Roseau cane (Phragmites australis (Cav). Trin. ex Steud.) is the dominant plant species of the Mississippi River Delta in Louisiana, USA, and protects marsh communities from erosion and storm-related impacts, maintaining shipping channels and oil infrastructure. Widespread dieback and thinning of P. australis were noted in the Mississippi River Delta in the fall of 2016. Invasive populations of the roseau cane scale (Nipponaclerda biwakoensis Kuwana), which is native to Asia, were found at outbreak levels in the stands. Subsequent research has implicated N. biwakoensis in the diebacks. To provide a safe, cost-effective, and long-term approach for the management of this invasive scale, classical (or importation) biological control methods are being considered. As the first step to developing biological control, research evaluating the natural enemies of the N. biwakoensis in the native range is necessary. This study investigated the population dynamics of the N. biwakoensis and its associated parasitoids on P. australis in Taiwan from July to November 2019 at five study sites. The scale densities increased over the season and peaked in September (85.77±3.63 scale per meter stem). An average of 14.50% adult female scales was parasitized and parasitism was highest in July. The parasitism rate of the immature scales was overall low (2.25%). Four endoparasitoid wasp species were collected and identified as Astymachus lasallei, Boucekiella depressa, Neastymachus japonicus, and Aprostocetus sp. Astymachus lasallei was the dominant wasp species parasitizing immature scales and N. japonicus was the dominant species found parasitizing adult female scales. Multiparasitism frequently occurred between A. lasallei and B. depressa and the latter has been reported as a hyperparasitoid. Overall, this study provides information on the identity and role of parasitoids of N. biwakoensis in its native range, which will aid in developing a potential classical biological control as a management tool for the invasive N. biwakoensis.

Study sites

The study was carried out during the summer and fall, when first- and second-generation scales were present. Field sampling was performed four times (July 08, August 19, September 28, and November 09, 2019) at five different sites along the western coastline of Taiwan. The collection sites were selected so that they were geographically separated by at least 20 km except for the two sampling sites in Taichung City. The study sites were usually wet during the rainy period (May to September). Mean water depth across the sites at the time of sampling was 5 cm with an average of 3 ppt salinity. The study sites included Daan District, Taichung City (24.351813, 120.567663), Qingshui District, Taichung City (24.321669, 120.551705), Shengang Township, Changhua (24.158311, 120.459115), Fuxing Township, Changhua County (24.027550, 120.398925), and Mailiao Township, Yunlin County (23.747655, 120.241417).

Field sampling

Phragmites specimens were sampled along 30-meter transect lines. Each transect was divided into three intervals from 0-10 m, 11-20 m, and 21-30 m. Ten samples of scale-infested plants were randomly collected within each interval. The stems were cut just above the ground. Clipped stems taken from each transect were kept separate by site, date of collection, and transect interval. The water depth (from the top of the soil to the water line) and salinity were recorded for each interval at each site. Water salinity at the bottom, midway between the bottom and water surface, and just below the water surface was determined using a portable salinity RF20 refractometer (FLIR Commerical Systems, Inc). To determine the health status of the P. australis stands within the sampling sites, three sections of stem densities were measured within each interval along the transect within 0.25 m² quadrats (as in Knight et al., 2020) and live and dead P. australis plants were counted.

Laboratory analysis

Phragmites specimens were taken back to the laboratory for subsequent analysis. For each sample plant the height and the proportion of the stem that was green were measured. The stem height was measured from the base of the stem to the apical meristem of the shoot. Percentage green stem was defined as an index of the extent of plant senescence and was estimated visually as the percentage of the stem covered by green leaf sheath tissue (as in Knight et al., 2020). The proportion of plants infested with scales was analyzed from 30 stems haphazardly collected from within each interval. Leaf-sheaths were peeled back carefully from the stem and total count of live scales, including adult (>5 mm) and immature (<5 mm) scales, and the scales with evident parasitoid wasp larvae and pupae were counted on each stem. Since it is not yet known which scale instar is the earliest that can be parasitized, for this study all stages of development were considered susceptible. Counts were made within a week of collection. Adult female scales on the collected specimens were segregated into parasitized and unparasitized groups using a binocular microscope (Olympus SZX12). Scales with developing parasitoids could be identified without dissection because parasitoid wasp larvae or pupae were visible inside the scale bodies. Scales that were fully desiccated, preyed upon, or with visible parasitoid exit holes were presumed dead and were not considered for analysis (as in Kaneko, 2005). Live scales were removed gently using soft forceps or a soft bristle paintbrush and placed in a petri dish. Afterward, the scales were put singly in a gel capsule to allow easy visual examination and labeled based on the site and date of collection and transect interval. The samples were placed in the growth chamber and reared at an ambient temperature of 26℃, 75% relative humidity, and 16:8 hour day/night cycle photoperiod. Potentially parasitized scales were monitored every other day to record any parasitoid emergence. For six weeks following collection, the number and sex of parasitoids that emerged from the reared scales were recorded. The wasps were preserved in 80% ethanol upon detection of the parasitoid species. Ethanol-preserved specimens were dehydrated through increasing concentrations of ethanol and transferred to hexamethyldisilazane (HMDS) (Heraty & Hawks, 1998) before card-mounting. The overall parasitism rate of the scales was calculated as the proportion of parasitized scales to the total number of scales.

Reared A. lasallei were identified using Noyes & Higashiura (2020) and with comparison specimens in the Smithsonian Institution National Museum of Natural History. Identifications of N. japonicus (Tachikawa) and B. depressa Hoffer were corroborated using Noyes & Hayat (1984) and Tachikawa (1970). Specimens reared as part of this study are deposited in the National Museum of Natural History, Washington, D.C.

Data analysis

Data presented in Results are shown with arithmetic means and their standard errors. Data on stem heights, percentage of the stems with green tissue, and dead and live stem densities were collated, tabulated, and analyzed with one-way analysis of variance using the SPSS statistical software package version 20.0 (IBM Corp., Armonk, NY). The means were compared using the Least Significant Difference (LSD) test (P<0.05) to assess differences between means.