In response to warming minimum temperatures, Avicennia germinans is encroaching poleward on the Texas Gulf Coast (TGC) into saline marshes dominated by Spartina alterniflora and Batis maritima. Increased Avicennia cover provides greater protection from soil subsidence and shoreline retreat. However, intense freeze disturbances cause widespread mangrove mortality reversing succession, and increasing the risk of soil subsidence and shoreline retreat due to the loss of below-ground biomass. We conducted a “natural experiment of opportunity” to measure below- and above-ground biomass allocation in Avicennia recovering from catastrophic disturbance caused by the 2021 Winter Storm Uri at sites along a freeze-disturbance gradient across the South and Central TGC. Port O’Connor (28.46°N) was the most severely affected site, Cohn Preserve on Mustang Island (27.71°N) was moderately affected, and Laguna Atascosa National Wildlife Refuge (26.35°N) was minimally affected (min. temp. °C ~ -9.0, ~ -7.4, ~ -5.5 respectively). A second freeze event occurred in December 2022 that severely affected Port O’Connor and moderately affected Cohn Preserve (min. temp. °C ~ -6.6, ~ -5.8 respectively). In an additional methods experiment, we quantified differences in root productivity in in-growth cores containing either peat moss or local substrate at each site. Multiple root ingrowth cores were inserted near the canopy edge of isolated Avicennia shrubs (n=6; 5 at Port O’Connor) at the three sites and collected at 4-month intervals (total cores = 102). Root productivity (g * (m-2 day-1)) assessed in a one-way ANOVA and Tukey multiple comparisons, increased with increasing freeze-disturbance effects (F2,28 = 6.386, p<0.01, Port O’Connor: mean = 0.192, sd = 0.188, Cohn Preserve: mean = 0.065, sd = 0.081, Laguna Atascosa: mean = 0.047, sd = 0.089). Using the below-and above-ground relative growth rates to assess the root:shoot biomass allocation ratio, we found an increasing ratio (greater roots to shoots) with increased freeze disturbance (one-way ANOVA and Tukey HSD tests, F2,12 = 6.049, p<0.05, Port O’Connor: mean = 0.191, sd = 0.118, Cohn Preserve: mean = 0.061, sd = 0.066, Laguna Atascosa: mean = 0.015, sd = 0.01). Further, we found no clear trend in quantity or variability in root productivity between native and peat moss substrate types in root ingrowth cores for root biomass at any site (F2, 53 = 0.021, p=0.8). However, peat moss ingrowth cores did consistently have less root necromass after the Dec. 2022 freeze suggesting better survival or lower turnover. Peat moss ingrowth cores contained lower quantities of live root biomass indicating higher rates of root mortality or a stunting effect on root productivity post-second freeze. This finding suggests that cumulative impacts of two freezes occurring less than two years apart are greater than the effects of individual freezes. Increasing root productivity with greater freeze disturbance suggests that recovering standing root biomass may be important for the recovery of above-ground biomass in freeze-affected Avicennia. Rapid recovery of below-ground biomass will also contribute to ameliorating rates of soil subsidence and shoreline retreat. Lastly, we found that peat moss is a viable substrate type for future root ingrowth studies if the focus of the study is on total root biomass. However, if a second freeze event occurs during the root ingrowth study, there may be unequal effects between peat moss and local substrate.