Chemically Mediated Host-Plant Selection by the Milfoil Weevil: A Freshwater Insect–Plant Interaction
- 201 Downloads
The milfoil weevil Euhrychiopsis lecontei is a specialist aquatic herbivore that feeds, oviposits, and mates on the invasive freshwater macrophyte Myriophyllum spicatum. We characterized the weevil's preference for M. spicatum, and through bioassay-driven fractionation, isolated and identified two chemicals released by M. spicatum that attract E. lecontei. Mass spectrometry and nuclear magnetic resonance spectroscopy were used to identify the attractive compounds as glycerol and uracil. Dose-response curves for glycerol and uracil indicated that weevil preference increased as sample concentration increased. Weevils were attracted to a crude sample of M. spicatum-released chemicals from 0.17 to 17 mg/l, to glycerol from 18 to 1800 μM (0.0017–0.17 mg/l), and to uracil from 0.015 to 15 μM (0.00014–1.4 mg/l). Although glycerol and uracil are ubiquitous, weevils are likely responding to high concentrations that are released as a result of the rapid growth of M. spicatum. Uracil concentration was greater in the exudates of M.spicatum than other Myriophyllum spp. E. lecontei was attracted to glycerol at a concentration similar to that at which terrestrial insects are attracted to sugar alcohols. This is the first example of a freshwater specialist insect being attracted to chemicals released by its host plant. Analysis of the water milfoil–weevil interaction provides further understanding as to how insects locate their host plants in aquatic systems.
Key WordsFreshwater macrophyte host location Myriophyllum spicatum Euhrychiopsis lecontei specialist herbivore chemical attractant aquatic plant–herbivore interactions
Plants produce and release chemicals that can be used by herbivores to locate mates, food sources, and oviposition sites. There is a great variety of compoundsthat attract terrestrial insects, including ubiquitous chemicals such as ethanol, hexanal, and sugars, as well as species-specific chemicals, such as polyphenols and isothiocyanates (Metcalf and Metcalf, 1992; Bernays and Chapman, 1994). Some ubiquitous chemicals, such as sugars, are nutrients that act as phagostimulants (Chapman, 2003). Common volatiles, such as ethanol and six-carbon alcohols, can be used as long-range attractants (Bruce et al., 2005). However, in aquatic systems, natural products are transported through water rather than air. Therefore, their mobility and persistence in the environment can be different. Simple sugars and amino acids can be rapidly degraded by microorganisms. Nonvolatile chemicals located on leaf surfaces, which typically have low mobility in the air, may become mobile in water. The dynamics of plant–insect interactions will undoubtedly be similar in both aquatic and terrestrial systems, but the mechanism of the interactions and the chemicals involved inthose interactions may be different (for a review, see Hay and Steinberg, 1992).
In aquatic systems, studies of animal response to chemical stimuli have focused primarily on fish and noninsect invertebrates. Feeding stimulants are often common metabolites of low-molecular weight. For example, amino acids, such as glycine, attract lobster, shrimp, and several species of fish (Carr, 1988). Amino acids, nucleotides, simple sugars, and organic acids have been used to assess the attraction of marine invertebrates (primarily crustaceans) to their prey (see reviewby Weissburg et al., 2002).
Recent studies on the functional ecology of natural products in freshwater macrophytes have demonstrated the role they play in defense against generalist herbivores (Kubanek et al., 2001; Choi et al., 2002; Cronin et al., 2002; Walenciak et al., 2002). Furthermore, several weevil species have a specialist or close association with freshwater macrophytes (Buckingham and Bennett, 1995; Cronin et al., 1998; Solarz and Newman, 2001; Center et al., 2002). However, the role of chemistry in these specialist–macrophyte interactions is unknown.
The sole example of a specialist insect being attracted to a submersed freshwater macrophyte is the Myriophyllum spicatum L. (Haloragaceae)–Euhrychiopsis lecontei Dietz (Coleoptera: Curculionidae) system (Solarz and Newman, 1996, 2001). Eurasian water milfoil M. spicatum is an aquatic macrophyte exotic to North America that is typically abundant in the littoral zone of mesotrophic to moderately eutrophic lakes (Smith and Barko, 1990). It may remain unnoticed for years, then suddenly grow, forming a dense canopy that shades out other plants, impedes navigation, and adversely affects recreational activities (Smith and Barko, 1990; Boylen et al., 1999). Reductions in M. spicatum populations have occurred in lakes where E. lecontei was presentand in controlled experiments with E. lecontei (Creed and Sheldon, 1995; Sheldon and Creed, 1995; Newman, 2004).
