.meanCond for that individual (Cond - ctrl.meanCond), mol (H20) m^-2 s^-1

----------------------------------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_symptom_prevalence
 symptom tracking on BOB, OW, and D (2018 only for D) seedlings. On BOB plants, only inoculated leaves were tracked (all leaves in 2017, 5/plant in 2018)
----------------------------------------------------------------------------------------------------------------------------------------------------------
Number of variables = 16
Number of cases/rows = 3649
id = unique individual plant identification field
year = year of data collection
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
date = measurement date
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
number.healthy = number of healthy leaves
number.wilt = number of leaves showing any sign of wilting (turgor loss to desiccation), BOB NA
number.bronzing = number of leaves showing any sign of discoloration typical to oak wilt infections, BOB NA
number.abscissed = number of leaves abscissed
total.leaves = total number of leaves on the plant (maximum value of observed counts)
perc* fields = previous symptom status number fields divided by total leaves as the percent of each species canopy affected


==========================================================================================================================================================================================================
[note on spectra processing: dropped all leaf level spectra that <35% reflectance at 770nm as poor quality and whole canopy spectra < 20% reflectance, sensor jumps corrected, resampled and interpolated]
==========================================================================================================================================================================================================

-----------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_spectra_2017_leaf_10nm
2017 leaf level spectra, aggregated (mean value) from 4 measurements of leaves to each individual by date. Resampled to every 10nm
-----------------------------------------------------------------------------------------------------------------------------------
Number of variables = 208
Number of cases/rows: 483
sample_name = same as id, generated during spectrolab processing
date = measurement date
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
400:2400 reflectance values (0-1), based on regularly collected white references

---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_spectra_2018_leaf_10nm
2018 leaf level spectra, aggregated (mean value) from up to 3 measurements of each leaf type on each individual to each individual by date by leaf type combination. Resampled to every 10nm
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Number of variables = 210
Number of cases/rows = 1535
sample_name = id and leaftype combined, generated during processing
date = measurement date
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
leaftype = GREEN (not obviously symptomatic leaf), SYM (advanced symptomatic tissue), S.T (area of leaf tissue between symptomatic and green tissue on the same leaf), and S.G (green tissue on a symptomatic leaf)
n_spec: number of spectra measures averaged to create this aggregate value
400:2400 reflectance values (0-1), based on regularly collected white references

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_spectra_2018_leaf_symptomatic_10nm
2018 leaf level spectra from symptomatic leaf measurements only, aggregated (mean value) from all symptomatic leaf measurements on an individual plant to each individual by date combination. Resampled to every 10nm.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Number of variables = 212
Number of cases/rows = 275
sample_name = id with preceding string generated during spectrolab spectra processing
date = measurement date
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
n_spec: average number of spectra measures per leaf type used to create this aggregate value
S.G, S.T, SYM: binary if type of leaf used in average value (1 = included)
400:2400 reflectance values (0-1), based on regularly collected white references

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_spectra_2018_wholeplant_10nm
2018 whole plant spectra from each of up to four views of the plant canopy. Resampled to every 10nm and only including data up to 1890nm because of increased variance in light source after that point
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Number of variables = 158
Number of cases/rows = 575
sample_name: id with preceding string generated during spectrolab spectra processing
date = measurement date
id = unique individual plant identification field
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
week = experimental week (week 1 = start of inoculation or drought imposition)
exp.day = day from start of experiment (inoculation date)
pot.side = view of canopy (4 ordinal directions, repeatedly measured)
400:1890 reflectance values (0-1), based on regularly collected white references

---------------------------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: oakWiltSeedl_thermal_leaf
mean values of leaf adaxial surface temperature and variation from whole leaf thermal images 
---------------------------------------------------------------------------------------------
Number of variables = 6
Number of cases/rows = 136
species = ellip (Quercus ellipsoidalis, northern pin oak) or macro (Quercus macrocarpa, bur oak)
trt = OW (oak wilt), D (drought), C (control), or BOB (bur oak blight)
leaftype = GREEN (not obviously symptomatic leaf), SYM (advanced symptomatic tissue), S.T (area of leaf tissue between symptomatic and green tissue on the same leaf), and S.G (green tissue on a symptomatic leaf)
Leaf_mean = mean value of  within leaf temperature (all pixels on a single leaf), deg C
Leaf_sd = standard deviation of within leaf temperature (all pixels on leaf)
CV = coefficient of variation of within leaf temperature