This may seem surprising since farm workers are defined as "essential" and most essential industries have seen lower than average job loss - and some like grocery stores and delivery drivers have seen no loss in employment at all. In addition, a large share of farm workers do not qualify for enhanced unemployment benefits that have discouraged some lower-wage workers from seeking employment. So why such a large decline? A loss of this size is probably caused by a combination of factors.
Covid rates are high in farm labor communities. Imperial County has by far the highest Covid case per capita in the state, and many other big agricultural counties have seen big increases and increasingly rank among the hardest hit locations in the state. Crowded living and working arrangements and poor access to health care make Covid a particularly difficult challenge for farm worker communities. Illness among workers and their families, concern about dangerous working conditions, and increased caregiving responsibilities can all reduce the supply of people willing and able to work farm jobs. Changes to the flow of people across the U.S.-Mexico border during the pandemic are also likely contributing to changes in labor supply, although H-2a agricultural visas have been exempted from new immigration restrictions for now.
In addition, there have likely been changes to farm labor demand as food demand has shifted from restaurants to home cooking. Restaurant closures have substantially reduced the demand for fresh produce, and other labor intensive specialty crops. It is likely that these crops are not being harvested or planted at typical levels which reduces labor demand. Grocery store shelves suggest most people at home are making eggs, pasta, and rice - not salad.
I am sure there are more contributing factors to this historic shift in the data.
Finally, I should mention two notes of caution. First, these initial employment estimates are based on samples and are subject to future revision as full employment tax filings are tabulated. Historically, revisions have seemed higher in the farm sector than other areas so caution is always warranted. But I have never seen a decrease anywhere this magnitude, so it seems very likely that me that there will still be very large losses even if the numbers are eventually revised higher. Second, the graph I displayed above is seasonally adjusted data - so the decrease it shows is actually a smaller than usual seasonal increase in the unadjusted data. Farm jobs have increased by 30,000 since March, but that is much much lower than the typical increase in spring, and California farm employment is estimated to be 20% lower (almost 100,000 fewer jobs) than last May.