Water milfoils native to North America, such as the northern milfoil Myriophyllum sibiricum Komarov, are the presumed original hosts of E. lecontei, which has expanded its host range to include M. spicatum (Creed and Sheldon, 1994; Solarz and Newman, 1996). E. lecontei has shorter development times, greater mass, and higher survival on the invasive Eurasian water milfoil than on its native host plants (Newman et al., 1997; Solarz and Newman, 2001). Based on field observations, oviposition studies, and behavioral bioassays, Solarz and Newman (1996) hypothesized that there may be chemical cues that attract weevils to M. spicatum. In this study, we tested their hypothesis that the weevils are attracted to M. spicatum, and we isolated two chemicals involved in that interaction.
Methods and Materials
Plant and Insect Material
Plants were obtained from various lakes in Minnesota, USA. M. spicatum was collected from Lake Auburn (Carver County), Cedar Lake, Lake of the Isles, and Smith's Bay in Lake Minnetonka (all in Hennepin County), M. sibiricum from Christmas Lake (Hennepin County) and Smith's Bay, M. spicatum × M. sibiricum hybrid from Otter Lake (Anoka County) and White Bear Lake (Washington County), Myriophyllum alterniflorum DC from Lake Sissabagamah (Aitkin County), Myriophyllum tenellum Bigel from Graham Lake (Carlton County) and West Twin Lake (St. Louis County), and Ceratophyllum demersum L. (Ceratophyllaceae) from Lake Auburn. The identities of M. spicatum × M. sibiricum hybrid, M. alterniflorum, and M. tenellum were confirmed by sequence comparison of nuclear and chloroplast markers (analysis performed by M. Moody, University of Connecticut; see Moody and Les, 2002 for methodology).
Preparation of Exudates
All plants were collected from stands with no apparent damage from the aquatic weevil E. lecontei. Stems were held in lake water until laboratory processing, which occurred within 24 hr of collection. The top 20 cm of the plants was inspected for herbivore damage and cleaned ofalgae and detritus. Plants were then placed in well water (characterized byOseid and Smith, 1974) to avoid contamination with chlorine and other additives in tap water. Plants were incubated for 2–4 d in direct sunlight or under constant illumination in a laboratory light bank that delivered 1065 μmolm−2 sec−1 photosynthetically active radiation. Temperature was at or below 30°C. After incubation, plants were removed and spun in a salad spinner for 15 sec to remove excess water. Plant material was weighed and frozen for future use. The water containing the milfoil exudates was filtered through cheesecloth, stored in plastic bags (Ziploc™), and frozen at −20°C until processing. Adult E. lecontei were collected biweekly from several of the same area lakes and held in a 0.545-m3 stock tanks (61 × 183 × 76 cm) located in an outdoor setting or in 76-l aquaria on fresh M. spicatum.
Chemicals and Other Materials
Ultrafiltration membranes used to concentrate extracts were obtained from Millipore, Bedford, MA, USA, and the ultrafiltration cell was obtained from Diaflo Amicon, W. R. Grace & Co., Beverly, MA, USA. Diethylaminoethyl (DEAE) ion exchange and Sephadex G-10 resins and the Superdex Peptide HR10/30 column used in the isolation of attractants were purchased from Amersham Biosciences, Piscataway, NJ, USA. Several additional high-performance liquid chromatography (HPLC) columns, C18, C8, and Source columns were obtained from the same manufacturer. The Magic C-18 column for liquid chromatography–mass spectrometry was from Michrom BioResources, Auburn, CA, USA. Uracil, used as a standard, was from Aldrich Chemical Company, Milwaukee, WI, USA. Glycerol was purchased from Spectrum Chemical Manufacturer, Gardena, CA, USA. The milfoil polyphenol, tellimagrandin II [β-1,2,3-tri-O-galloyl-4,6-(S)-hexahydroxydiphenoyl-d-glucose], was a gift from E. Gross, Limnological Institute, University of Konstanz, Germany.
Adult E. lecontei preferences were determined using a two-way choice test performed in Y-tubes according to published procedures described in Solarz and Newman (1996, 2001). Assays were designed to maximize the behavioral analysis with small amounts of material. The experimental apparatus consisted of a glass Y-tube, 1.2-cm diam with 14-cm-long arms and stem. Each end of the tube was sealed with a cork, and the Y-tube was filled with well water. The control material was randomly placed into one arm of the Y-tube. Control material was either a plastic plant or a blank tube when fresh milfoil was used or a 6-mm diam filter paper saturated with well water or solvent for the chemical choice tests. Sample material, similarly prepared, was introduced into the second arm, and a weevil into the stem of the Y-tube. The Y-tube stem was covered with an opaque cloth and placed at a 10° incline (Solarz and Newman, 1996). A choice was indicated by the weevil swimming to or within 1 cm of either the control or sample material within a 10-min period. Choice and the time required to make that choice were recorded. No choice weevils were excluded from analyses. Weevils usually made a choice within 5 min. There was no difference in weevil choice between a plastic plant and a blank tube (N = 20, P = 0.37); therefore, either one was used as a control. As an alternative to the plastic plant or blank tube, exudates from a common co-occurring aquatic plant, C. demersum, were prepared as above and also tested against M. spicatum exudates and isolated fractions. Y-tubes were rinsed with 95% ethanol, then rinsed three times with well water after every three to six trials to ensure that previous test weevils did not affect the behavior of subsequent ones, and that chemicals did not diffuse into both arms of the Y-tube (Solarz and Newman, 2001). Throughout the isolation, Y-tube choice tests were used to determine whether weevils were attracted to the fractions.
