Sustained high yields of rice (Oryza sativa L.) are required in intensively cropped, irrigated systems to ensure food security and avoid expansion of cultivation into natural ecosystems. We examined whether good management practices and adapted cultivars could maintain high yields during 15 years (2001–2015) of continuous rice cultivation with three crops per year in the dry season (DS), early wet season (EWS), and late wet season (LWS). Results are for the highest-yielding fertilizer N rate of 195 kg N ha^–1 in DS and 90 kg N ha^–1 in EWS and LWS at the long-term continuous cropping experiment (LTCCE) in the Philippines. Our study addressed two critical questions remaining after previously reported research in the LTCCE: Can DS yield be sustained at >80 % of the climate- and genotype-specific yield potential, and can yield be increased in wet seasons? Management in DS, when cropping was synchronized with surrounding fields, controlled diseases and maintained yield at >80 % of simulated yield potential during 15 years. In EWS and LWS, when cropping was not synchronized with surrounding fields, tungro viruses were transmitted by insects from surrounding fields into the experiment. Management and adapted cultivars within the experiment did not prevent yield loss associated with reduced spikelet filling, which was attributed largely to tungro disease. In EWS, when planting was earlier than surrounding fields, yields were about 60–80 % of yield potential. In LWS, when planting was later than surrounding fields, yields declined to <65 % of yield potential. Enzyme-linked immunosorbent assay confirmed high infection in EWS and LWS with rice tungro spherical virus, which was not accompanied by symptoms of plant yellowing and stunted growth. In DS, when tungro was negligible, good management with adapted cultivars consistently minimized yield loss from nutrient deficiencies, soil problems, pathogens, and pests. However, good management with adapted cultivars could not maintain a small yield gap when tungro viruses were transmitted from infected surrounding fields into the experiment. In such case, closing the yield gap requires management interventions to reduce the pathogen and vector in surrounding fields. Negligible visible symptoms on most tungro-infected plants highlight the ease with which a biotic stress can be overlooked in experiments and growers’ fields.

精耕细作的灌溉系统需要持续高产的水稻（Oryza sativa L.），以确保粮食安全并避免耕作扩大到自然生态系统。我们研究了良好的管理规范和适应的栽培品种是否可以在连续15年的水稻种植（旱季（DS），湿季初（EWS）和湿季末（15）（2001-2015）期间保持高产（ LWS）。结果表明，在DS中最高产量的肥料氮含量为195 kg N ha ^–1，在氮肥中为90 kg N ha ^–1在菲律宾的长期连作实验（LTCCE）中，在EWS和LWS中进行了研究。我们的研究解决了先前在LTCCE中进行的研究后仍然存在的两个关键问题：DS产量能否维持在超过气候和基因型特定产量潜力的80％以上，并且在雨季可以增加吗？在DS中的管理，当与周围田地同步种植时，控制了疾病，并在15年内将产量保持在模拟产量潜力的80％以上。在EWS和LWS中，当耕作与周围田地不同步时，通果病毒被昆虫从周围田地传播到实验中。试验中的管理和改良品种不能防止与小穗充实减少有关的产量损失，这主要归因于通格罗氏病。在EWS中，当播种早于周围的田地时，单产约为单产潜力的60–80％。在LWS中，当播种时间晚于周围田地时，单产下降到小于单产潜力的65％。酶联免疫吸附试验证实水稻冬虫球形病毒在EWS和LWS中有很高的感染率，但并没有伴随植物发黄和生长发育迟缓的症状。在DS中，当通果草可忽略不计时，采用适应性栽培品种的良好管理可将营养缺乏，土壤问题，病原体和害虫造成的产量损失降至最低。但是，当通哥罗病毒从受感染的周围田地传播到实验中时，采用合适的栽培品种进行良好管理无法保持较小的产量缺口。在这种情况下 缩小产量差距需要采取管理干预措施，以减少周围田地中的病原体和病媒。在大多数受通果蝇感染的植物上，可见的症状几乎可以忽略不计，这凸显了在实验和种植者的田地中可以轻松忽略生物胁迫。