A Summary of Seven Years of Blueberry Research Trials in Tennessee

James B. Wills, David W. Lockwood, Gary Honea 

Introduction 

            There has been a growing interest in producing blueberries across the state of Tennessee in recent years.  Blueberry consumption has not been as high in Tennessee as in some northern states.  However, with the influx of people from northern areas into Tennessee who bring with them a taste for blueberries as well as Tennessee natives who are finding out about the many good qualities of the blueberry that relate to health, the potential for increasing blueberry production is very good.  Research in Tennessee was needed to determine what varieties of blueberries would grow well in the varied climates and soil types across the state and to determine the best cultural practices that would permit high plant survival rates as well as high fruit production. A research program was begun in 2000 to obtain research based information on blueberry production that would be useful to current and future growers. 

Materials and Methods 

Plot Establishment 

            Two varieties of blueberries were selected for the trial that were expected to do well in Tennessee.  The Rabbiteye variety selected was Tifblue and the Highbush variety was Bluecrop.  The Bluecrop plants were obtained from a  Michigan nursery with a ball root system.  The selected pollinator was Duke.  The Tifblue were obtained from a nursery in North Carolina and were bare root plants. The selected pollinator was Centurion.  The experimental design was randomized complete block with a split plot factorial arrangement of treatments.  There were five mulch treatments: (1) raised bed with no mulch (None), (2) raised bed with decomposed sawdust mulch (Organic), (3) raised bed with black woven polyethylene ground cover (Black), (4) raised bed with black woven polyethylene ground cover and sawdust on top (Black/Organic), and (5) raised bed with black woven polyethylene ground cover painted white (White).  All treatments were on raised six inch beds four feet wide with trickle irrigation and the capability to fertigate (irrigate and apply fertilizer) through the irrigation system. Beds for planting were formed with a Rain-Flo Model 2600 Raised Bed Plastic Mulch Layer. The plastic mulch was five feet wide and was applied at time of bed formation. Drip tape for fertigation was installed at the same time. The five treatments were replicated four times.  Pollinators were placed in border rows on each side of the plots, at one end of each treatment rows and approximately midway down the row of each treatment.  Each treatment row consisted of five plants of each variety (a total of 10 plants) and one pollinator for each variety for a total of 12 plants.  Research blueberry plots were located at three UT research and education centers: (1) Plateau Research and Education Center (PREC) at Crossville, (2) Highland Rim Research and Education Center HRREC) at Springfield, and the (3) Middle Tennessee Research and Education Center (MTREC) at Spring Hill.  Plots at each station were basically the same with minor variations commensurate with the limits of the individual locations.  At each location a total of 240 record plants were transplanted with additional plants in the borders for replacement of mortalities. The goal was to evaluate survivability and productivity of the two varieties under the five different treatment systems.  The plants were transplanted in late February and early March of 2000 after sawdust mulch was worked into the soil and sulfur added to bring the pH into the desired range of 5.0 to 5.2. 

Temperature and Moisture 

            Each plot was instrumented with thermocouples to measure soil temperature and Water Mark moisture sensors to measure soil moisture at a 5 inch depth ( ~ 12 cm).  Both temperature and moisture sensors were connected to a data logger to record continuous temperature and moisture readings at 15 locations in the each of the three sites ( 5 treatments and 3 replications per treatment). The temperatures from all three replications in each treatment, in each variety were averaged to arrive at the representative soil temperature and moisture.  Temperature was measured in degrees Centigade and the soil moisture was determined from the WaterMark moisture sensors by measuring the resistance and converting it to a matric potential in centibars (cb)or kilopascals (kPa), the higher the matric potential the drier the soil. 

Instrumentation for soil temperature and moisture measurements was completed in 2002 at all three locations. A database of records on soil moisture and soil temperatures has been maintained for all instrumented locations from the time sensors were installed to present. However, due to the sensitive nature of the instrumentation, interruptions in data collection were encountered periodically at all locations due to lightning, varmints, extreme temperatures, damage to connecting wires by weeding equipment and other problems. Therefore, all data records are not available at all locations for all dates. 