For Y-tube bioassays, an aliquot from a known concentration of sample was infiltrated into filter paper disks, then used in the choice test. For dose–response curves, a range of concentrations of M. spicatum exudates and each isolated chemical was infiltrated. An estimate of the dose perceived by the weevils is the applied amount (in mg or mol) divided by the volume of one arm of the Y-tube (12 ml). This value in mg/l or molar amounts is indicated in the appropriate figure captions, and can be considered an average concentration that the weevils perceived during a trial. Likely, weevils were able to detect lower concentrations as the material dispersed throughout the Y-tube. Approximately 50 (range 20–127) randomly selected weevils were tested at each concentration level.
Chemical Isolation and Identification
Plant exudates and partially purified fractions were kept frozen at −20°C. Exudates were thawed, filtered through Whatman No. 1 cellulose filter paper, and lyophilized. The powder was resuspended in Milli-Q water and used in the Y-tube bioassay to confirm attraction. The resuspended exudate was fractionated into large and small molecules by using ultrafiltration. Ultrafiltration was performed with a 76-mm YM1 membrane (MWCO 1000) in a model 8400 Diaflo Amicon ultrafiltration cell under a 70-psi N2 atmosphere. Both the retentate (mw > 1000) and the filtrate (mw < 1000) were tested vs. water in the Y-tube bioassay.
Active filtrate was loaded onto a DEAE-anion exchange column (3.5 × 10cm column) that had been previously equilibrated with Milli-Q water. The column was eluted with Milli-Q water at a flow rate of at 1.7 ml/min. The effluent was monitored at 254 nm, and fractions were collected. A single UV-absorbing peak was observed. Fractions from this peak were pooled, concentrated by lyophilization, and tested in the Y-tube bioassay. Material that was eluted with 0.5 M NaCl was also collected, but showed no activity in the bioassay.
The DEAE pool was further purified by size exclusion chromatography on a G-10 Sephadex column (30 × 2.5 cm) that had been previously equilibrated with Milli-Q water. The column was eluted with water at a flow rate of 0.3 ml/hr, the effluent monitored at 254 nm, and fractions collected. Four peaks were eluted, and fractions from each were pooled separately, concentrated by lyophilization, and tested in Y-tube bioassays.
APCI analyses were performed on a ThermoElectron LCQ Classic Ion Trap mass spectrometer (Waltham, MA, USA) fitted with an APCI ionization source. The APCI heated capillary temperature was set at 450°C and controlled at 10 μA. Xcaliber™ (ThermoFinnigan Corp.) was used to acquire the data. Several LC-MS runs with C18, C8, and Source columns correlated the absorption spectrum to the molecular weight, but did not further separate components of the active fraction.
Samples for NMR spectroscopy were dissolved in D2O. NMR spectra were performed on Varian Inova 500, 600, and 800 spectrometers (University of Minnesota) using standard pulse sequences and water suppression. Identity of the active compounds was confirmed by comparison to standards using 1H, 13C, 1H–1H correlated spectroscopy (COSY), 1H–13C COSY, MQ COSY, heteronuclear multiple quantum coherence (HMQC), and heteronuclear multiple-bond correlation (HMBC) NMR spectra.
