Biocontrol Science and Technology

ISSN: 0958-3157 (Print) 1360-0478 (Online) Journal homepage: http://www.tandfonline.com/loi/cbst20

Biological control of water hyacinth in California’s Sacramento–San Joaquin River Delta: observations on establishment and spread

R. Patrick Akers, Rebecca W. Bergmann & Michael J. Pitcairn

To cite this article: R. Patrick Akers, Rebecca W. Bergmann & Michael J. Pitcairn (2017) Biological control of water hyacinth in California’s Sacramento–San Joaquin River Delta: observations on establishment and spread, Biocontrol Science and Technology, 27:6, 755-768, DOI: 10.1080/09583157.2017.1342220 To link to this article: http://dx.doi.org/10.1080/09583157.2017.1342220

Published online: 03 Jul 2017.

Submit your article to this journal

Article views: 33

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=cbst20

Download by: [University of Florida] Date: 24 October 2017, At: 09:01 BIOCONTROL SCIENCE AND TECHNOLOGY, 2017 VOL. 27, NO. 6, 755–768 https://doi.org/10.1080/09583157.2017.1342220

RESEARCH ARTICLE Biological control of water hyacinth in California’s Sacramento–San Joaquin River Delta: observations on establishment and spread R. Patrick Akers, Rebecca W. Bergmann and Michael J. Pitcairn California Department of Food and Agriculture, Biological Control Program, Sacramento, CA, USA

ABSTRACT ARTICLE HISTORY Water hyacinth ( (Martius) Solms-Laubach) is a Received 28 December 2016 serious invasive weed in the Sacramento–San Joaquin River Delta Returned 6 June 2017 of California. Three : eichhorniae Warner and Accepted 10 June 2017 Neochetina bruchi Hustache (Coleoptera: Curculionidae) and KEYWORDS Niphograpta (=Sameodes) albiguttalis (Warren) (: – Neochetina bruchi; ) were released during 1982 1987 at four locations for Neochetina eichhorniae; the biological control of water hyacinth. Observations in 1985 Niphograpta albiguttalis; suggested that all three species had established. By 2002, water microsporidia hyacinth populations in the Delta still required an aggressive chemical control campaign and the status of the biological control agents was in question. In late 2002, a field survey to determine the distribution and abundance of the released insects was performed. Water hyacinth plants were collected by boat in the main water channels and from land at smaller sloughs and examined for insects. In total, 27 sites with water hyacinth distributed across the Delta were examined of which 21 had weevils. Weevil abundance ranged from 0 to 10.9 weevils per plant, with an average of 0.93 (±0.47 SEM) adult weevils per plant. All weevils (n = 518) were identified as N. bruchi.NoN. eichhorniae were recovered and no larvae or evidence of larval feeding by N. albiguttalis were observed. A total of 322 weevils were examined for microsporidia and none was found infected, indicating an infection rate of less than 1%. These results suggest that N. bruchi may be the only established biological control agent of water hyacinth in the Delta and that infection by microsporidia does not appear to be limiting its population

Downloaded by [University of Florida] at 09:01 24 October 2017 abundance.

Introduction The Sacramento–San Joaquin River Delta (hereafter, the ‘Delta’) is a large freshwater estuary formed by the confluence of two major rivers that drain the Sacramento and San Joaquin Valleys and surrounding mountain ranges of California. These combined watersheds represent 40% of the land area of California and convey almost half of the annual surface water generated statewide (California Department of Water Resources, 2010). The meeting of these two great rivers has resulted in a complex maze of

CONTACT Michael J. Pitcairn [email protected] © 2017 Informa UK Limited, trading as Taylor & Francis Group 756 R. P. AKERS ET AL.