Irrigation 

            When the blueberry bushes were planted at each location, irrigation was installed to assure plant survival and growth. As blueberries are a shallow rooted plant, adequate moisture is needed in the upper 6 to 8 inches of soil.

            Drip tape was buried two inches deep in each plant row. The drip tape used was a 15 mil wall thickness, which is much heavier than the ten mil drip tape used on annual crops. 15 mil drip tape was projected to last for three to five years and would be replaced with a more substantial watering system as funding permitted. Water used for irrigation was city water at Spring Hill and Crossville. Water at Springfield was pumped from a farm pond and required sand filtration to remove biologicals such as algae and sediment. Plants were irrigated an average of two times per week as needed to supply a total of approximately two inches of irrigation water per week throughout the growing season.  A matric potential of 150 cb was used as a guide to begin irrigation.

            In the third year of the project, drip tape irrigation was replaced with 3/4 inch diameter plastic polyethylene pipe. The plastic pipe was suspended 18 inches vertically above the rows running the length of the rows. Treated posts (4" x 4" x 4') were concreted into the ground at the end of each row 24 inches deep with 24 inches of post above ground level. High tensile fencing wire was stretched between the posts at a height of 18 inches and tightened with a ratchet made for high tensile wire. The plastic irrigation pipe was attached to the high tensile wire with plastic cable ties approximately every six feet. Concrete rebar (reinforcing steel for concrete) posts one-half inch diameter and four feet long were installed vertically every 12 feet along the wire to support the wire and prevent wire sagging. The high tensile wire was attached to the rebar posts using plastic electric fencing wire connectors. Rebar posts were driven into the ground with approximately two inches of posts extending above the wire level.

            Drip emitters ( 0.5 gallons per hour) were installed in the plastic pipe at each plant approximately eight inches on each side of the plant to provide irrigation water in a zone around the plant. Initially the emitters dripped water from a height of 18 inches onto the row by the plants. Moisture sensor readings indicated that rows with plastic mulch were not receiving as much irrigation water as the other rows. Spray stakes were installed through the plastic mulch and small plastic tubes were extended from the emitters to the spray stakes to insure irrigation water reached the soil beneath the plastic mulch. 

Canopy Volume Index 

Since yield data was not collected until 2003, plant growth in the interim period was estimated using a canopy volume index for the years 2001 through 2003.  A canopy volume index was also estimated again in 2006 at the end of the yield trials.   

To determine the canopy volume index two photographs were taken of each plant at right angles to each other against a scaled background.  Square foot area for each of the two views was measured along with the height of the plant.  Dividing the area of each of the two views by the height, an average width and depth of the plant was obtained.  Then the height, average width, and average depth were multiplied to obtain an estimated canopy volume index.   

Mor ta lity 

Mortality of plants was of interest to determine if survivability was influenced by type of row cover or variety.  Mortality was tracked from 2000 through 2003 and then again in 2006.

Yield 

Yield data was collected at each of the three locations from 2003 through 2006.  The plants were transplanted in 2000 and began to bear in 2002 during the third year after transplanting.  However, to promote optimum plant establishment and growth the blooms and fruit were stripped from the plants at all locations until 2003.  At each location two representative average record plants were identified for each variety in each of the five treatments in all three replications.  The weights of harvested berries were recorded and an average weight per plant determined. 

Plant Netting

After the plants began to bear it was realized that birds were a significant problem.  In fact during some years the earlier ripening Bluecrop had no production due to birds eating the berries.  To get an idea of the extent of the problem in 2006 several of the Tiftblue plants at the HRREC were covered with netting in four of the treatments and compared to the un-netted record plants in the same treatment and variety.  