Uracil content of partially purified plant exudates (DEAE pool) was measured with a QTrap LC/MS/MS hybrid triple quadrupole/linear ion trap mass spectrometer (Applied Biosystems, Inc., ABI, Foster City, CA, USA). A 5-μl sample was withdrawn with a Agilent 1100 Micro Autosampler and transferred by an Agilent 1100 Capillary Pump to a Magic C18 column (150 × 0.2 mm) with 5-μm, 200-Å pore size particles. Components were sequentially eluted with a gradient. Solvent A was 98:2 water/acetonitrile containing 0.1% formic acid. Solvent B was 5:95 water/acetonitrile containing 0.1% formic acid. Using a flow rate of 5 μl/min, the gradient went from 100% A to 75% A in 4 min, was held for 3 min at 75% A, then ramped over 3 min to 25% A. The LC system was in line with the QTrap equipped with a TurboIonSpray ion source. The electrospray voltage was set at 5500 V. Samples were analyzed using a multiple reaction monitoring scan mode. Ions with a m/z of 113.10 in positive polarity were detected in quadrupole 1, autofragmented in quadrupole 2, and a m/z of 69.90 (afragment of uracil) was monitored in quadrupole 3. Analyzst 1.3.1 (ABI) wasused to acquire and analyze the data. A standard solution of 1.0 mg/ml of material that eluted from the DEAE column was spiked with uracil to a final concentration ranging from 0.09 to 900 μM to generate a standard curve. These samples were used to estimate the concentration of uracil present in DEAE pool samples from M. spicatum, M. sibiricum, M. spicatum × M. sibiricum hybrid, M. alterniflorum, M. tenellum, and C. demersum. Peak areas from standards were used to determine uracil concentrations.
Y-tube data were tested with a χ2 analysis. Dose–response curves for exudates and each isolated attractant were developed using logistic regression based on the amount of sample infiltrated into filter disks used in Y-tube bioassays. Comparisons among models were made using χ2 analyses of model deviances. Weevil preference was determined in the Y-tube bioassay. Glycerol and uracil were combined in Y-tube bioassays to test for possible synergistic effects. Differences in uracil content between M. spicatum, M. sibiricum, and M. spicatum × M. sibiricum hybrid were determined with an analysis of variance.
To determine the chemical identity of the weevil attractant(s), exudates of M. spicatum were tested for weevil preference, lyophilized, and retested for weevil preference (Table 1B). Weevils were attracted to certain M. spicatum exudates (P < 0.001) but not to M. sibiricum exudates (P = 0.62) or C. demersum exudates (P = 0.41). The M. spicatum exudates were separated into low-molecular weight (<1000 amu) and high-molecular weight (>1000 amu) compounds. Weevils were attracted to the low-molecular weight fraction (P < 0.001) and preferred it to C. demersum exudates (P = 0.017; Table 1C). The low-molecular weight fraction was purified using an anion exchange column, and weevils were attracted to the one major peak that passed through this column (P < 0.001) and preferred that peak to M. sibiricum exudates (P = 0.028; Table 1D). The anion exchange column pool was purified by size exclusion chromatography, and weevils were attracted to the third (P = 0.001) of the four peaks that eluted from the column (Table 1E). The active G-10 pool contained several components as determined by MS, and was further purified by HPLC. Mass spectral analysis indicated that the attractant consisted of two major components.
Chemical Isolation and Identification
Uracil and glycerol were identified as active components of the M. spicatum. Monitoring of column effluents by UV absorption led to the isolation of uracil, which has an UV absorption maximum at 259 nm. Glycerol does not absorb in the ultraviolet spectrum; however, its presence in the attractive fraction of the M. spicatum exudates was apparent in NMR spectra of the final fraction. Uracil was identified by low-resolution positive ion APCI and by high-resolution chemical ionization mass spectral analysis. Direct infusion APCI of the isolated fraction showed a base peak ofm/z 113.1 [M + H]+, which, upon fragmentation, yielded peaks at 95.9 and 69.9 m/z. Under high-resolution CI, the molecular ion of 113.0346 [M + H]+ was observed consistent with a molecular formula of C4H4N2O2. Electron ionization MS also showed the fragmentation pattern characteristic of glycerol (61.0, 43.0, and 31.0 m/z values).
The presence of uracil and glycerol was confirmed by NMR spectroscopy. 1H-NMR spectra showed an aromatic region with peaks at 5.4 and 7.4 ppm corresponding to uracil and a multiplet at 3.6 ppm and two doublet of doublets at 3.4 and 3.5 ppm corresponding to glycerol. The 13C-NMR spectra revealed peaks at 165.5, 153.4, 143.7, and 101.2 ppm corresponding to uracil and 72.2 and 62.6 ppm corresponding to glycerol. The COSY 1H-NMR spectra showed that the aromatic protons coupled at 5.6 and 7.4 ppm were consistent with thespectral properties of uracil. The MQ COSY spectra confirmed the coupling of the 3.4- and the 3.5-ppm doublet of doublets with each other and the multiplet consistent with the spectrum of glycerol. HMQC and HMBC spectra permitted the correlation of protons and carbons. The identities of uracil and glycerol wereconfirmed by comparisons to standards. A third compound was also present in some samples, but could not be separated from other components or identified.
