Reducing Pesticide Use in Vegetable Production by Calculating Soil-borne Disease Risk

2016 Annual Report for GNE15-103

Project Type: Graduate Student
Funds awarded in 2015: $14,633.00
Projected End Date: 12/31/2016
Grant Recipient: Cornell University
Region: Northeast
State: New York
Graduate Student:
Faculty Advisor:
Dr. Sarah Pethybridge
Cornell University

Reducing Pesticide Use in Vegetable Production by Calculating Soil-borne Disease Risk


The Northern root-knot nematode Meloidogyne hapla is an important soilborne pathogen of potatoes and other vegetables in New York.  This pathogen reduces tuber yield and quality.  Management decisions for nematodes need to be made prior to planting.  Information regarding nematode populations is critical in determining the implementation of management tactics, but is not readily available to most growers.  Decisions made therefore are often risk averse, and assumptions of the ubiquitous presence of the nematode frequently leads to unnecessary application of chemical treatments.  These chemistries may be environmentally and economically costly, raising the cost of production and have negative impacts on the environment. 

The holistic objective of my project is to reduce unnecessary applications by increasing informed decision making through the development of a new quantitative risk algorithm for M. hapla that will estimate tuber damage at harvest based upon initial pathogen populations.  This tool will aid in decision making by incorporating information on nematode populations prior to planting, host plant susceptibility, and other edaphic factors to determine a risk category for the field, from which management decisions may be directed.   

Objectives/Performance Targets

The goal of this multi-year project is to develop a new quantitative risk algorithm for M. hapla that will predict tuber damage and crop yield at harvest based upon initial populations as measured by pathogen DNA in the soil.  To achieve this goal, specific objectives of the project include determining a suitable method for extracting DNA from soil and also the development of species-specific quantitative PCR primers and probes.  The results of these objectives were reported in the previous annual report.  An additional objective of this project was to test the relationship between M. hapla populations and tuber damage.  To accomplish this, three commercial potato fields in New York with high populations of M. hapla and the lesion nematode Pratylenchus spp. were selected in collaboration with two participating growers.  Two fields were located in Springville and were planted with the fresh market variety ‘Eva’ and the third field was located in Wayland and was planted with the chipping variety ‘Lamoka’.  

The fields were intensively sampled soon after planting and prior to emergence (May 2016; initial population at planting).  Soil samples (~ 1 kg) were collected from the hills at a depth of approximately 6 inches at each point.  Samples were collected on a 10 by 10 point grid, giving 100 total points per field.  Each point was marked with a flag and the location recorded by GPS.  Distance between the points varied from 20 to 50 feet, depending on the size and shape of the field.  The soil samples were divided into smaller portions for use in three quantification experiments.  First, a subsample was subjected to manual counting using a modified Whitehead tray to extract live, motile-stage nematodes.  An aliquot of the extraction suspension was then quantified under the microscope to quantify populations by identifying nematode genera.  Secondly, a subsample was used in a tomato “bait plant” experiment.  Field soil was placed into 5 inch clay pots, and a tomato seedling was planted into the soil.  Plants were maintained for seven weeks, after which they were up-rooted, the roots washed, and root galling severity scored.  Bait plants are a common research method for quantifying Meloidogyne spp. populations.  A final subsample was prepared for storage and DNA extraction. 

The fields were sampled again after vine kill and prior to harvest (September/October 2016, final population at harvest).  Soil samples were collected from the same points as the initial sampling.  As before, approximately 1 kg of soil was collected from the hills at a depth of about 6 inches.  Soil was portioned into two subsamples, the first for manual counting by use of a modified Whitehead tray, and another for storage and DNA extraction.  Additionally, one plant was collected at each sampling point.  Data on total yield (kg), number of tubers, diameter of tubers, root mass, root galling severity, and tuber damage were collected. 

Data was analyzed using a paired Student’s t-test to compare the initial and final populations as determined by the manual quantification of live motile nematodes from the Whitehead trays.  Additionally, correlation analysis and regression models were used to model the relationships between initial populations and the various measures of yield and crop damage within the R Studio statistical platform (R Core Team, version 3.1.1). 

