Managing woody species that have encroached into grasslands and can resprout following top-kill has become an increasing problem. Treatments that yield complete mortality (i.e., “root-kill”) such as mechanical grubbing are often too costly at landscape scale. Less expensive treatments that only top-kill are problematic if treatments are not applied frequently to maintain regrowth suppression. More sustainable strategies are needed. Previous research found that a low rate (0.28 kg·ha-¹) of clopyralid-only herbicide root-killed some honey mesquite (Prosopis glandulosa) but importantly caused partial top-kill (PTK) in surviving trees that maintained apical dominance and prevented basal sprouting. This management approach, referred to as “vertical brush sculpting,” increased herbaceous productivity compared with top-killed mesquite with dense resprouting. Here we determined if higher rates of clopyralid could increase mesquite root-kill and maintain the PTK physiognomy in surviving plants. We evaluated effects of clopyralid at 0.28, 0.42, and 0.56 kg·ha-¹ (C28, C42, and C56, respectively) and a mixture of clopyralid (0.28 kg·ha-¹) + triclopyr (0.28 kg·ha-¹) (C + T) on mesquite physiognomy. Averaged over two sites at 6 yr post treatment (YPT), root-kill was 0%, 20.0%, 42.8%, 58.9%, and 65.6% in untreated, C28, C42, C56, and C + T, respectively. Among surviving trees, percent PTK with no basal sprouts was greatest in clopyralid-only treatments; percent top-killed with basal sprouting was greatest in C + T. Grass production, measured in untreated, C28, and C + T, ranged from 110% to 215% greater in treated than untreated at 2 YPT. We developed a model to project stand-level mesquite competitive impact on grasses in all treatments over 30 yr. Mesquite competitive impact values were lowest in C42 and C56 due to high root-kill, and most survivors had noncompetitive PTK physiognomy. We viewed C42 as the best option due to lower cost. Vertical brush sculpting offers a method to reduce woody competition with grasses yet maintain landscape heterogeneity.