The presence of uracil in the exudates of M. spicatum, M. sibiricum, M. alterniflorum, M. tenellum, and C. demersum was determined using a Qtrap LC-MS. Uracil was present in the exudates of all of the aquatic plants tested (Figure1). Concentration of uracil was higher in M. spicatum than in M. sibiricum or the M. spicatum × M. sibiricum hybrid (ANOVA, F2,10 = 11.63, P= 0.004).
Y-Tube Choice Tests of Isolated Attractants and the Polyphenol, Tellimagrandin II, Found in M. spicatum
C. demersum exudates
Glycerol and uracild
The slopes of the curves in Figures 2–4 can be used to indicate whether the weevil is responding differently to increasing doses of the different chemicals. Glycerol and uracil curves are significantly different (χ2 goodness of fit test, χ2= 12.1, df = 2, P = 0.002), but the curve of the exudates is not significantly different from that of either uracil or glycerol. This suggests that the exudates attract the weevil at a level between that of glycerol and uracil.
This study together with the previous work by Solarz and Newman (1996) provides evidence that a fully aquatic specialist insect is attracted to plant-released chemicals. Solarz and Newman (1996) demonstrated that E. lecontei was attracted to M. spicatum in the dark as well as in the light. Using bioassay-driven fractionation, we identified uracil and glycerol as two compounds released from M. spicatum that attract E. lecontei. Although herbivore attraction to plant-released chemicals is a well-known phenomenon in terrestrial systems, this interaction has not previously been shown for a submersed freshwater plant and insect herbivore.
In two-way choice tests, E. lecontei preferred M. spicatum to a plastic plant (control), its native host M. sibiricum, and the co-occurring unrelated macrophyte C. demersum. Although historically associated with M. sibiricum (Creed and Sheldon, 1994), the weevils used in this experiment did not prefer M.sibiricum or its exudates to either a control or to C. demersum. This lack of preference for M. sibiricum may be a response to the host plant on which weevils developed (collected from M. spicatum at Lake Auburn). Solarz and Newman (2001) found that larvae reared on or adults exposed to M. spicatum preferred M. spicatum, but larvae reared on M. sibiricum had no preference. Alternatively, M. sibiricum and E. lecontei are presumed to have coevolved, and, therefore, M. sibiricum may exude deterrents or lack a specific attractant. The weevil's response to one co-occurring species, C. demersum, and three additional milfoils, the M. spicatum × M. sibiricum hybrid, the morphologically similar M. alterniflorum, and the morphologically dissimilar M. tenellum, all of which have some uracil in their exudates, was variable. Only the M. spicatum × M. sibiricum hybrid attracted weevils.
Attraction of E. lecontei adults to M. spicatum is a result of chemical compounds in exudates that were released by M. spicatum (Table 1B). Myriophyllum spp. release chemicals, including sugars, proteins, cyanogens, tannic substances, and polyphenols, into the water (Lyr and Streitberg, 1955; Godmaire and Nalewajko, 1990; Nakai et al., 2000; Gross, 2003). By testing M. spicatum exudates, we avoided weevil exposure to chemicals that may be present in the plant tissue, but are not released into the water (Hay et al., 1998).
Like terrestrial plants that release volatiles as a result of the oxidative degradation of leaf lipids (Bernays and Chapman, 1994; Paré and Tumlinson, 1999), aquatic plants release hydrophilic metabolites as a by-product of rapid growth (Nalewajko and Godmaire, 1996). Peptides with water-soluble groups were suggested by Decho et al. (1998) to be good attractants in aquatic systems. Glycerol and uracil are two ubiquitous, low-molecular weight compounds that fit this profile.
Glycerol is a common metabolite and can also protect against abiotic stresses. The identification of glycerol-specific transporters in cell membranes of Arabidopsis thaliana indicates that it plays a role as a general osmolyte (Weig and Jakob, 2000; Eastmond, 2004). Glycerol content in the leaves of several agriculturally important plants such as maize and alfalfa typically rangesfrom roughly 0.1–0.4 mM (10–39 μg/g wet weight) to as high as 6 mM (554 μg/g wet weight) in Zea mays after a rain event, which may increase the demand for intracellular osmolytes (Gerber et al., 1988). Carbohydrates are also found on leaf surfaces of terrestrial plants at concentrations up to 0.01 mM at which point they act as feeding stimulants for insects (Bernays and Chapman, 1994). Carbohydrates located on leaf surfaces of aquatic plants will dissolve in water and may then function as foraging kairomones that attract herbivores (Ruther et al., 2002). Glycerol produced by M. spicatum attracts E. lecontei at0.018- to 1.8-mM range, which is similar to the range that attracts insects to similar compounds in terrestrial systems (Bernays and Chapman, 1994).
