Hydraulic fracturing has been proven useful for improving the gas production efficiency of low-permeability hydrate reservoirs. However, no specific fracturing fluid system exists for marine hydrate reservoirs. Applying conventional water-based fracturing fluids in low-temperature (<293.2 K) marine hydrate reservoirs can lead to difficulties in gel-breaking and hydrate instability caused by external additives. In this study, we developed two water-based fracturing fluid systems that exhibited good gel-breaking performance at low temperatures (278.2–288.2 K), one comprising ammonium persulfate (APS), ferrous sulfate (FeSO4), and citric acid (CA), the other APS and triethanolamine (TEA). In addition, we investigated the influence of crosslinking and gel-breaking reactions of the fracturing fluids on the methane hydrate phase equilibrium. The breaking time of the APS-FeSO4-CA system at 278.2K, 283.2K, and 288.2K was 30 min, 75 min, and 120 min respectively. On the other hand, the APS-TEA system accomplished the same task in 3 h, 6 h, and 10 h respectively. Furthermore, methane hydrate phase-equilibrium data showed that the fracturing fluid's three states, base fluid (BF), fracturing fluid gel (FFG), and gel-breaking liquid (GBL), all exerted an inhibitory effect on the phase equilibrium, with the hydrate suppression temperature (ΔT) ranging from 0.4 to 1.0 K. The differences in their inhibitory effect were primarily related to the micromorphology of the hydroxypropyl guar gum (HPG) molecules and the number of active hydroxyl (-OH) groups in the solution. Cryo-scanning electron microscopy (SEM) measurements revealed that the micromorphology of HPG in BF, FFG, and GBL transformed from regular ribbon-like structures to 3-D network structures and then to hole-containing, short ribbon-like structures. FFG formed a regular 3-D network structure after adding crosslinking agents to the BF. The ΔT values reversed from FFG < BF to FFG > BF when the HPG concentration increased from 0.3 wt% to 0.5 wt%, possibly due to the difference in the HPG molecules' water adsorption capacity before and after crosslinking. The broken HPG molecules with OH− groups in the solution were transformed into residues after gel-breaking, decreasing the GBL's inhibitory effect. Therefore, our results indicated that micromorphology variations could cause differences in the HPG molecules' water adsorption capacity, ultimately leading to ΔT differences. Based on our findings, the gel-breaking fracturing fluid systems we developed demonstrated potential for optimizing hydraulic fracturing in marine hydrate reservoirs.