Managing Cropload with the Pollen Tube Growth Model In Organic Apple Orchards

Cropload Management in Apples

By nature, apple trees often bear heavy crops of fruit, resulting in small, poor quality fruit and reduced bloom in the following year.  In commercial orchards, therefore, it is important to "thin" flowers or small fruitlets and reduce the cropload.  Non-organic growers often spray chemical thinning agents to manage cropload.  Without access to synthetic chemical thinning agents, organic apple growers often thin by hand, which is labor-intensive and ineffective. 

Alternatives to hand-thinning in organic apple orchards include fruitlet thinning sprays such as Protone, which can be sprayed when fruitlets are approximately 10 mm in diameter, and bloom thinning sprays which can be sprayed during bloom to kill flowers and reduce fruitset.  Several bloom thinning spray materials are available to organic growers:

1. Liquid lime sulfur (LLS) is a common bloom thinning agent, sometimes used together with fish oil or another type of spray oil. A downside of liquid lime sulfur is its high toxicity to handlers and applicators - in concentrated form the product is highly caustic and corrosive and can cause irreversible eye damage and skin burns. LLS pesticides are marked with the Danger signal word which denotes highly toxic products.  We have not used LLS in our orchard.
2. Potassium bicarbonate has been shown to be an effective bloom thinner in European trials in Poland and Switzerland. Although this practice has not been widely used in the U.S. to my knowledge, there is some discussion of it on the Holistic Orchard Network. We have used potassium bicarbonate in our orchard in past years and it has appeared moderately effective.  Typically we have used 2-3 applications of potassium bicarbonate applied at 15 lbs/100 gallons water per acre.
3. Regalia (a plant based OMRI-listed plant disease resistance inducer) has been shown to be an effective bloom thinner when mixed with spray oil in several studies from Virginia. However, after those studies were conducted the manufacturer modified the formulation of Regalia, and the manufacturer believes that the newer formulation is not an effective bloom thinner. We and other growers have observed that the current Regalia formulation when applied together with spray oil causes obvious visible petal burn and accelerates petal drop and seems to provide some amount of fruit thinning as well.  Dan Kelly, an organic apple grower at Blue Heron Orchard in Missouri reported excellent thinning when he sprayed Regalia (presumably new formulation) with oil according to the PTGM.
4. Various sprays oils have been shown effective for bloom thinning in Slovenia and China. To my knowledge, spray oils are not often used alone for thinning in the U.S., although they are often mixed with other thinners such as LLS or Regalia.

Pollen Tube Growth Model

The Pollen Tube Growth Model (PTGM) uses weather data to time bloom thinning sprays.  A detailed description of the PTGM was published in the New York Fruit Quarterly in 2018.  Here is a brief overview:

Apple fruits develop from flowers; a bee deposits pollen on the stigma of a flower, and pollen grains germinate into pollen tubes which grow into the style of the flower until they fertilize an ovule (a seed precursor) at the bottom of the style. Only flowers with fertilized ovules form fruit.  Most bloomtime thinning sprays presumably work because they prevent ovule fertilization by killing floral organs before fertilization occurs.

The premise of the PTGM is that it is key to begin bloom thinning sprays when the ideal number of flowers have already been fertilized and additional fertilization is not desired, and then continue regular sprays until bloom has ended. If the first thinning spray is made too early, cropload will be low; if it is made too late, cropload will be high. The pollen tube growth model uses weather data and farmer observations of bloom timing to predict when thinning sprays should be made.  The model is based on detailed experimental measurements of how fast pollen tubes grow at various temperatures.  These growth rates vary by variety, and the model has been calibrated for use with seven common commercial varieties: Fuji, Gala, Golden Delicious, Granny Smith, Honeycrisp, Red Delicious, and Pink Lady. 

Although apple bloom occurs quickly in a relatively short period of time, flowers do not all open at that same time.  To use the model, the grower must determine (by educated guess) when enough flowers are open on the trees to produce the desired crop in a particular orchard block.  For example, if the grower wants 100 fruits per tree, the grower would estimate when 100 flowers per tree are open.  The goal is then to time sprays to allow ovule fertilization in those 100 flowers, but not to allow fertilization in any flowers which open after those 100 flowers.  The grower sets the start time for the PTGM on NEWA when the desired number of flowers per tree are open.  In addition, the grower measures average style length in the orchard block of interest and enters that information into the NEWA model.  From thereon, the model estimates, based on temperature and apple variety, the distance that pollen tubes have grown into the flower styles.  The model displays the tube growth as a percent of the average style length.  For example if the average style length in a block is 10 mm, 20% pollen tube growth means that pollen tubes have grown 2 mm into the styles.  The goal is to make the first thinning spray at 100% pollen tube growth measured from the start time of the model.  At this stage, the early-opening flowers from which fruit set is desired should be fertilized, but the thinning spray should prevent flowers which opened subsequently from being fertilized and setting fruit.  A second spray should made before 100% pollen tube growth since the time of the first spray.  Additional sprays should be made if needed according to the same principle until the end of bloom.

The PTGM has not gained widespread use in the eastern U.S.  Non-organic apple growers already thin successfully by spraying synthetic growth regulators in the weeks after bloom and many have not been motivated to try bloom thinning based on the PTGM. Organic growers generally lack effective post-bloom thinning sprays and often hand-thin fruitlets, but hand thinning must be completed in a brief window of time to provide significant benefits and it is extraordinarily labor-intensive. Some organic growers have used bloom thinning sprays with inconsistent results, but the PTGM has not been widely used to time these sprays, although many commercial apple growers are comfortable using similar computer models which predict insect pest and disease development based on weather data.