Uracil is effective at a much lower concentration. Weevils were significantly attracted to uracil at 0.015–15 μM. At these concentrations, E. lecontei adults are responding at a concentration more indicative of a plant-specific attractant than a nutrient. For example, the terrestrial weevil Ceutorhynchus assimilis adults were attracted to a 120 μM phenylacetonitrile, a metabolite common in their host plant Brassica napus (Bartlet et al., 1997). In marine systems, shrimp Palaemonetes pugio were attracted to amino acids and nucleotides at a concentration of 0.1–1000 μM, a range that Carr (1988) suggests is intermediate between background noise and the concentration in prey tissue. Although it is not a species-specific attractant, the presence of uracil in the water may indicate either a damaged plant or an actively growing one suitable to E. lecontei for feeding and oviposition.
Like glycerol, uracil is also ubiquitous in plants. We found it in exudates of all plants tested, including all Myriophyllum spp. and C. demersum. However, its presence in exudates raises the question as to why a plant would release anexpensive metabolite. In plants, uridine nucleosides and nucleotides are the second most common in cells after adenosine. In Arabidopsis, ureide permease transporters have been identified as the main transporters of uracil in phloem (Schmidt et al., 2004). Uracil in exudates may be a function of its abundance and ubiquity in rapidly growing cells, such as the apical meristems, where pyrimidine salvage is needed to support the high demand for nucleotides. Of the pyrimidines, plants can only salvage uracil (Wagner and Backer, 1992).
An alternative use for uracil in plants is the production of alkaloids and pseudoalkaloids (sensu Waterman, 1998). Plants that use uracil in this capacity have pools of the pyrimidine for rapid synthesis (Brown and Turan, 1995). Ostrofsky and Zettler (1986) reported that alkaloids were present in M. spicatum. However, no alkaloids or pseudoalkaloids were detected in our collections (Stermitz, Colorado State University, unpublished data, see Marko and Stermitz, 1997 for methods), suggesting that uracil-based natural products are not produced in these aquatic plants, or that they are found in low concentrations.
The concentration of uracil was highest in M. spicatum exudates, intermediate in M. spicatum × M. sibiricum hybrid, and lowest in M. sibiricum. M. spicatum grows faster than M. sibiricum and starts growing earlier in the season (Madsen et al., 1991). Therefore, it would produce and release glycerol and uracil sooner than other co-occurring plants.
The presence of uracil in aquatic plant exudates may also be a result of damage because of abiotic stressors such as UV radiation that can damage nucleic acids. The sun-exposed meristem tips of M. spicatum form dense mats at the water surface. This greater surface exposure may lead to more UV stress than that experienced by other milfoils located below the surface, potentially resulting in the loss of cell contents, including uracil. Ultraviolet radiation can also induce the production of polyphenols in aquatic plants (Meijkamp et al., 1999; Rozema et al., 1999). The higher concentration of polyphenols found in M. spicatum compared to M. sibiricum (Spencer and Ksander, 1999; Marko etal., unpublished results) suggests that it is under greater UV stress.
Polyphenols in M. spicatum deter feeding by herbivores (Leu et al., 2002; Li et al., 2004). Li et al. (2004) found that M. spicatum had higher phenolic content than co-occurring aquatic plants, and that in two-way choice tests, the snail Radix swinhoei ate less M. spicatum than five of the seven tested plants. One polyphenol in M. spicatum, tellimagrandin II, is known to inhibit the growth of periphyton and deter the growth of the lepidopteran herbivore Acentria ephemerella (Gross et al., 1996; Choi et al., 2002). However, tellimagrandin II was neither attractive nor deterrent to E. lecontei at the concentrations tested. This suggests that the weevil does not respond to all M. spicatum-released chemicals, but rather that it is specifically attracted to glycerol and uracil.
Host plant attraction often involves a mixture of chemicals, and the effective concentration of one attractant can be modified by a specific mixture (Roseland et al., 1992; Bartlet et al., 1997). For example, C. assimilis adults were more attracted to phenylacetonitrile plus a mixture of isothiocyanates than to phenylacetonitrile alone (Bartlet et al., 1997). Synergism between glycerol and uracil was not detected in the M. spicatum–E. lecontei system. Similar percentages of E. lecontei adults were attracted to M. spicatum exudates, uracil, and a combined fraction of uracil + glycerol.
