In order to study the water temperature distribution characteristics of crab ponds under micro flow recirculating ecological culture mode, a multi-layer water temperature real-time monitoring system based on an IOT (Internet of things) cloud was  designed. Six temperature sensors are fixed to collect the water temperature of different layers and upload them to the IOT cloud platform, which is designed to manage the uploaded data from the different monitoring points and sensors. The  results of the experiment  that, the system can  monitor thewater temperature of different water layers in the pond in real time, at every 5 minutes intervals. The micro flow makes the extreme value of the water temperature 2 °C lower, and the temperature difference between the upper and lower layers is 2.5°C smaller, which can effectively reduce the risk of water temperature stratification. Meanwhile, the micro flow recirculating mode can effectively improve the pond water temperature balance level, resulting in an average value of water temperature balance of 0.893 for the whole day, which is 10.17% higher than that in the traditional mode (0.811). The experiment verified the feasibility of the system and the effectiveness of the micro flow recirculating culture mode.