The block reward is composed of two parts, the block subsidy and the transaction fees. While the block subsidy is fixed respective to the height of a block, the transaction fees are a product of the blockspace market. If Bitcoin's exchange rate dropped precipitously, the profitability for miners would drop of course. Since each miner has a different situation, it would not be just one spot at which every operation suddenly becomes unprofitable, but rather a range, though.
As the exchange rate drops, some miners would stop mining. This would reduce the block cadence. Then, as transactions get processed more slowly, we'd see feerates of unconfirmed transactions go up as users vie to be included in the next block. Between the added revenue from transaction fees and the immense hardware investment miners have fronted, hopefully there would be enough incentive for some miners to continue mining. Then with the next difficulty adjustment, the difficulty would drop and mining profitability would increase again, likely causing some of the miners that had paused their operations to return.
Note that it would not take 2016 blocks for the first difficulty adjustment in most cases, as the exchange rate could drop at any point in a difficulty epoch. If the hashrate dropped significantly in the middle of an epoch, the first difficulty adjustment would likely not reflect the full drop, though.
You can see a similar effect around halvings: after the block subsidy dropped from 12.5 BTC to 6.25 BTC on May 12 2020, the difficulty dropped by 6.0% and 9.3% in the two following adjustments.