approximately 1020 km of natural and man-made channels and water bodies (Gibson, Pratt, & Holcomb, 2002), that interlace a low-lying area of 299,000 hectares. Prior to the mid-1800s, it was a brackish estuary and peat marshland, but now consists mostly of fresh water due to impoundments on upstream tributaries and hydrologic engineering projects within the Delta (Ingebritsen & Ikehara, 1999). It serves as a source of drinking water for 25 million Californians and irrigation water for three million acres of agricultural land. The Delta, itself, contains over 218,000 hectares of agricultural land and is home to a half a million residents. It is heavily utilised for agriculture, commerce, and recreation (California Department of Water Resources, 2010). Water hyacinth, Eichhornia crassipes (Martius) Solms-Laubach (Liliales: Pontederia- ceae), is a floating macrophyte with shiny green leaves, inflated spongy petioles, and attractive purple and yellow flowers that has become very abundant on the waterways of the Delta. It is exotic to , having originated in the upper watershed of the Amazon River in South America (Julien, 2001). It was first introduced into the United States in 1884 as an ornamental plant for water gardens (Center, 1994). The first record of water hyacinth growing in the waters of California is a herbarium specimen collected in 1904 in a slough near the town of Clarksburg in Yolo County, a small com- munity in the northern part of the Delta (Bock, 1968). It was first reported as an economic nuisance in 1920 when plants were found in a small tributary of the Kings River in Fresno County, with the observation that the infestation ‘…extends down the slough for about a mile’ (Johnson, 1920). Water hyacinth reproduces clonally through the formation of stolons that give rise to daughter plants and sexually through viable seed. It grows well in the Delta, attaining heights of up to 1 m, and can completely obstruct the passage of boats in small sloughs and forms large mats along the edges of main channels that limit navigation and rec- reational use, reduce water flow, and clog pumps essential for agricultural and domestic water supplies (Anderson, 1990; Mercado, 1983; Thomas & Anderson, 1984). The thick mats shade out light which limits submerged plant growth and alters water quality measures such as dissolved oxygen (Villamagna & Murphy, 2010). A concerted control effort using herbicides addresses the most damaging populations, covering ca 1000– 2000 ha per year, but it is costly, exceeding $6 million USD annually (Edward Hard, Div- ision of Boating & Waterways, California Department of Parks and Recreation, personal communication). Downloaded by [University of Florida] at 09:01 24 October 2017 Biological control through the use of host-specific natural enemies may provide a more cost-effective and sustainable method of controlling water hyacinth in the Delta. The use of biological control successfully reduced the abundance of water hyacinth in several countries worldwide, including Papua New Guinea (Julien & Orapa, 2000), India (Jayanth, 1988), Lake Victoria, (Mailu, 2001), Benin (De Groote, Ajuonu, Attignon, Djessou, & Neuenschwander, 2003), Australia (Wright, 1980), and Mexico (Aguilar, Camarena, Center, & Bojorquez, 2003) and in Louisiana, Florida, and Texas (Center, Cofranceso, & Balciunas, 1989; Cofrancesco, Stewart, & Sanders, 1985; Goyer & Stark, 1984). In California, three species of biological control insects were released in 1982– 1987 for the biological control of water hyacinth in the Delta. The results of this project were reported in a technical report by Stewart, Cofrancesco, and Bezark (1988) and are recapitulated here. BIOCONTROL SCIENCE AND TECHNOLOGY 757

Initial releases of biological control agents: a brief summary During 1982–1985, the Aquatic Plant Control Research Program of the US Army Corps of Engineers Waterways Experiment Station, Vicksburg, Miss., and the Biological Control Program of the California Department of Food and Agriculture (CDFA) in Sacramento, California, entered into a cooperative effort to introduce combinations of two South American weevils, Neochetina eichhorniae Warner and N. bruchi Hustache (Coleoptera: Curculionidae), and a South American , Niphograpta (formerly Sameodes) albiguttalis (Warren) (Lepidoptera: Crambidae), at four locations in the Delta (Stewart et al., 1988). All three had previously been released into Florida, Louisi- ana, and Texas in 1972, 1974, and 1976, respectively. The objectives of the California project were to establish founder colonies in the Delta, perform short-term analyses of their effectiveness, and determine if some combinations of agents were more effective than others. Releases of the three biological control agents in the Delta occurred between 1982 and 1987. The two Neochetina species were collected from a field population near the Trinity River at Wallisville, Texas. Weevils were shipped to the United States Department of Agri- culture’s Agricultural Research Service (USDA-ARS) quarantine facility at Albany, CA, where they were examined for pathogens and parasites, separated to species, then held on living water hyacinth until release in the field, usually within a few weeks. In these col- lections, N. eichhorniae was the more abundant species (Stewart et al., 1988). The moth, N. albiguttalis, came from pupae collected from field sites in Florida and Louisiana and shipped to the USDA-ARS facility. The adult that emerged from these pupae were held as male–female pairs in quarantine. Once the female laid eggs, the parents were checked for internal pathogens. Offspring of pathogen-free parents were used to initiate a moth colony maintained by the CDFA Biological Control Program in a green- house in Sacramento. Only N. albiguttalis from the CDFA colony were released at field sites in the Delta (Stewart et al., 1988). There were four release sites in the Delta: White Slough, Trapper Slough, Veale Tract, and Old River (Figure 1). The sites were chosen for the stability of their water hyacinth populations, to allow for variances in environmental conditions, and to assist in the dis- persion of the agents throughout the area. Water hyacinth plants growing at Old River, White Slough, and Veale Tract were confined by a floating beam to prevent the

Downloaded by [University of Florida] at 09:01 24 October 2017 breakup of the large mat of plants. The amount of area occupied with water hyacinth at each release site was: Old River, 30.5 m × 122 m (3721 m2); White Slough, 18 m × 220 m (3960 m2); 30.5 m × 975 m (29,737 m2); Veale Tract, 36.5 m × 146 m (5329 m2). The numbers of insects and the combination of species released varied by site (Table 1). In total, 7500 adult N. eichhorniae, 2823 adult N. bruchi, and 12,385 N. albiguttalis larvae were released from July 1982 through August 1984 (Stewart et al., 1988). Later in 1987, the CDFA cleared out its colony of N. albiguttalis and made a one-time release of 2300 larvae at Trapper Slough (Bezark, Villegas, Ball, & Casanave, 1988). All four sites were mon- itored every four months for establishment and population growth of the insects and several parameters of plant growth in an attempt to measure impact. When monitoring was discon- tinued in August 1985, all three species were reported to have established in at least one release site (Table 1): N. bruchi at Veale Tract and Old River and N. eichhorniae and N. albiguttalis at White Slough. No agent showed signs of establishment at Trapper 758 R. P. AKERS ET AL. Downloaded by [University of Florida] at 09:01 24 October 2017