Several weaknesses to the PTGM are apparent.  First, the model was developed for seven specific varieties.  Some of these, such as Granny Smith or Pink Lady, are very rare in organic apple orchards of the North Central region, whereas common organic varieties such as Liberty are not included in the model.  Varieties do differ significantly in pollen tube growth rates (see graph), and so it is difficult to apply the PTGM to varieties other than the seven varieties specifically included in the model. 

Second, the model assumes that flowers are pollinated immediately upon opening and that bloom thinning sprays will contact and kill all open flowers.  Both of these assumptions are probably untrue in most real-world situations. 

In addition, growers with many apple varieties are concerned about the time required to measure style lengths, run the model, and make separate thinning sprays for different varieties. 

Objectives

For our study in 2022-2023, we opted to use Regalia and oil as a bloom thinner because of the previous successful reports for this combination.  None of the varieties in our orchard is specifically included in the PTGM.  Therefore we decided to use the model for Golden Delicious, a variety with relatively average pollen tube growth rates compared to the other varieties in the model. We also tried different timings of the initial thinning spray because delaying the initial spray later than what is recommended by the PTGM will address growers' concerns about slow pollination and and also concerns that the actual varieties in this study may have slower pollen tube growth rates than Golden Delicious.

The goals of our project were (i) to measure the effectiveness of Regalia and the PTGM for crop thinning in varieties not explicitly included in the PTGM and using three variations on initial spray timing and (ii) to measure the time required to implement the model. 

Weather Station

In spring 2022, we installed a Rainwise AgroMET Weather Station in our orchard.  The weather station collects hourly data on temperature (as well as other weather data).  Our weather station, as well as the weather stations from many other orchards, are connected to the NEWA network (https://newa.cornell.edu/).  The NEWA system uses this orchard-specific weather data to run the pollen tube growth model (PTGM) and other models related to cropload management and insect and disease control.  The NEWA website has a tutorial video on how to use the pollen tube growth model.

Experimental Design

We conducted this research in our certified organic apple orchard. The overall apple orchard was 2.5 acres in size in 2022.  All trees are on dwarfing roostocks (mostly G.41, G.11, and G.16), and spaced 6' apart in the row.  Between row spacing varies from 10.5' to 12', for a density of 605-690 trees per acre.  We use hardwood bark mulch to maintain a weed-free strip 5-6' wide in the tree row.  Trees are individually staked using Best Angle 10' metal stakes and irrigated with drip irrigation.  Our target yield varies by variety but is generally 30-60 lbs per tree (approximately 500-1000 bushels per acre).  Since 2017, actual annual yields per bearing acre have ranged from 525-925 bushels/acre.

In spring 2022 we selected ten blocks of trees to include in the experiment, each consisting of 35-50 similar-aged trees of a single variety. The blocks included seven varieties: Liberty, Goldrush, Winecrisp, Williams Pride, Pristine, Priscilla, and Sansa, all disease-resistant varieties common in organic orchards. We included multiple varieties because varieties differ in optimal cropload and response to thinning treatments. Blocks had been planted between 2012 and 2019.   Young and mature trees can differ in their optimal cropload and response to thinning treatments.

Table. Blocks Used in Experiment

Within each block, we selected 16 trees for the experiment, avoiding trees which obviously differed greatly from the others in bloom density in spring 2022.  We divided the sixteen trees into four plots, which were randomly assigned to one of four treatments.  One treatment was a control treatment, with no early bloom thinning sprays applied (by "early bloom" we mean sprays during spur bloom - see the results section for a discussion of late bloom on 1-year old wood).  In the other three treatments, we applied early bloom thinning sprays with a tractor-mounted Pakblast 50 airblast sprayer with Regalia and JMS Organic Stylet Oil (both at 1% volume:volume concentration) according to the Pollen Tube Growth Model for the variety Golden Delicious.  Spray volume was 100 gallons per acre, which is a relatively high volume spray in our setting, and which we expect to provide excellent spray coverage, especially during bloom when the canopy is still small.  This Regalia concentration has worked well in published research results. The three spray treatments differed in the timing of the initial thinning spray: in one treatment we aimed to apply the initial spray at the 100% pollen tube growth stage, as recommended by the PTGM, in the second treatment at the 125% stage, and in the third treatment at the 150% stage.  The rationale for sprays at 125% and 150% was to try more "conservative" thinning sprays and to allow for more flower fertilization if, for example, a variety had slower pollen tube growth rates than Golden Delicious or if insect pollination of flowers was slow.

Pollinators

Our orchard is primarily surrounded by corn and soybean fields.  Although we have established pollinator habitat on our 12 acre farm, native pollinator populations are modest.  Therefore we have purchased bumblebee colonies to ensure adequate pollination.  In both 2022 and 2023 we purchased two bumblebee quads from Koppert Biological Supply, and these bees foraged on two acres of bearing apples plus an additional 0.25 acres of gooseberries and 0.25 acres of currants.  (Gooseberries in particular bloom simultaneously with apples and are very attractive to foraging bumblebees.)

Data Collection

We measured the following data: