Biological Control 49 (2009) 169–179
Contents lists available at ScienceDirect
Biological Control
journal homepage: www.elsevier.com/locate/ybcon
Biology of the galling wasp Tetramesa romana, a biological control agent of giant reed q
Patrick J. Moran *, John A. Goolsby
United States Department of Agriculture, Agricultural Research Service, Beneficial Insects Research Unit, 2413 E Hwy 83, Weslaco, TX 78596, USA article info abstract
Article history: The biology of the gall-forming wasp Tetramesa romana Walker (Hymenoptera: Eurytomidae) from Received 6 August 2008 southern France and Spain was studied to determine its suitability as a biological control agent of giant Accepted 28 January 2009 reed (Arundo donax L.), an exotic and invasive riparian weed in the US and Mexico. Females produced eggs Available online 5 February 2009 parthenogenetically and deposited them into shoot tips. Eggs were 1.26 mm long and hatched within 5 days of oviposition. The larvae completed three instars within 23–27 days, with the first and second Keywords: instars lasting 5 days each at 27 °C and the third instar lasting 12 days. Larvae increased in length from Galling wasp 0.9 mm to 4.1 mm between the first and last instars. The total generation time averaged 33 days. Adults Eurytomidae were 6.8 mm long with antennae and lived 3.7 days. Spanish wasps lived 1.3 days longer than French Development Reproduction wasps. Most (80%) parthenogenetic reproduction involved wasps that were less than 5 days old. A total Oviposition behavior of 36% of French wasps tested and 65% of Spanish wasps produced offspring. Individual Spanish wasps Exotic invasive grass produced an average of 26 (up to 66) progeny, displaying an intrinsic capacity for increase of 0.081 and a population doubling time of 9.2 days. Over 90% of exit holes made by emerging wasps were located within two nodes of the shoot tip. Reproductive wasps made 2.5-fold more probes than did wasps that failed to gall stems, and probing behavior was 13% more prevalent in the afternoon than in the morning. The Spanish genotypes of T. romana are particularly suitable for mass production to support field releases. Published by Elsevier Inc.
1. Introduction native plants and animals (Bell, 1997; Herrera and Dudley, 2003), leading, for example, to the possible extinction of a Mexi- 1.1. Invasion of giant reed in North America can fish species (Etheostoma segrex Norris & Minckley) (McGaugh et al., 2006). Infestations of giant reed cause similar problems in Giant reed (Arundo donax L., Poaceae), along with many other southern California (Bell, 1997; Dudley, 2000; Quinn and Holt, food and fiber plants, was introduced to North America ca. 400 2008) and in the Sacramento-San Joaquin Delta in northern Cali- years ago from Mediterranean Europe (Dunmire, 2004), with pos- fornia (Spencer et al., 2006). Giant reed was brought to the New sible introductions from other native areas, including northern World for use in thatch housing and fences, to make woodwind Africa, the Middle East, the Arabian Peninsula, India and Nepal musical instruments (Perdue, 1958), and for erosion control (Perdue, 1958). Giant reed has invaded riparian habitats in the (Tracy and DeLoach, 1999). Several A. donax cultivars and other southern US and northern Mexico (Bell, 1997), forming dense Arundo species are occasionally used as ornamental grasses and infestations along 600 river-miles of the Lower Rio Grande in as forage for exotic mammals in the southern US (Greenlee, Texas (Everitt et al., 2005), with additional populations through- 1992; Oakes, 1993). Giant reed is a source of industrial cellulose out the Lower Rio Grande Basin (LRGB), which includes Texas (Perdue, 1958) and a potential biofuel (Szabo et al., 1996). How- and five Mexican states. Giant reed consumes water from the ever, the economic and ecological damage caused by giant reed Rio Grande and its tributaries that is needed for agricultural outweighs the potential benefits of this species. Physical (burn- and domestic use. Large thickets alter stream flow patterns, in- ing), mechanical (mowing or mulching) and chemical control crease erosion and flood damage (Frandsen and Jackson, 1994), methods are available (Tracy and DeLoach, 1999), but may not fuel wildfires along rivers (Rieger and Kreager, 1989), limit river have sufficient impact, in part because giant reed regrows access (Hughes and Mickey, 1993) and displace populations of vigorously from its well-provisioned rhizomes (Decruyenaere and Holt, 2001; Spencer and Ksander, 2006).