Figure 1. Release locations (circles) for three biological control agents on water hyacinth in the Sacra- mento–San Joaquin River Delta from 1982 to 1987. The square outline identifies the area of the central Delta where water hyacinth was surveyed for biological control agents in 2002. Triangles identify locations of other sloughs visited during the land-based surveys.

Slough. Dispersion of the three species away from the release sites was monitored but the results showed little or no movement away from release sites by August 1985 when the study ended (Stewart et al., 1988). BIOCONTROL SCIENCE AND TECHNOLOGY 759

Table 1. Number of adult biological control insects released at four sites in the Sacramento–San Joaquin River Delta from 1982 to 1984 (from Stewart et al., 1988). White Old River Slough Trapper Slough Veale Tract Species Totals N. bruchi Number released 1414 0 850 559 2853 Status in August Established – Not established Established 1985 N eichhorniae Number released 0 3969 200 3331 7500 Status in August 1985 – Established Not established Not established N albiguttalis Number released 3530 3500 1855+2300 (1987) 3500 14,685 Status in August 1985 Not established Established Not established Not established Note: The status of the introduction as of August 1985 is indicated. An additional 2300 N. albiguttalis were released by the California Department of Food and Agriculture at Trapper Slough in 1987.

Post-release observations and objectives In 1985 when the project had ended, the opinion was that all three insects had established (Stewart et al., 1988). After 1985, efforts to document the status of the released biological control agents were discontinued. Later observations found weevils on water hyacinth plants in the Delta but the species were not identified and no quantitative information on abundance or spatial spread was obtained. Spencer and Ksander (2004), citing Stewart et al. (1988), reported that both N. eichhorniae and N. bruchi were present in the Delta but there was no mention of efforts for species-specific identification. Here we report on a survey in 2002 of water hyacinth plants growing in the central Delta to determine the status of the three biological control organisms. The objectives of the survey were: (1) determine those species that successfully established; (2) examine the amount spread away from the original release sites; and (3) determine if the popu- lations of the biological control agents were infected with microsporidia, an internal pathogen that may limit their effectiveness. Accurate identification of those biological control agents that continued to persist in the Delta is necessary to determine if the release efforts were successful. For the two Neochetina weevils, differences in their diag- nostic characters are subtle and accurate identification is best confirmed by an expert taxonomist. If several species established, an assessment of their relative abundance would allow us to determine if one species was under-represented for which we may consider additional releases. Movement away from release sites would help us determine if there is a need to further re-distribute the biological control agents throughout the Downloaded by [University of Florida] at 09:01 24 October 2017 Delta. At the time of the survey, the common perception was that at least one species of Neo- chetina established but populations were too low in abundance to produce significant damage to the water hyacinth plants growing in the Delta (Anderson, 1990). One reason for this suggested that the weevils were infested with microsporidia, internal patho- gens of insects that can cause reduction in longevity and fecundity in infected individuals (Joudrey & Bjornson, 2007; Siegel, Maddox, & Ruesink, 1986). Studies of the Neochetina weevils in Florida showed that infection rates average about 9% for N. eichhorniae and 4% for N. bruchi (Center & Rebelo, 2001). Laboratory studies on the impact of Nosema infec- tion on the two Neochetina spp. showed that adult survival rates were reduced by 30% in both species and fecundity was reduced by 72% in N. eichhorniae and 62% in N. bruchi (Goettel & Inglis, 2006). At the time of the original releases in the Delta, weevil shipments were screened for microsporidia and Nosema sp. was noted in at least some shipments, but 760 R. P. AKERS ET AL.

the overall rate of infection was very low and only disease-free insects were released at field sites (Stewart et al., 1988). Conditions in the Delta are dynamic and the presence and abundance of water hyacinth and its biological control insects will vary year to year. However, documenting the results for objectives one and two (determining successful establishment and spread from release sites) benefits greatly from knowledge of the date of the observations. Thus, we hope the results of this survey serve as a baseline from which to compare results from future surveys.

Methods Efforts to examine the status of the biological control agents released on water hyacinth in the Delta occurred during September through December 2002. Field observations showed that the weevils were most abundant in late summer and early fall (unpublished data) and surveys during this time period provided the best chance of finding biological control agents. At this time of year, water hyacinth in the Delta achieves its peak annual biomass (Spencer & Ksander, 2005). Survey efforts consisted of two kinds of activities: a survey of plants in the main water channels using watercraft and a survey of plants in small tributaries and sloughs from shore.