In both aquatic and terrestrial systems, there is considerable variation in herbivore response to chemical attractants, and sometimes, it is the presence of a trace chemical that drives the plant–insect interaction. For example, in terrestrial systems, Hylobius pales (Curculionidae) adults show a dose-dependent response to some chemicals, but not others (Salom et al., 1994). Furthermore, weevil response can depend on whether the study is conducted in the field or in the lab. In field studies, C. assimilis exhibited a dose response to known attractants (Smart et al., 1997), but not under laboratory conditions (Bartlet et al., 1997). Bartlet et al. (1997) suggest that the ratio of chemicals and the presence of trace compounds may ultimately affect behavior in the field. Variation in E.lecontei behavior to M. spicatum populations was observed throughout these experiments. Certain populations of M. spicatum did not attract the weevils, and environmental variables such as weather also affected weevil response. Because M. spicatum is a nonindigenous plant, it may release chemicals that attract E.lecontei, whereas native milfoils may have been selected to release fewer attractant chemicals. Undoubtedly, the attraction of E. lecontei to M. spicatum is a complex process that involves a variety of factors.
In terrestrial systems, ubiquitous plant volatiles, such as phenylpropanoids, fatty acid derivatives, and isoprenoids, are involved in the attraction and deterrence of insects to their host plants (Bruce et al., 2005). In aquatic systems, common low-molecular weight compounds, such as amino acids, sugars, and nucleotides, perform the same function (Weissburg et al., 2002). These water-soluble compounds may be reversibly bound to leaf surfaces or released from trichomes. Aquatic herbivores probably rely on taste receptors rather than odor for host location (Spänhoff et al., 2005). Additionally, because many of these chemicals are readily taken up by bacteria, they may act as short-term signals. Alternatively, compounds released by plants may be degraded by epiphytes, and the degradation products may serve as semiochemicals. Comparisons of chemical mobility and persistence in different environments will provide an understanding of the ecological and evolutionary relationships among aquatic insects and their host plants.
We thank Elisabeth M. Gross for the gift of tellimagrandin II and Frank R. Stermitz for analysis of alkaloid content. We are grateful to Tom Krick, Letitia Yao, Beverly Ostrowski, and LeeAnn Higgins for assistance and advice with the mass spectral and NMR analyses. We thank M. Moody for molecular identification of Myriophyllum spp. Assistance with specimen collection and analyses were provided by D. Ward, C. Lemmon, S. Daugherty, S. Coloso, J. Dehn, K. Eichstaedt, C. McCollum, and K. Mann. We thank LeeAnn Higgins and two anonymous reviewers for their comments on the manuscript. This work is sponsored by the Minnesota Sea Grant College Program supported by the NOAA Office of Sea Grant, United States Department of Commerce, under grant no. NOAA-NA16-RG1046. The U.S. government is authorized to reproduce and distribute reprints for government purposes, not withstanding any copyright notation that may appear hereon. Additional support was provided by the Minnesota Agricultural Experiment Station and the University of Minnesota Graduate School. This paper is journal reprint no. JR 511 of the Minnesota Sea Grant College Program.
- Bernays, E. A., Chapman, R. F. 1994Host Plant Selection by Phytophagous InsectsChapman & HallNew YorkGoogle Scholar
- Buckingham, G. R., Bennett, C. A. 1995Host range studies with Bagous affinis (Coleoptera: Curculionidae), an Indian weevil that feeds on hydrilla tubersEnviron. Entomol.27469479Google Scholar
- Carr, W. E. S. 1988
The molecular nature of chemical stimuli in the aquatic environmentAtema, J.Fay, R. R.Popper, A. N.Tavolga, W. N. eds. Sensory Biology of Aquatic AnimalsSpringer-VerlagNew York327Google Scholar
- Center, T. D., Dray, F. A., Jubinsky, G., Grodowitz, M. J. 2002Insects and Other Arthropods that Feed on Aquatic and Wetland PlantsUSDA, Agricultural Research ServiceFort Lauderdale, FLGoogle Scholar
- Creed, R. P., Sheldon, S. P. 1994Aquatic weevils (Coleoptera, Curculionidae) associated withnorthern watermilfoil (Myriophyllum sibiricum) in Alberta, CanadaEntomol. News105 98102Google Scholar
- Creed, R. P., Sheldon, S. P. 1995Weevils and watermilfoil: Did a North American herbivore cause the decline of an exotic plant?Ecol. Appl.511131121Google Scholar
- Cronin, G., Lodge, D. M., Hay, M. E., Miller, M., Hill, A. M., Horvath, T., Bolser, R. C., Lindquist, N., Wahl, M. 2002Crayfish feeding preferences for fresh water macrophytes: The influence of plant structure and chemistryJ. Crustac. Biol.22708718Google Scholar
- Hay, M. E., Steinberg, P. D. 1992
The chemical ecology of plant–herbivore interactions in marine versus terrestrial communitiesRosenthal, G. A.Berenbaum, M. R. eds. Herbivores: Their Interactions with Secondary Plant Metabolites, 2EAcademic PressSan Diego, CA371413Google Scholar
- Hay, M. E., Stachowicz, J. J., Cruz-Rivera, E., Bullard, S., Deal, M. S., Lindquist, N. 1998
Bioassays with marine and freshwater macroorganismsHaynes, K. F.Millar, J. G. eds. Methods in Chemical Ecology, Vol. 2, Bioassay MethodsChapman and HallNew York39141Google Scholar
- Kubanek, J., Hay, M. E., Brown, P. J., Lindquist, N., Fenical, W. 2001Lignoid chemical defenses in the freshwater macrophyte Saururus cernuusChemoecology1118Google Scholar
- Lyr, H., Streitberg, H. 1955Die Verbreitung von Hydropoten in verschiedenen Verwandtschaftskreisen der WasserpflanzenWiss. Z.–Martin-Luther-Univ. Halle-Wittenb., Math.-Nat.Wiss. Reihe4471484Google Scholar
- Madsen, J. D., Hartleb, C. F., Boylen, C. W. 1991Photosynthetic characteristics of Myriophyllum spicatum and six submersed aquatic macrophyte species native to Lake George, New YorkFreshw. Biol.26233240Google Scholar
- Meijkamp, B., Aerts, R., Staaij, J., Tosserams, M., Ernst, W. H. O., Rozema, J. 1999
Effects of UV-B on secondary metabolites in plantsRozema, J. eds. Stratospheric Ozone Depletion: The Effects of Enhanced UV-B Radiation on Terrestrial EcosystemsBackhuys PublishersLeiden, The Netherlands7199Google Scholar
- Metcalf, R. L., Metcalf, E. R. 1992Chemical Ecology of Plant KairomonesChapman and HallNew York, NYGoogle Scholar
- Nalewajko, C., Godmaire, H. 1996Extracellular products of Myriophyllum spicatum L. as a function of growth phase and diel cycleArch. Hydrobiol.127345356Google Scholar
- Newman, R. M., Borman, M. E., Castro, S. W. 1997Developmental performance of the weevil Euhrychiopsis lecontei on native and exotic watermilfoil host plantsJ. North Am. Benthol. Soc.16627634Google Scholar
- Ostrofsky, M. L., Zettler, E. R. 1986Chemical defenses in aquatic plantsJ. Ecol.74279287Google Scholar
- Rozema, J., Staaij, J., Olof Bjorn, L., Bakker, N. 1999
Depletion of stratospheric ozone and solar UV-B radiation: Evolution of land plants, UV-screens and functions of polyphenolicsRozema, J. eds. Stratospheric Ozone Depletion: The Effects of Enhanced UV-B Radiation on Terrestrial EcosystemsBackhuys PublishersLeiden, The Netherlands119Google Scholar
- Ruther, J., Meiners, T., Steidle, J. L. M. 2002Rich in phenomena-lacking in terms. A classification of kairomonesChemoecology12161167Google Scholar
- Salom, S. M., Carlson, J. A., Ang, B. N., Grosman, D. M., Day, E. R. 1994Laboratory evaluation of biologically-based compounds as antifeedants for the Pales weevil, Hylobius pales (Herbst) (Coleoptera: Curculionidae)J. Entomol. Sci.29407419Google Scholar
- Sheldon, S. P., Creed, R. P. 1995Use of a native insect as a biological control for an introduced weedEcol. Appl.511221132Google Scholar
- Smart, L. E., Blight, M. M., Hick, A. J. 1997Effect of visual cues and a mixture of isothiocyanates on trap capture of cabbage seed weevil, Ceutorhynchus assimilisJ. Chem. Ecol.23889902Google Scholar
- Smith, C. S., Barko, J. W. 1990Ecology of Eurasian watermilfoilJ. Aquat. Plant Manage.285564Google Scholar
- Spänhoff, B., Kock, C., Meyer, A., Meyer, E. I. 2005Do grazing caddisfly larvae of Melampophylax mucoreus (Limnephilidae) use their antennae for olfactory food detection?Physiol. Entomol.30134143Google Scholar
- Spencer, D. F., Ksander, G. G. 1999Phenolic acids and nutrient content for aquatic macrophytes from Fall River, CaliforniaJ. Freshw. Ecol.14197209Google Scholar
- Wagner, K. G., Backer, A. I. 1992Dynamics of nucleotides in plants studied on a cellular basisInt. Rev. Cyt.134184Google Scholar
- Waterman, P. G. 1998
Chemical taxonomy of alkaloidsRoberts, M. F.Wink, M. eds. Alkaloids: Biochemistry, Ecology, and Medicinal ApplicationsPlenum PressNew York87106Google Scholar
- Weig, A. R., Jakob, C. 2000Functional identification of the glycerol permease activity of Arabidopsis thaliana NLM1 and NLM2 proteins by heterologous expression in Saccharomyces cerevisiaeFEBS Lett.281293298Google Scholar