q Mention of trade names or commercial products in this article is solely for the 1.2. Potential for biological control purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. * Corresponding author. Fax: +1 956 969 4888. Biological control may provide sustainable control of giant reed E-mail address: [email protected] (P.J. Moran). in North America and in other areas, such as South Africa (Rossa
1049-9644/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.biocontrol.2009.01.017 170 P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179 et al., 1998). Although grasses have been considered difficult tar- control of Cordia curassavica (Jacq.) R. & S. in Malaysia (Simmonds, gets for biological control with arthropods (Evans, 1991) because 1980). of their relatively simple architecture and reduced loads of chem- ical deterrents, toxins, and stimulants (Strong et al., 1984), 1.4. Biology of T. romana monophagy and oligophagy are in fact common among endopha- gous herbivores such as shoot-borers feeding on perennial grasses Little information is available about the biology of Tetramesa (Tscharntke and Greiler, 1995). Evaluations of potential non-native species, in particular the number and duration of immature stages, agents (some adventive) of another invasive perennial grass, the and the fecundity and longevity of adults. Female reproduction Eurasian form of common reed, Phragmites australis (Cav.) Trin likely involves thelytokous parthenogenesis, as in many galling ex. Steudel, are ongoing (Tewksbury et al., 2002; Häfliger et al., wasps in the Cynipidae (Askew, 1984), since males are rare or un- 2005, 2006). In the case of giant reed, 21 mostly polyphagous in- known in most Tetramesa species (Phillips, 1936; Claridge, 1961). A sect herbivores were identified (Tracy and DeLoach, 1999), with recent DNA analysis indicated that geographic populations of T. oligo- or monophagous species expected among the endophagous romana are genetically variable (D. Tarin, A. Pepper, J. Manhart, fauna. More recent surveys have focused on Mediterranean Europe Texas A&M University, College Station, Texas, unpublished data), as a source for candidate agents (Kirk et al., 2003), and a classical and could therefore differ in life history parameters. Tetramesa biological control program for A. donax is in development. romana may sometimes cause no visible gall (Steffan, 1956; Cla- ridge, 1961). However, galls are visible as swollen protuberances on side shoots in southwestern France and in Spain (Kirk et al., 1.3. Selection of Tetramesa romana for biological control 2003). New adult wasps emerge by chewing exit holes from the gall chamber to the outer surface of the stem. Field observations Biological control of non-native weeds using exotic arthropod (Kirk et al., 2003; Dudley et al., 2006) indicate that galls are found herbivores has led to partial or complete control of over 30 inva- at shoot tips. Past studies of other Tetramesa spp. (Dubbert et al., sive weeds in North America since 1946 (Nechols et al., 1995; Coul- 1998) and an adventive population of T. romana in California, son et al., 2000). The decision to release an agent into the field USA (Dudley et al., 2006) suggest that females perceive external must be grounded on a thorough understanding of its biology cues that provide information about stem diameter. Ovipositor (development, feeding, survival, and reproduction) in order to pre- probing behavior in stems likely exposes females to plant spe- dict the agent’s role in introduced environments (Price, 2000). Bio- cies-specific chemical cues (Askew, 1984). Wasp oviposition may logical and ecological information from the native range can be vary according to light quantity or daylength, as in the case of used to prioritize agents (Myers, 1987; van Klinken and Raghu, other herbivores evaluated for biological control, such as the saltc- 2006). Suitable agents must meet modern host specificity require- edar leaf beetle Diorhabda elongata Brulle subspecies deserticola ments, which are based on a thorough understanding