Water-based survey of the main water channels The survey of water hyacinth plants growing in the water channels of the central Delta occurred from 30 September through 3 October 2002. A sample was obtained by driving an airboat or a conventional boat alongside a floating mat of plants and individual plants were lifted into the boat by hand or with a hook on a pole. Because plants repro- duced clonally, an individual plant was identified as those not connected by a stolon to another plant. All plants brought into the boat were examined for the released biological control insects and feeding damage. Efforts were directed at finding adult weevils for species identification and examination of microsporidia. Adult weevils are inactive during the day and hide in the centre of the plant among the furl of new leaves (Julien, 2001). To obtain adults, plants were pulled apart and examined. We did not look for weevil larvae or pupae. We also examined the plants for N. albiguttalis larvae or feeding Downloaded by [University of Florida] at 09:01 24 October 2017 damage (tunnels, chewed gaps, adult emergence holes, and larval frass on leaf petioles). All suspect insects were returned to the laboratory for identification to species, counting, and examination for microsporidia. In addition, the sample location (lat/long coordi- nates), patch size, and the number of plants examined was recorded. Within the main water channels, we sampled 23 locations with water hyacinth. The area examined covered most of the navigable waters in an area of 600 km2 (230 square miles). Sample locations were chosen to provide a separation of several kilometres between sites to provide better spatial representation of the area. Due to water flow and the state-operated herbicide program to control water hyacinth, we would often travel several kilometres along some channels without seeing water hyacinth, and at many other locations, there were often only small scattered mats of plants, 1–2 m in diameter, available for sampling. For small patches, all plants were examined; for large patches, a sample of 30–50 plants were examined. BIOCONTROL SCIENCE AND TECHNOLOGY 761

Land-based survey of the smaller canals and sloughs Original release sites Efforts were made to visit each of the four original release sites (Figure 1) to see if any of the released agents were present. Plants near the White Slough site were examined as part of the main channel survey. The release sites at Veale Tract, Old River, and Trapper Slough were visited from land. At Veale Tract and Trapper Slough, water hyacinth plants were pulled to shore using a hook on a pole and plants were examined as above for the water-based survey. All insects were returned to the laboratory for identification and examination for microsporidia. At the Old River release site, plants were being sprayed with herbicide at the time of the survey visit, so no plants were examined and no insects were obtained.

Other small sloughs Sevenmile Slough and Walthall Slough (Figure 1) had a high abundance of water hyacinth during fall 2002 and were examined for evidence of biological control insects in the first half of December 2002. These sloughs were upstream from the central area of the Delta and represented areas further distant from the original release sites. The sloughs were checked every 0.5 km along their lengths, where access to the shoreline was possible. At Sevenmile Slough, 11 locations along a 5 km section infested with water hyacinth were examined. At Walthall Slough, plants at 10 locations were examined. At each sample location, plants were pulled to shore with hooks and examined for insects and feeding damage. All insects were returned to the laboratory for identification and examination for microsporidia.

Species identification of adult Neochetina Weevils Initial identification of the Neochetina weevils was based on scale pattern on the elytra. Adult N. bruchi often display a distinctive chevron pattern of scales on the wing covers which is absent with N. eichhorniae. Weevils were tentatively identified as N. eichhorniae if they did not have the distinctive chevron pattern. These putative N. eichhorniae were further compared to keys and descriptions and to voucher specimens of the two species to confirm their identity. Accurate identification depends upon key

Downloaded by [University of Florida] at 09:01 24 October 2017 characters on the antennal segments, rostrum, coxae, and reproductive organs and exam- ination of these characters indicated that most of the weevils without the chevron scale patterns appeared to be N. bruchi. For confirmation of these results, representative material as well as all specimens of uncertain identification were examined by Dr Charles O’Brien, a specialist on the Family Curculionidae who had produced the most recently published key on the Neochetina (1976).

Bulk collections and microsporidia screening of adult Neochetina Weevils Adult weevils were examined for microsporidia and other pathogens by Dr Gerard M. Thomas of Consulting Diagnostic Service, a private insect pathologist located in Ber- keley, California. The submitted specimens were examined by microscopic study of tritu- rated tissue smears under phase microscopy. Special attention was given to the detection of 762 R. P. AKERS ET AL.

microsporidia. The sample of adult weevils examined consisted of weevils from survey sites where 10 or more weevils were collected (n = 13 sites), except for the one high- density site where 56 (out of 109) weevils were submitted. To further increase sample size, a large collection of adult weevils was obtained September 24–27 in an effort separate from the water-based and land-based surveys above. These adult weevils were collected from four large rafts of plants within water-based survey area and an additional 60 weevils were included with the specimens submitted for pathogen screening.

