Methods - Seasonal labour demand

Agricultural crops and products are not like manufacturing. The timing of seasonal growth cycles and when produce is available for harvest is strongly influenced by the weather. This is particularly the case for intensive farming like aquaculture, fruit and vegetables. This includes mangoes, one of the most valuable agricultural products in the 'Seasonal labour demand case study', worth about $50 million per year. Labour demands during the peak weeks when mangoes need to be picked and packed each year are substantial (equivalent to more than 1 in 50 workers resident in the Seasonal labour demand case study), but the exact weeks when workers are most needed varies from year to year.

To assist rural producers in meeting the challenge of seasonal labour demands Australia grants special visas to allow workers from select nearby countries to do seasonal work in Australia each year. However, in speaking to local grower representatives during this project, the inflexibility of current visa conditions was raised as one of their major concerns. For example, under current arrangements, return air tickets must be booked before visa applications can be lodged. In practice this means that employment periods have to be fixed in advance of when the fruiting season and labour requirements can be forecast. The consequence is that in some years there are few pickers available during periods of high demand and in other situations, pickers may be available but the fruit is not ready to be picked.

Historical daily weather data were used to estimate year-to-year differences in when mango fruiting and ripening occurred in the 'Seasonal labour demand' case study and when labour was required to pick and pack fruit. For each year, the date for the onset of flowering was estimated from a cold trigger based on consecutive days of minimum temperatures below threshold levels (three days at or below 15^oC, or five days at or below 18^oC). Once the flower budding date was calculated, fruit development and ripening was estimated using a ‘heat sum’ approach (NT Government 2013). This involved keeping a running total of heat accumulation units until a threshold value was reached, where the amount of heat units added each day were calculated by subtracting 12^oC from the daily average temperature. Three mango varieties were considered with different heat sums to ripening as follows: Kensington Pride (1600), Florida Types (1680), and R2E2 (1800) (ordered from the earliest to latest ripening variety).

Seven weather stations, covering the mango-growing locations in the Darwin region, were used in the analyses (station numbers 14041, 14044, 14058, 14160, 14207, 14215 and 14247; see table below). Historical daily weather data for each location were downloaded from the SILO web delivery platform¹⁰ and used to estimate peak fruiting dates for each combination of mango variety and location for each year from 1960 to 2017 (Jeffrey et al. 2001).

Weather stations used to calculate fruiting dates for mangoes in the Seasonal demand case study

Based on industry data for mango production for the area (AMIA 2019), labour demand was calculated as a six week period around the estimated peak fruiting date for each variety with two weeks of peak demand in the middle, when most fruit needed picking, a week either side of reduced labour demand, and another week either side where a smaller amount of fruit was available.

For the insight featured in the case study (Seasonal labour case study), the data shown are for the Kensington Pride mango variety (the most common variety grown in the study area) using weather data from Middle Point weather station 14041 (used by the mango industry as a representative indicator for the Darwin region). Data for other combinations of variety and weather station showed a similar variation in peak fruiting and labour demand between years (noting that some varieties consistently ripen earlier than others).

More detail on methods is provided in the methods report (Stokes et al. 2019).