**Background:** Citrus Huanglongbing (HLB), caused by 'Candidatus Liberibacter asiaticus' (CLas) and vectored by Diaphorina citri, is the most devastating disease of citrus worldwide. In Guangdong Province, China, HLB has been present since at least the 1870s and currently affects approximately 59,700 hectares (about 25% of total citrus planting area). Despite decades of research, no studies had systematically investigated annual HLB prevalence within specific orchards over multiple years in the last 30 years. This study aimed to characterize the space–time dynamics of HLB in six orchards under different management regimes.
**Methods:** Six orchards in Zhaoqing, Huizhou, Guangzhou, and Meizhou cities (Guangdong Province) were surveyed annually from 2019 to 2021. Orchards ranged from 5 to 20 hectares. Two orchards (1 and 2) used conventional management; three orchards (3, 5, and 6) used comprehensive control; orchard 4 had both conventional and comprehensive management sections, including a screen house region (A1). Visual inspections were conducted in autumn and winter by trained investigators, with PCR confirmation (RT-PCR using CLas4G/HLBr primers and HLBp probe) of at least 20 symptomatic and 20 asymptomatic leaf samples per day. Disease severity was scored on a 0–3 scale. Spatial data were recorded using the 'two-step-path' app and analyzed with LocaSpaceViewer4. CLas genetic diversity was assessed using six gene loci (three prophage loci, MITE region, and two short tandem repeat genes). Fruit quality parameters (vitamin C, TSS, TA, fruit weight, peel color) were measured from 20 healthy and 20 diseased trees inside and outside the screen house.
**Key Results:** Comprehensive management dramatically reduced HLB incidence. In two 'Gonggan' orchards in Deqing, the conventionally managed orchard saw HLB rise from 15.58% (830/5326) to 30.65% (1289/4205) over three years, while the comprehensively managed orchard rose from 0% to only 0.18% (21/11,623). The fingered citron orchard (orchard 1) under conventional management increased from 6.49% (1165/17,964) to 15.78% (2753/17,447) before being destroyed. In orchard 4, the screen house region (A1) had a three-year average HLB incidence of 1.05 ± 0.48% compared to 6.58 ± 2.85% in the open-air region (A2). The disease index was 4.36 in A1 versus 8.70 in A2. Yield of HLB-affected trees in the screen house was significantly higher (15.37 ± 0.86 kg/tree) than outside (4.56 ± 0.21 kg/tree, p < 0.001). No significant differences were found in fruit quality parameters (vitamin C, TSS, TA, TSS/TA ratio, fruit juice rate) or morpho-physiological parameters between screen house and open-air fruits, though peel redness was lower inside the screen house (32.51 ± 1.02 vs. 36.96 ± 0.57, p = 0.003). Spatial analysis revealed that HLB incidence was higher in northern sections of flat orchards and at field edges, with a 'marginal effect' observed. Diseased trees were clustered, with scale 3 trees predominantly at region edges. CLas genetic analysis showed that populations from the same city clustered together, but geographically distant Huizhou and Meizhou populations were genetically similar.
**Clinical Implications:** This study provides strong field evidence that comprehensive management—combining disease-free seedlings, timely removal of infected trees, and psyllid control—can achieve 95.53–99.34% relative control efficacy, consistent with prior reports. The 'three-pronged' approach remains the most effective strategy. Screen houses effectively exclude D. citri vectors, reduce HLB spread, and maintain fruit quality while increasing yield of infected trees, making them a viable option for high-value citrus production. The spatial patterns identified (north-to-south spread, edge effects, clustering) can inform orchard layout and monitoring strategies. The finding that CLas populations from distant locations can be genetically similar suggests that seedling movement may contribute to disease spread, highlighting the importance of nursery stock certification.