Geostatistics was used to assess spatial patterns of nematode populations and determine the degree of spatial dependence between two points in space.  Manual count data at each sampling point was interpolated to estimate populations at non-sampled locations, i.e. locations between sampling points.  In the case of spatial dependence, locations closer together are assumed to be more similar in characteristics than locations farther apart.  To accomplish the interpolation, a semivariogram for each field × species × sampling time combination was calculated.  This quantifies the degree of spatial correlation between observations at different sampling locations (semivariance vs. separation distance).  From the semivariogram, ordinary kriging was employed to estimate the populations over the entire sampling area.  Ordinary kriging uses a function of weighted averages of known values (the sampling points) to predict the unknown values (locations between the sampling points) (Cressie, 1988).  Semivariograms were calculated, fit, and kriging was performed in R Studio using the package ‘gstat’. 


Counting of M. hapla and Pratylenchus spp., across all sampling locations within each of the fields found high localized variability in populations at planting and harvest (Table 1).  Populations of M. hapla at harvest were significantly higher than at planting in two of the three potato fields. In contrast, Pratylenchus spp. populations were significantly higher at harvest than at planting in one of the three fields (Table 1).  Manual counting of nematodes from individual locations were also used to estimate the nematode populations over the entire sampling area at both pre-plant and pre-harvest sampling times (Figure 1).  This indicated that nematode populations were highly variable within the fields, in a generally aggregated spatial pattern.  At many locations, high populations of M. hapla and Pratylenchus spp. did not necessarily occur together. 

Regression analyses of bait plant root galling severity with populations of M. hapla at planting identified a significant positive linear correlation, indicated that root galling severity in the bait plants was highly associated with the populations of M. hapla in the soil.  This supports the hypothesis that this common technique for quantifying M. hapla populations is accurate, albeit prohibitively laborious for use as a pre-plant risk assessment for growers (Figure 2). 

No significant association was found between total weight (kg) of tubers per plant and M. hapla and Pratylenchus spp. populations at planting within any of the three fields (Table 2).  This finding suggests that ‘Eva’ and ‘Lamoka’ may be good hosts for these two nematodes but may also have some tolerance.

A significant positive association was identified between the total weight (kg) of tubers per plant and M. hapla and Pratylenchus spp. populations at harvest (Table 2).  This finding suggests that a higher root volume may simply be providing support for the multiplication nematode populations and supports the potential for tolerance to these nematodes in these varieties.  Annual Report 2016 Figures – A. Gorny

Impacts and Contributions/Outcomes

Stakeholder engagement is a key facet of this project.  Two potato growers in western and central New York were involved for the on-farm sampling.  The graduate student and faculty mentor leading this project met with participating growers on a bi-monthly basis to inform and present findings from the project.  Participating growers received information about the types and relative populations of plant-parasitic nematodes present in their fields, and potential management tactics discussed.  Each grower expressed high interest and satisfaction with the research taking place on their farms. 

Results from soil sampling were presented at the Fresh Market Potato Variety and Pest Management Meeting on August 25, 2016, an extension based event held in Marion, NY.  Attendees included growers, crop consultants, and Cornell Cooperative Extension personnel.  The presentation included a “take home” outline of the results, and a robust question and answer session.

Results from the broader project were presented at national and regional scientific conferences, including the American Phytopathological Society Annual Meeting in Tampa, FL (July 30 – August 3) and the Northeast Division Meeting of the American Phytopathological Society in Ithaca, NY (October 19 – 21). 

Work conducted since receiving the Northeast SARE Graduate Student Grant has consisted of selecting and validating an optimal DNA isolation method from soil, development of a qPCR primer and probe set for use in molecular quantification, and field sampling to establish the relationship between nematode pre-plant populations with crop damage or yield observed at harvest.  Future work consists of repeating the field sampling with three additional commercial potato fields to establish the robustness of these relationships.  Future work also consists of quantifying variation in varietal susceptibility to M. hapla and Pratylenchus spp.  



Cressie N  (1988)  Spatial prediction and ordinary kriging.  Mathematical Geology  20:405-421. 

R Core Team  (2014)  R: A language and environment for statistical computing.  R Foundation for Statistical Computing, Vienna, Austria.  URL



Dr. Sarah Pethybridge
Assistant Professor
Cornell University
630 West North Street
Geneva, NY 14456
Office Phone: 3157872417