Results Survey of the main water channels Water hyacinth plants growing in the main water channels were examined at 23 locations within the central area of the Delta (Figure 2). A total of 354 adult weevils were found at 18 of the 23 sites. Abundance ranged from 0 to 10.9 weevils per plant, averaging 0.93 (±0.47 SEM) weevils per plant. All weevils were identified as N. bruchi;noN. eichhorniae were found. No plants with damage consistent with feeding by N. albiguttalis larvae were observed suggesting that the moth failed to establish at these sites. Local abundance of weevils was not associated with patch size (linear regression, weevil abundance (weevils per plant) against patch size (m2), r2 = 0.01, p = .67) (Figure 3). Downloaded by [University of Florida] at 09:01 24 October 2017

Figure 2. Survey results for the occurrence of biological control agents on water hyacinth in the central area of the Sacramento–San Joaquin River Delta, September–December, 2002. Abundance of N. bruchi is reported as number of weevils per plant. BIOCONTROL SCIENCE AND TECHNOLOGY 763

Survey of smaller channels and sloughs Original release sites Adult weevils were found at the three original release sites for which a sample was obtained: White Slough (n = 17 adult weevils), Trapper Slough (n = 8 adults), and Veale Tract (n = 12 adults). All weevils were identified as N. bruchi. There was no evidence of larvae or larval feeding by N. albiguttalis. No sample was obtained from the Old River site due to chemical herbicides being applied to plants at the time of the visit. At White Slough, only N. eichhorniae were released and at cessation of monitoring efforts in August 1985, this weevil was thought to have established. In 2002, only N. bruchi was col- lected there.

Other small sloughs No weevils were found at 9 of 10 locations sampled along Walthall Slough. One adult weevil was found at one location. There was no evidence of moth larvae or larval feeding damage at any sample location. At Sevenmile Slough, we sampled 11 locations along the 5 km section infested with plants. At all locations the plants had no signs of weevil or moth feeding damage and we found no insects (Figure 2).

Species identification of adult weevils and spread from release sites A total of 518 adult Neochetina weevils were collected during this survey. All were ident- ified as N. bruchi. Representative specimens were submitted to Dr Charles O’Brien, a Neo- chetina specialist, to confirm identification. Among these were five specimens we tentatively identified as N. eichhorniae, however, all were identified as N. bruchi by Dr O’Brien. Thus, it appears that, of the three biological control agents released into the Delta, only N. bruchi established and is now widespread in the central area of the Delta Downloaded by [University of Florida] at 09:01 24 October 2017

Figure 3. The relationship between abundance (weevils per plant) of adult N. bruchi and patch size (m2) of water hyacinth plants. 764 R. P. AKERS ET AL.

(Figure 2), occurring at 21 of 27 sampling locations (including Walthall Slough) (78% of locations). Spread of the weevil away from release sites may occur passively through movement of infested plants downstream or actively through adult flight. The furthest location with weevils downstream from an original release site was 15.8 km from the Veale Tract, one of the two locations N. bruchi was reported to have established in 1985. Interestingly, White Slough is the location furthest upstream from all three locations where N. bruchi was released and its occurrence there suggests active movement to this location. White Slough is 23.3 km from Veale Tract and 18.2 km from Trapper Slough where N. bruchi was released but was not reported to have established. Given that these observations are 20 years after initial release, it is difficult to estimate a rate of spread. However, the spatial dispersal of locations where N. bruchi was collected do show this weevil to occur throughout an area of 704 km2, with locations both upstream and downstream from release sites, indicating active and passive movement, respectively.

Microsporidia screening of adult weevils A total of 322 weevils were submitted for microsporidia screening: none were found infected. The pathologist noted that he saw no signs of any other pathogens, including some common commensals. It is important to note that the absence of finding an infected weevil does not completely rule out the presence of microsporidia in the weevil population in the Delta. If we made an assumption that the true infection rate in the Delta was 1% and that the chance of infection among weevils is independent, there would be a probability of 0.039 of collecting 322 weevils that had no microsporidia (p = .99322), and thereby falsely concluding there was no disease. In other words, if microsporidia were present, the infec- tion rate was estimated to be less than a 1%.

Discussion The biological control water hyacinth in the California Delta began with the release of three agents, N. eichhorniae, N. bruchi, and N. albiguttalis, in 1982. When field monitoring ended in 1985, it was reported that all three species had established at, at least, one release location (Stewart et al., 1988). Twenty years later, the only agent collected was N. bruchi. Downloaded by [University of Florida] at 09:01 24 October 2017 This is surprising as N. eichhorniae is often the more abundant species elsewhere in the southeastern United States. For example, Center and Dray (1992) reported that at 22 locations throughout southern Florida, N. eichhorniae made up 70% of the weevils col- lected. More recently, Tipping et al. (2014) reported that at four sample locations in Florida over 99% of weevil adults were N. eichhorniae. It should be acknowledged that it is possible that N. eichhorniae and N. albiguttalis do occur in the Delta but their abun- dance is very low and may have been missed. Alternatively, they may occur outside the survey area such as in the southern portion of the Delta where water hyacinth occurs in tributaries of the San Joaquin and Merced Rivers. If the results of the 2002 survey are accurate, that only N. bruchi established, we may consider why N. eichhorniae and N. albiguttalis did not. The reasons for failure of estab- lishment have been grouped into four categories: (1) mortality due to predation, parasit- ism, or disease; (2) competition; (3) host incompatibility; and (4) climate (Crawley, 1989). BIOCONTROL SCIENCE AND TECHNOLOGY 765