Results and Discussion 

Soil Temperature 

Soil temperature at the five inch depth was collected for the years 2003 through 2006.  As previously mentioned equipment problems with continued exposure to the elements, etc. resulted in less than complete data sets especially for the years 2005 and 2006.  However, the data sets for 2003 and 2004 are mostly complete and are representative of the entire period of the trail and were therefore selected for inclusion in this report.  The average yearly mean temperature varied only slightly from one treatment to the other at all three locations (Table 1), little more than one degree Centigrade for all treatments.  At HRREC and PREC the black treatment tended to be the warmest at the five inch depth.  It was initially thought that the white treatment would be cooler and might aid in keeping the soil cooler during the early part of the season and delay the plants from coming out of dormancy during warmer than normal periods in late winter or early spring.  However, temperatures with white mulch were not very different from the other treatments. 

Table 1.  Temperature comparisons for treatments at all three locations.

 When soil temperatures are considered for a month at a time the temperature differences become more pronounced as indicated in the Figure 1.  The graph depicts the fluctuation of soil temperature at HRREC (Springfield) for the month of July for 2004, which is during the peak of blueberry production especially for the Tiftblue variety.  From the representative graph the temperature can be seen to vary from 2 to 4 degrees Celsius. 

The treatment “none” has the greatest variation from night to day, getting the hottest during the day and the coolest during the night.  The “back” gets almost as hot as“none” hot but does not cool down as much.  “Organic” and “organic/black” are the least variable. 

 Figure 1.  Average soil temperature at Springfield in July 2004. 

Soil Moisture (Matric Potential) 

Table 2 lists the average annual soil moisture at all three locations for the years 2003 and 2004 for which the most complete data set exists.  A trickle irrigation system was installed at each of the three locations that could be turned on manually as required based on the soil moisture information.  The goal with respect to irrigation was to keep the matric potential less than 150 centibars.   

Although all treatments at each location normally had ample moisture, the “none” treatment, as expected, tended to be the driest, unless a rainy period was encountered, then it would tend to be wettest.  

Table 2.  Moisture comparisons for treatments at all three locations.

 Figure 2 depicts the moisture fluctuations at HRREC (Springfield) for July 2004.  Black/organic and white are the least moist.  Irrigation during July 2004 was not as frequent as desired and the “organic/black” treatment was the driest.  The combination of mulch and ground cloth acted as a barrier to precipitation.  “White” was the next driest, probably because the ground cloth has been painted with white latex paint to provide the white color.  The white paint tended to make the ground cloth less permeable to rainfall.  

Figure 2.  Average soil moisture at Springfield in July 20. 

The sharp drops in moisture levels correspond to precipitation or irrigation events.  The “none” responds more to precipitation, however the black typically maintains a higher moisture level especially in lower irrigation levels. 

Canopy Volume Index 

Each year the Tiftblue plants continued to outgrow Bluecrop plants at an increasing rate and were always significantly more robust (Table 3). 

Table 3.  Canopy volume index by variety and year for MTES, HRES, and PES.

Means within a column followed by the same letter are not significantly different at the 0.05 lever of probability, Duncan’s multiple range tests. 

The differences among treatments for both varieties averaged together were not very different the first year (2001); however, the “none” treatment had significantly less growth (Table 4).  In 2002 and 2003 black and “black/organic” had significantly better yield than the remaining treatments with the “none” treatment showing the least canopy development.   

Table 4.  Canopy volume index by treatment and year for MTREC, HRREC, and PREC combined.

Means within a column followed by the same letter are not significantly different at the 0.05 lever of probability, Duncan’s multiple range tests. 

Mortality 

  For Tiftblue at HRREC and MTREC there was no loss of plants for the black row covering treatment and for the black/organic at HRREC (Table 5).  Over the entire period the “none” treatment had the greatest mortality at the HRREC location but there was little difference at the other two locations.  For Bluecrop the “none” treatment again had the greatest mortality at HRREC and MTREC while white generally had the least mortality (Table 6).  Plants in the “none” treatment were smaller and less healthy in appearance at all locations. 

