Soil seed banks – living seeds in the soil profile and on the soil surface – represent primary sources of regenerative potential and buffering capacity in dryland ecosystems. A recent study by Ma et al. (2021) presented the first conceptual framework that describes soil seed bank dynamics during state transitions from primary states, through degradation processes towards ecological state thresholds, to degraded alternate states. Our on-going study across an ecological state gradient on the Jornada Experimental Range provides a timely opportunity to test this conceptual framework. During the 2020 field season, 408 soil seed bank samples were collected across 17 sandy and shallow-sandy ecological sites, including primary black grama (Bouteloua eriopoda) grasslands, mesquite-encroached shrub-invaded grasslands, and alternate state mesquite (Prosopis glandulosa) shrublands. Paired shrub-island and interspace seed bank samples were collected alongside aboveground vegetation community data. Soil seed bank samples have been quantified using greenhouse emergence methods for 17 weeks to date. Preliminary results show that seed bank densities indeed follow similar patterns along a degradation gradient described by Ma et al. (2021). Across our sites seed bank densities range from approximately 600 seeds m-2 in primary grassland states where densities are evenly distributed between shrub-islands and interspaces, 900 to 3000 seeds m-2 in shrub-invaded grasslands where seed densities tend to be greater in interspaces than shrub-islands, and 200 to 1200 seeds m-2 in alternate state shrublands where seed densities concentrate in shrub-islands. Through further investigation of seed bank dynamics across our sites, and their shrub-island and interspace pairings, we will continue to develop our understanding of this important conceptual framework through empirical testing.