For N. eichhorniae, predation, parasitism, and disease may be an unlikely explanation because what may kill N. eichhorniae will likely kill N. bruchi. Competition is also unlikely as an explanation for establishment failure as the two weevil populations did not reach population levels where water hyacinth became a limiting resource. Host incompatibility may be a factor. Heard and Winterton (2000) reared N. eichhorniae and N. bruchi on water hyacinth with three levels of nitrogen in their leaf tissue and found that feeding by both species reduced growth of water hyacinth but that N. bruchi had a significantly greater impact on growth than N. eichhorniae. This was attributed to N. bruchi obtaining relatively more benefit than N. eichhorniae from the higher nitrogen levels in the leaf tissue. Similarly, Center and Dray (2010) showed that N. bruchi performed better (higher egg production and population growth rate) than N. eichhorniae when reared on plants exposed to variable fertiliser regimens. For water hyacinth plants in the Delta, Spencer and Ksander (2004) found that, for the majority of leaf lamina samples (67%), tissue nitrogen levels exceeded 2.93%, the medium level for Heard and Winterton (2000) and concluded that tissue nitrogen was more than sufficient for weevil growth. It may be that the high nitrogen levels in the water hyacinth in the Delta may have pro- vided a benefit to N. bruchi over N. eichhorniae. Crawley (1989) identified climate as the most common explanation in establishment failure for released biological control agents. The climate of the Delta is Mediterranean and experiences 1–2 months of light frost (temperatures below 0°C are infrequent and limited to a few hours at night). The three biological agents are adapted to warm climates as found in tropical and subtropical climates. There is no information on the thermal requirements for N. bruchi and N. eichhorniae (lower developmental threshold or degree-days required for development from egg to adult). Deloach and Cordo (1976) provided data on oviposition and egg development at several constant temperatures for both weevils. Their results showed that adult N. bruchi were able to deposit eggs at 10°C and 15°C and those produced at 15°C successfully developed. Adult N. eichhorniae deposited no eggs at 10°C and only 1 egg at 15°C and it was not viable. These data suggest that N. bruchi has a higher tolerance for cold temperatures than N. eichhorniae which may allow it to better tolerate the abiotic conditions of the Delta. Among the three agents released, N. bruchi appeared to be performing best at the end of the initial release effort in the Delta (Stewart et al., 1988). The failure of N. albiguttalis to persist may be due to predation, host incompatibility, Downloaded by [University of Florida] at 09:01 24 October 2017 and climate. For part of their life cycle, immature N. albiguttalis feed externally and are vulnerable to being fed on by birds and generalist invertebrate predators (Julien, 2001). Interestingly, the thermal requirements for N. albiguttalis suggest that this insect has some tolerance for cold temperatures and, in South Africa, N. albiguttalis successfully established at several locations in the Highveld region which has a temperate climate and several months of cold temperatures (May & Coetzee, 2013). Lastly, it is reported that N. albiguttalis prefers the bulbous form of water hyacinth (Center & Durden, 1981; Julien, 2001), which is uncommon in the Delta and may have further contributed to estab- lishment failure. The combined results of the land- and water-based surveys in the central area of the Delta showed that N. bruchi was present and common throughout much of the central Delta. It was unusually abundant at one location (over 10 weevils per plant), but the overall average was about one weevil per plant. Patch size clearly did not affect the 766 R. P. AKERS ET AL.

density of the weevils. Patches of plants with weevils can get broken up by winds and water currents, carried downstream and recombined with plants from other areas of the Delta and this can lead to passive dispersal of weevils downstream from release sites. Conversely, N. bruchi appears to have actively spread upstream from the three release sites where it was originally introduced. The spatial spread of locations where N. bruchi was collected shows this weevil to occur throughout an area of 704 km2 of the Delta. One of the primary goals of this survey effort was to collect a large sample of adult weevils (at least 300 adults) to screen for microsporidia. This sample size was determined by considering the likely infection rates and the probability of falsely concluding that there was no microsporidian disease. In this survey, we submitted 322 weevils for examination for infection by microsporidia. All were negative, suggesting that, if present, the infection rate was less than 1%. In summary, N. bruchi is firmly established in the Delta, while it appears that N. eichhorniae and N. albiguttalis either did not establish or occur at such low levels that they were not detected. The infection rate of deleterious microsporidia appears to be extremely low and does not appear to be limiting the reproduction and survivorship of the weevil population. It is recommended that future research be directed at the follow- ing: (1) obtain description of the thermal requirements for N. bruchi and N. eichhorniae; (2) expand agent survey in California to areas further south where temperatures are warmer than in the central Delta; (3) consider re-releasing N. eichhorniae in areas with better climatic match, such as in southern California, and N. albiguttalis in areas with bulbous plants; (4) release new agents better adapted to the climate of the California Delta, either new biotypes of N. bruchi and N. eichhorniae or new agent species.