of agent host Chen Desbrochers (Bean et al., 2007). Past studies on temperate location, feeding behavior and immature and mature survival Tetramesa species indicated that wasps complete only one or two (Heard, 2000; Sheppard et al., 2005). Candidate agents must show generations per year (Phillips, 1936; Claridge, 1961). The objective efficacy on the target weed, at least when protected from natural of this study was to characterize the immature development, gen- enemies (McFadyen, 1998; Pearson and Callaway, 2003; Sheppard eration time, and adult survival, reproductive output and oviposi- and Raghu, 2005; McClay and Balciunas, 2005; Goolsby et al., tion behavior of T. romana from France and Spain under 2006), and they should be compatible with climatic conditions in laboratory conditions. the highest-priority ecoregions of the weed’s invasive range (Bri- ese, 2004), as well as with the genotypes and demographies of the target weed that are prevalent in these ecoregions (Müller- 2. Materials and methods Schärer and Schaffner, 2008). In the case of giant reed, one of the agents selected for evaluation on the basis of safety criteria, broad 2.1. Experimental organisms geographic distribution and abundance (Kirk et al., 2003) is the stem-galling wasp Tetramesa romana Walker (Hymenoptera: 2.1.1. Arundo donax Eurytomidae). Arundo donax is a rhizomatous perennial grass, up to 8 m tall, The wasp T. romana has been reported from southern France that thrives in southern temperate and subtropical North America (Steffan, 1956), Italy and Egypt (Claridge, 1961), Spain, Greece, Tur- in riparian zones along river systems, reservoirs, and canals, form- key and the northern coast of Africa (Kirk et al., 2003). Other mem- ing impenetrable thickets in floodplains (Everitt et al., 2005). It can bers of the genus Tetramesa are known to be highly host-specific grow on wide range of soils, from loose gravelly and sandy sub- (Phillips, 1936; Claridge, 1961), restricted to a single host species strates to heavy clays and river sediments (Perdue, 1958; Spencer or genus in the family Poaceae, and they have developed associa- et al., 2008), of varying fertility (Quinn et al., 2007; Spencer and tions with inquilines and parasitoids that show similar levels of Ksander, 2006), and can tolerate a broad range of environments specificity (Dawah et al., 1995, 2002; Dubbert et al., 1998; Mat- (Rossa et al., 1998; Decruyenaere and Holt, 2001; Quinn and Holt, sumoto and Saigusa, 2001). Gall induction by Tetramesa spp. wasps 2008). can cause damage leading to economic losses in crops. Lodging and New shoots emerge throughout the year in warm climates such yield loss occur in North America and Spain in wheat (Triticum aes- as southern Spain and the LRGB of Texas and Mexico. Both rainfall tivum L.) galled by the wheat jointworm T. tritici (Fitch) and the and temperature influence shoot production by rhizomes, which wheat strawworm T. grandis (Riley), barley (Hordeum vulgare L.) often constitute over half of total biomass (Spencer et al., 2006; galled by the barley jointworm T. hordei (Harris), and rye (Secale Thornby et al., 2007; Spencer et al., 2008). Clonal populations ex- cereale L.) galled by the rye jointworm T. secale (Fitch) (Davis, pand rapidly as rhizomes grow and produce new shoots (Thornby 1918; USDA, 1954; Holmes and Blakeley, 1971; Sterling, 1976; et al., 2007). Rooting from plagiotropic (horizontally layered) stem Martin and Harvey, 1982; Cantero-Martínez et al., 2003; Shanower canes and broken stem fragments is also common (Wijte et al., and Waters, 2006). No other Tetramesa species has been evaluated 2005; Boland, 2006). Rhizomes and canes establish new popula- for biological weed control, although two species in the genus tions after being uprooted and dispersed by flooding (Bell, 1997). Eurytoma have been examined (Kleinjan and Edwards, 2006; Neser, Shoots become woody within the first year of growth and produce 2008), and one species (Eurytoma attiva Burks) has contributed to many axillary shoots. Giant reed produces large (30–60 cm) pani- P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179 171 cles from main shoots and occasionally side shoots throughout the men-curling indicative of oviposition probing behavior, and also year, but in the invaded range the seeds are sterile (DiTomaso and shows the egg, larval and pupal stages of T. romana. Healy, 2003). Panicle formation leads to shoot senescence. Senesc- Our native range collections (Kirk et al., 2003) and laboratory ing and dead shoots remain standing in many cases. Fire kills host range studies (Goolsby and Moran, 2009) indicate that T. shoots and breaks down dead biomass, but also favors generation romana is host-specific to Arundo spp. Stimuli provided by the par- of new shoots (Rieger and Kreager, 1989). ent wasp or by eggs lead to the development of gall tissues (Price, Despite the clonal nature of the giant reed invasion of North 2008), as the gall begins to form before the larvae hatch. Extra-oral America, multiple introductions may have occurred, and candidate digestion may occur, as the larvae appear to secrete liquids onto natural enemies that are most likely to be adapted to local geno- the gall tissue and then use their mandibles to rasp and collect gall types should be selected (Briese, 2004; Goolsby et al., 2006). A tissue (A. Cohen, Insect Diets and Rearing Research, Raleigh, North population genetics analysis (D. Tarin, A. Pepper, J. Manhart, Carolina, personal communication). At least two parasitoids attack unpublished data) is underway to determine the geographic points T. romana in the Mediterranean region: the pteromalid Gugolzia of origin of the small number of A. donax genotypes present in the harmolitae Delucchi & Steffan and the eurytomid Eurytoma steffani LRGB. Initial results indicated points of origin within Spain. Tetra- Claridge (Steffan, 1956). mesa romana was therefore collected from Spanish locations, as well as from a population near the USDA-ARS, European Biological 2.2. Plant and insect cultures Control Laboratory (EBCL) in coastal southern France. 2.2.1. Plants 2.1.2. Tetramesa romana Arundo donax rhizomes were collected in Laredo and San Juan, The wasp T. romana (Fig. 1A) is in the family Eurytomidae (Cla- Texas, at sites adjacent to the Rio Grande. The majority (70%) of ridge, 1961). Noyes (2004) describes the Family Eurytomidae, plants used in developmental and reproduction studies were from focusing mostly on Palearctic fauna. The subfamily Eurytominae Laredo, while observations of oviposition behavior involved plants includes Tetramesa (Lotfalizadeh et al., 2007). Claridge (1961), from San Juan. Rhizomes of A. donax were either potted immediately Noyes (2004), and Lotfalizadeh et al. (2007) provide morphological or stored at 10 °C at the USDA Animal and Plant Health Inspection features to separate the genus Tetramesa from other eurytomid Service (APHIS)-Pest Detection, Diagnostics and Management Labo- genera. Illustrations of the adult stage may be found in Zerova ratory (PDDML) in Edinburg, Texas. Rhizomes were potted in 4-L (1976). Diagnostic information for the adult stage was provided pots containing 80% peat moss soil with perlite (‘Sunshine Mix No. by G. Delvare (CIRAD, Montferrier-sur-Lez, France, unpublished 1’, Sun-Gro Horticulture, Bellevue, WA). Pots were placed in a green- data) based on examinations of a lectotype in the Natural History house regulated between 25–32 °C and 70–90% relative humidity Museum (London) and adults collected from France, Spain and (RH) with ambient light (12–14 h daylight per day). Pots containing Morocco. Observations on T. romana larval morphology and feed- one to three A. donax shoots were transferred to the quarantine lab- ing behavior were made by A. Cohen, Insect Diet and Rearing Re- oratory at the PDDML for exposure to T. romana after 4–6 weeks of search, Raleigh, North Carolina (personal communication). Fig. 1 growth, at which time shoots were 0.5–1.0-m tall with 0.7–1.3 cm shows photographs of the adult female at rest and exhibiting abdo- thick main shoots and no lateral shoots.
