Summary: In 2014, the rust fungus Puccinia komarovii var. glanduliferae, native to the foothills of the Himalayas from Kashmir to Western Nepal, became the first fungal agent to be released into Europe for the classical biological control of a non‐native weed. The target, Impatiens glandulifera or Himalayan balsam, is a prolific invader of riparian habitats in Europe and North America. During the period 2015–2018, a strain of the rust from India was released at a total of 36 sites, in 17 counties in England and Wales. There was limited field infection in 2015 and inoculation experiments conducted under controlled conditions revealed significant variation in the susceptibility of plant populations to the rust, with some showing immunity. Subsequently, a second strain of the rust from Pakistan was released in 2017 and was found to infect a different cohort of Himalayan balsam populations. The rust mass production methodology, and field inoculation and monitoring protocol, are detailed, and plants were tested for susceptibility to both rust strains prior to field release. Levels of foliar infection at selected sites in 2017 and 2018 are presented, as well as seedling infection rates in the spring of 2019. The results show that the rust is able to overwinter and establish populations in stands of Himalayan balsam in England. The issues involved with measuring the impact of the biological control agent are discussed.

Impatiens glandulifera, or Himalayan balsam, is a prolific invader of riverine habitats. Introduced from the Himalayas for ornamental purposes in 1839, this annual species has naturalised across Great Britain (GB) forming dense monocultures with negative affects across whole ecosystems. In 2006 a programme exploring biocontrol as an alternative control method was initiated and to date, two strains of the rust fungus Puccinia komarovii var. glanduliferae have been released. To better understand the observed differences in susceptibility of GB Himalayan balsam stands to the two rust strains, inoculation studies were conducted using urediniospores and basidiospores. Experiments revealed large variation in the susceptibility of stands to urediniospores of the two rust strains, with some resistant to both. Furthermore, the infectivity of basidiospores was found to differ, with some stands fully susceptible to the urediniospore stage, being immune to basidiospore infection. Therefore, before further rust releases at new sites, it is necessary to ensure complete compatibility of the invasive stands with both urediniospores and basidiospores. However, for successful control across GB it is essential that plant biotypes are matched to the most virulent rust strains. This will involve additional strains from the native range to tackle those biotypes resistant to the strains currently released.

Background : The red-stemmed parrot’s feather, Myriophyllum rubricaule , was recently described as a new plant species following confusion over its identity and more broadly within the genus. In a molecular study using Myriophyllum samples collected as part of a classical biological control programme against M. aquaticum in 2021, two populations of a different species were found in Great Britain (GB). This species was thought likely to be M . rubricaule ; however, it had not been verified by molecular analysis, nor was this species reported to exist in the wild at that time. Methods : Sequence data from the rDNA-ITS, rbcL , trnH-psbA , trnK-matK ( trnK 5 intron, matK coding-region and trnK 3 intron) and trnL-trnF regions were compared between suspected M . rubricaule collected from GB and other Myriophyllum species, including M . rubricaule from the Netherlands (NL) and Chile. Phylogenetic analyses of the Myriophyllum samples using combined datasets of four regions (rDNA-ITS, rbcL , trnH-psbA and matK coding-region) and five regions (the rDNA-ITS, rbcL , trnH-psbA , trnK-matK and trnL-trnF ) were undertaken using a maximum likelihood method. Results : The sequencing revealed that the two field-collected samples and one from an aquatic plant nursery in GB were identical to M . rubricaule from NL. Phylogenetic analysis based on the four multilocus datasets showed the three GB samples grouped with M . rubricaule from NL and Chile and were clearly separated from M . aquaticum . Further phylogenetic analysis based on the five regions also showed that the three GB M . rubricaule were distinct from other Myriophyllum species, confirming their identification as M . rubricaule . Conclusion : This study has confirmed the first occurrence of M . rubricaule in the wild in GB through molecular identification. The potential invasive risk of this species given its apparent establishment at sites in GB and close relationship to M . aquaticum is also discussed.