Acknowledgements The authors thank Patrick Moran and Paul Pratt, USDA Agricultural Research Service, Albany, CA, and two anonymous reviewers for comments on an early draft of this manuscript. The authors also thank Robin Breckenridge of the California Department of Food and Agriculture’s Noxious Weed Program who piloted the boat and assisted during the survey efforts and Sean Farnum for creating the maps in Figures 1 and 2.

Downloaded by [University of Florida] at 09:01 24 October 2017 Disclosure statement No potential conflict of interest was reported by the authors.

Funding This work was supported by the California Department of Boating and Waterways [agreement number 02-105-061].

References Aguilar, J. A., Camarena, O. M., Center, T. D., & Bojorquez, G. (2003). Biological control of water- hyacinth in Sinaloa, Mexico, with the weevils Neochetina eichhorniae and N. Bruchi. BioControl, 48, 595–608. BIOCONTROL SCIENCE AND TECHNOLOGY 767

Anderson, L. W. J. (1990). Aquatic weed problems and management in North America. In A. H. Pieterse & K. J. Murphy (Eds.), Aquatic weeds: The ecology and management of nuisance aquatic vegetation (pp. 371–391). Oxford: Oxford University Press. Bezark, L., Villegas, B., Ball, J., & Casanave, K. (1988). 1987 Progress report for natural enemy dis- tribution projects (Technical Report BC-88-2). California Department of Food and Agriculture, 28 p. Bock, J. H. (1968). The water hyacinth in California. Madroño, 19, 281–283. California Department of Water Resources. (2010). Sacramento-San Joaquin delta overview. Retrieved from http://baydeltaoffice.water.ca.gov/sdb/tbp/deltaoverview/index.cfm Center, T. D. (1994). Biological control of weeds: waterhyacinth and waterlettuce. In D. Rosen, F. D. Bennett, & J. L. Capinera (Eds.), Pest management in the subtropics - a Florida perspective (pp. 481–521). Wimborne: Intercept Ltd. Center, T. D., Cofranceso, A. F., & Balciunas, J. K. (1989). Biological control of aquatic and wetland weeds in the southeastern United States. In E. S. Delfosse (Ed.), Proceeding of the VIIth International Symposium on the Biological Control of Weeds (pp. 239–262). East Melbourne: CSIRO Publications. Center, T. D., & Dray, A. F. (1992). Associations between waterhyacinth weevils (Neochetina ech- horniae and N. bruchi) and phenological stages of Eichhornia crassipes in southern Florida. The Florida Entomologist, 75, 196–211. Center, T. D., & Dray, A. F. (2010). Bottom-up control of water hyacinth weevil populations: Do the plants regulate the insects? Journal of Applied Ecology, 47, 329–337. Center, T. D., & Durden, W. C. (1981). Release and establishment of Sameodes albiguttalis for the biological control of water hyacinth. Environmental Entomology, 10,75–80. Center, T. D., & Rebelo, T. (2001). Microsporidia and Neochetina. Water Hyacinth News, 4,5–7. Cofrancesco, A. F., Stewart, R. M., & Sanders, D. R. ( 1985). The impact of Neochetina eichhorniae (Coleoptera: Curculionidae) on waterhyacinth in Louisiana. In E. S. Delfosse (Ed.), Proceedings of the VIth International Symposium on the Biological Control of Weeds (pp. 525–535). Ottawa: Agriculture Canada. Crawley, M. J. (1989). The successes and failures of weed biocontrol using insects. Biocontrol News and Information, 10, 213–223. De Groote, H., Ajuonu, O., Attignon, S., Djessou, R., & Neuenschwander, P. (2003). Economic impact of biological control of water hyacinth in southern Benin. Ecological Economics, 45, 105–117. Deloach, C. J., & Cordo, H. A. (1976). Life cycle and biology of Neochetina bruchi, a weevil attacking waterhyacinth in Argentina, with notes on N. Eichhorniae. Annals of the Entomological Society of America, 69, 643–652. Gibson, J. A., Pratt, W. B., & Holcomb, J. C. (2002). Sacramento – San Joaquin delta boating needs assessment 2000-2020. Report to the California Department of Boating and Waterways. Retrieved from http://dbw.ca.gov/deltaindex.asp Downloaded by [University of Florida] at 09:01 24 October 2017 Goettel, M. S., & Inglis, G. D. (2006). Methods for assessment of contaminants of invertebrate bio- logical control agents and associated risks. In F. Bigler, D. Babendreier, & U. Kuhlmann (Eds.), Environmental impact of invertebrates for biological control of (pp. 145–165). Wallingford: CABI International. Goyer, R. A., & Stark, J. D. (1984). The impact of Neochetina eichhorniae on waterhyacinth in southern Louisiana. Journal of Aquatic Plant Management, 22,57–61. Heard, T. A., & Winterton, S. L. (2000). Interactions between nutrient status and weevil herbivory in the biological control of water hyacinth. Journal of Applied Ecology, 37, 117–127. Ingebritsen, S. E., & Ikehara M. E. (1999). Sacramento-San Joaquin Delta: The sinking heart of the state. In D. Galloway, D. R. Jones, & S. E. Ingebritsen (Eds.), Land subsidence in the United States (pp. 83–94). Circular 1182. Reston, VA: U.S. Geological Survey. Jayanth, K. P. (1988). Successful biological control of water hyacinth (Eichhornia crassipes)by Neochetina eichhorniae in Bangalore, India. Tropical Pest Management, 34, 263–266. Johnson, E. 1920. Fresno county will fight water hyacinth. Monthly Bull. Calif. State Dept. Agr. 9, 202–203. 768 R. P. AKERS ET AL.

