Timing is everything

The timing of apple flowering, or flowering phenology, is dependent on temperature conditions from autumn to spring. Climate change may therefore affect flowering, exposing orchardists to problems such as:

• Inadequate chill – this can cause light and uneven flowering and a protracted flowering period leading to uneven maturity and poor fruit yields;
• Cross pollination asynchrony – when pollinators flower at different times to the main variety resulting in reduced fruit set;
• Changes to frost risk – may result in positive or negative outcomes for flowering.

If growers can predict flowering timing more accurately then adaptation strategies to ensure compaction of flowering, even fruit set and synchronised flowering of pollinating cultivars can be undertaken in the orchard.

Cold or hot, or a bit of both?

Growers know that floral buds need a combination of cold temperatures through winter and warm temperatures through spring to promote flowering. What isn’t known is how these temperatures combine to result in flowering.

A common model is to assume flowering occurs as a result of a winter chilling phase followed by a spring warming phase. The winter chill phase operates up until the chilling requirement is met, representing that dormancy has been broken.

This sets the beginning of the warming phase, which operates up until the heat requirement is met, predicting flowering. These two phases are treated independently and sequentially with the model often called the ‘sequential model’.

Recently, an alternate model has been described. This model is similar to the sequential model whereby meeting the chilling requirement is initially needed. However, subsequent to reaching this minimum chilling requirement, the heat needed to promote flowering is reduced depending on additional chill accumulated, or, over-chill. A minimum heat requirement is still required regardless of the over-chill amount. This model is known as the ‘chill overlap model’.

In their national project, Crossing the Threshold: Adaptation tipping points for Australian fruit trees, the researchers investigated which of these two models better explained flowering behaviour.

Flowering data

They collected Cripps Pink flowering data from growers and researchers to test the ability of the two models to predict full bloom dates over the range of climates found in Australia.

Two datasets were used to evaluate the models:

1  Grower data: A dataset comprising historical observations of full bloom dates recorded by orchardists from six growing regions.

2  Researcher data: A dataset comprising observations of flowering at four locations taken by researchers as part of this project.

For both models, winter chill was estimated using chill portions (CP) and heat was calculated using growing degree hours (GDH).

Model performance

Sequential model

In comparing the predicted and observed day-of-year of flowering in each location, the sequential model performed poorly. Of particular note were large deviations at the warm winter location (Donnybrook) and cool winter location (Batlow). The model predicted earlier flowering than observed for Donnybrook and later flowering than observed at Batlow.

Chill overlap model

Results using the chill overlap model were a marked improvement compared to the sequential model. This model did not show bias at cool or warm winter sites and recorded a statistical error of 6.9 days, a big improvement compared with the sequential model (14.7 days).

The new understanding

These results pose the question: Why does the chill overlap model better represent flowering timing in response to temperature?

The sequential model has fixed chill and heat requirements that do not interact. As such, average conditions tend to be modelled well, and cool and warm conditions are treated as deviations.

However, the chill overlap model allows for the interaction of additional chill accumulation to lower the heat requirement. In a cool winter site, such as Batlow, a lot of over-chill is accumulated. This then lowers the heat required to reach full bloom, accelerating the timing of full bloom. In a warm winter site, such as Donnybrook, less over-chill is accumulated and therefore more heat is required to meet conditions for flowering, taking more time.

Adapting to a changing climate

As the climate continues to warm, anticipating potential changes to flowering timing and therefore cross pollination synchrony and frost risk, is important to ensure orchard investments made now remain profitable. Growers need stable and reliable models that are based on known physiological relationships to predict flowering across different climates. Models that accurately predict flowering will provide growers with information to help them make decisions around changing cultivars, cross-pollinators, crop types or moving growing regions. The results from the research for Cripps Pink indicate the chill overlap model is suitable to conduct such analyses.

Source: Rebecca Darbyshire, Ian Goodwin and Susie Murphy White on behalf of Apple and Pear Australia Ltd.