Table 5.  Mortality of Tiftblue blueberry plants at all three locations.

 Table 6.  Mortality of Bluecrop blueberry plants at all three locations.

Blueberry Yield

 MTREC(Spring Hill)

For Tiftblue at MTREC for years 2003 and 2004, the treatments black and black / organic were statistically the same and had the highest yield as can be seen in Table 7.  In 2005 and 2006 black, black / organic, and white were statistically the same with the highest yields.  For Bluecrop, yields were about one third of the Tiftblue yields, but the differences in Bluecrop yields for all five treatments were not so evident.  Some of this is due to bird problems that seem to be especially severe on Bluecrop which ripened about one month ahead of the Tiftblue.  2005 had the highest average yield when yield from all treatments are considered together.  The yield for 2006 was only the third highest yield.

 Table 7.  Blueberry Yields at Middle Tennessee Research and Education Center

    Note:  “No data” was due to bird problems.

 HRREC

 At the Highland Research and Education Center again Tiftblue significantly out-produced Bluecrop as indicated by Table 8.  Based on average for all five treatments 2004 was the best year and 2006 was second best.  For Tiftblue in 2003 and 2004 the black and black /organic treatments yielded statistically the same and were higher than the other three treatments.  In 2005 all treatments were statistically the same.  However, in 2006 organic, black, and black/organic were highest and were statistically the same.  

Table 8.  Blueberry Yields at Highland Rim Research and Education Center

Note:  “No data” was due primarily to bird problems. 

Since birds seemed to be a problem in the blueberries particularly at the Highland Rim and Middle Tennessee Research and Education centers, in 2006 it was decided to cover several plants with netting at the Highland Rim Research and Education Center to ascertain the degree of the problem with birds eating the barriers.  Since the Bluecrop portion of the crop was already picked over by birds, the Tiftblue variety was selected to cover several of the plants and compare yields to unnetted plants in the same replication and treatment (Table 9).  That data is graphically presented in Figure 3.  Yields from un-netted plants were reduced to roughly 38% to 20% of the netted yield. 

Table 9.  Actual yield for Tiftblue blueberries comparing netted and un-netted.

Figure 3.  Comparison of netted and un-netted blueberries at HRREC, 2006 by replication and treatment.

 PREC 

At PREC Tiftblue significantly out-yielded Bluecrop as can be seen in Table 10.  All treatments yielded significantly better for Tiftblue.  Bluecrop yielded poorly; however the black row covering treatment yielded the best in all four years and bare ground, in most cases yielded the least.  Based on average yield for all five treatments, 2004 was the best year followed by 2005, with 2006 being third. 

Table 10.  Blueberry Yields at Plateau Research and Education Center

Conclusions 

Yield for all years and all locations was greatest from the Tiftblue variety.  Birds were a very real problem especially for the Bluecrop variety, which ripened about a month earlier than the Tiftblue.  Concerning treatment, the highest yield came from “black” and “black/organic”. The canopy volume index (CVI) correlated well with yield for both treatment and variety.  CVI was highest for “black” and “black/organic” and for the Tiftblue variety.  Higher moisture levels correlated well with higher yields at Crossville  for “black” and “black/organic” during the production months of June, July, and August for both 2003 and 2004.  The lowest yields were for the “none” treatment which was also the driest.  The results at the other two locations did not present as clear a correlation.  Temperature had little correlation to yield at any location; however, the “black” and the “none” treatment tended to get the warmest.  Mortality of plants was greatest for the “none” treatment and least for the “black” treatment.  

Email all comments and suggestions to ghonea@utk.edu
Copyright © 1999 by The University of Tennessee. All rights reserved.

This research represents one season's data and does not constitute recommendations.  After sufficient data is collected over the appropriate number of seasons, final recommendations will be made through research and extension publications.