Joudrey, P., & Bjornson, S. (2007). Effects of an unidentified microsporidium on the convergent lady , Hippodamia convergens Guérin-Méneville (Coleoptera: Coccinellidae), used for bio- logical control. Journal of Invertebrate Pathology, 94, 140–143. Julien, M. H. (2001). Biological control of water hyacinth with arthropods: A review to 2000. In M. H. Julien, M. P. Hill, T. D. Center, & D. Jianqing (Eds.), Biological control of water hyacinth, Eichhornia crassipes (pp. 8–20). ACIAR Proceedings No. 102, Canberra: Australian Centre for International Agricultural Research. Julien, M. H., & Orapa, W. (2000). Successful biological control of water hyacinth (Eichhornia cras- sipes in Papua New Guinea by the weevils Neochetina bruchi and Neochetina eichhorniae (Coleoptera: Curculionidae). In N. R. Spencer (Ed.), Proceedings of the Xth International Symposium on Biological Control of Weeds (pp. 138–139). Bozeman: Montana State University Press. Mailu, A. M. (2001). Preliminary assessment of the social, economic, and environmental impacts of water hyacinth in the Lake Victoria basin and the status of control. In M. H. Julien, M. P. Hill, T. D. Center, & D. Jianqing (Eds.), Biological control of water hyacinth, Eichhornia crassipes (pp. 130–139). ACIAR Proceedings No. 102, Canberra: Australian Centre for International Agricultural Research. May, B., & Coetzee, J. (2013). Comparisons of the thermal physiology of water hyacinth biological control agents: Predicting establishment and distribution pre- and post-release. Entomologia Experimentalis et Applicata, 147, 241–250. Mercado, M. (1983). The role of boating and waterways in aquatic plant management in the delta. Proceedings, 17th Annual Meeting, Aquatic Plant Control Research Program, US Army Corps of Engineers, Aquatic Plant Control Research Program, Misc. Paper A-83-A, pp. 2–4. O’Brien, C. W. (1976). A taxonomic revision of the New World subaquatic genus Neochetina (Coleoptera: Curculionidae: Bagoini). Annals of the Entomological Society of America, 69, 165–174. Siegel, J. P., Maddox, J. V., & Ruesink, W. G. (1986). Lethal and sub-lethal effects of Nosema pyrausta on the European corn borer (Ostrinia nubilalis) in Central Illinois. Journal of Invertebrate Pathology, 48, 167–173. Spencer, D. F., & Ksander, G. G. (2004). Do tissue carbon and nitrogen limit population growth of weevils introduced to control waterhyacinth at a site in the Sacramento-San Joaquin delta, California? Journal of Aquatic Plant Management, 42,45–48. Spencer, D. F., & Ksander, G. G. (2005). Seasonal growth of water hyacinth in the Sacramento/San Joaquin Delta, California. Journal of Aquatic Plant Management, 43,91–94. Stewart, R. M., Cofrancesco, A. F., & Bezark, L. G. (1988). Biological control of waterhyacinth in the California Delta: Final report (Technical Report A-88-7). Vicksburg, MS: US Army Engineer Waterways Experiment Station, 46 p. Thomas, L., & Anderson, L. (1984). Water hyacinth control in California. Aquatics, 6,11–15. Tipping, P. W., Martin, M. R., Pokorny, E. N., Nimmo, K. R., Fitzgerald, D. L., Dray Jr., F. A., & Downloaded by [University of Florida] at 09:01 24 October 2017 Center, T. D. (2014). Current levels of suppression of waterhyacinth in Florida USA by classical biological control agents. Biological Control, 71,65–69. Villamagna, A. M., & Murphy, B. R. (2010). Ecological and socio-economic impacts of invasive water hyacinth (Eichhornia crassipes): a review. Freshwater Biology, 55, 282–298. Wright, A. D. (1980). Biological control of water hyacinth in Australia. In E. S. Delfosse (Ed.), Proceedings of the Vth International Symposium for the Biological Control of Weeds (pp. 529–535). Melbourne: CSIRO.