Vegetation restoration in deserts is challenging due to these ecosystems' inherent fragility and harsh environmental conditions. One approach for active restoration involves planting native species, which can accelerate the recovery of ecosystem functions. To ensure the effectiveness of this process, carefully selecting species for planting is crucial. Generally, it is expected that a more diverse mix of species in the plantation will lead to the recovery of a greater number of ecosystem functions, especially when the selected species have complementary niche traits that facilitate maximum cooperation and minimize competition among them. In this study, we evaluated the planting of two native species from the hyper-desert of Taklamakan, China, which exhibit marked morpho-physiological differences: a phreatophytic legume (Alhagi sparsifolia) and a halophytic non-legume (Karelinia caspia). These species were grown in both monoculture and intercrop communities. Monoculture of the legume resulted in the highest biomass accumulation. Intercropping improved several ecosystem functions in the 50 cm-upper soil, particularly those related to phosphorus (P), carbon (C), and sulfur (S) concentrations, as well as soil enzyme activities. However, it also increased soil sodium (Na+) concentration and pH. Halophyte monocultures enhanced ecological functions associated with nitrogen concentrations in the upper soil and with P, S, C, and cation concentrations (K+, Ca2+, Mg2+, Cu2+, Fe2+, Zn2+, Co2+, Ni2+), along with enzyme activities in the deep soil. It also maximized Na+ accumulation in plant biomass. In summary, we recommend legume monoculture when the primary goal is to optimize biomass accumulation. Conversely, halophyte monoculture is advisable when the objective is to extract sodium from the soil or enhance ecosystem functions in the deep soil. Intercropping the two species is recommended to maximize the ecosystem functions of the upper soil, provided there is no salinization risk. When planning restoration efforts in desert regions, it is essential to understand the impact of each species on ecosystem function and how complementary species behave when intercropped. However, these interactions are likely species- and system-specific, highlighting the need for more work to optimize solutions for different arid ecosystems.