Fig. 1. Tetramesa romana biology. (A) Adult female (bar = 5.33 mm). (B) Abdomen-curling behavior indicative of stem probing and oviposition. (C) Swollen A. donax shoot tip indicative of galling with exit holes made by emerging adults. (D) Ovaries dissected from newly emerged female showing numerous eggs with stalk-like pedicels (bar = 1 mm). (E) Third-instar larva (bar = 5.95 mm). (F) Pupa (bar = 6.05 mm). 172 P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179
2.2.2. Wasps position noted relative to the shoot tip as marked at the time of Tetramesa romana wasps from Perpignan, France (accession parent wasp removal. number M06008), and the area surrounding Granada, Spain (acces- sion numbers M07064 and M07074) were reared on A. donax by 2.5. Oviposition behavior confining 2–4 newly emerged females in cylindrical cages (40 cm long by 12 cm wide) made of black silk organza with a metal wire To examine relationships between wasp oviposition behavior frame, secured to the A. donax shoot at the top and bottom with and reproductive success or time of day (morning or afternoon), twine. Two cotton balls, each saturated with purified water and individual T. romana wasps from Perpignan, France were observed one drop (ca. 0.5 g) of honey, were placed in leaf blade axils inside in cylindrical cages in a greenhouse maintained as above for up to the cage as an adult food source. Wasps were held on shoots for 1– 8 h per day for up to the first 5 days of their adult lives. Adults were 3 days in a quarantine greenhouse regulated between 20–30 °C transferred to new shoots daily, prior to the start of the next daily and 60–80% RH with ambient light (400–500 lmol mÀ2 sÀ1 at ca. observation period. The number of abdomen-curling events indic- 2 m above the ground at noon) supplemented between November ative of probing were counted, and time (minutes) spent in probing and March by sodium halide lights (total 14:10 L:D cycle). The behavior (Fig. 1B), resting on the stem, resting on the leaf, resting cages were lightly misted with water once per day. After wasp on the cage, and resting on a cotton plug containing honey and and cage removal, shoots were held in a growth chamber set to a water were summed across the wasp’s lifetime, and converted to constant 27 °C, 70% RH and 14:10 L:D cycle. Cages were placed proportions by dividing by the total time of observation for each on galled shoots 25 days after parent wasp removal. Newly wasp. Wasps were observed for a total of 177 h in the morning emerged wasps were collected daily for use in studies of T. romana (08:00–12:00) (mean ± SE, 4.8 ± 0.5 h per wasp over its lifetime) development, adult survival, reproductive output and behavior. and 267 h in the afternoon (13:00–17:00) (6.1 ± 0.5 h per wasp).
2.6. Data analysis 2.3. Development and generation time of T. romana All analyses used SAS software (Version 9.1.3) (SAS Institute, To determine the size and duration of the life stages of T. roman- 2004). Larval and pupal morphological measurements were aver- a, A. donax shoots (0.5–1 m tall) were each infested with 2–4 fe- aged by day and graphed. Decreases in average larval head capsule male wasps as for rearing, except that females were exposed to width and larval length, which interrupted the general upward shoots for only 1 day. The approximate time from oviposition to trends over time, were used to infer larval transition times and egg hatch was deduced from the appearance of larvae in dissected the duration of instars, based on the assumption that apolysis shoots 4–5 days after oviposition. Shoots were dissected daily from and ecdysis during molting interrupt larval growth in the pharate the 4th to the 33rd day after parent wasp removal (1–9 galled larva of the next instar (Chapman, 1982). Head capsule widths plants dissected each day, average (±SE) of 4 ± 2 plants per day, to- and larval lengths were compared across instars with multivariate tal of 155 plants dissected). Larval and pupal length (at 12.5Â or and univariate analyses of variance (ANOVAs), using measure- 25Â) and larval head capsule width (at 50Â) were measured with ments of 74 first instars, 113 second instars, and 520 third instars. an ocular micrometer installed in the right-hand lens of a Leica Wilk’s lambda was used to infer multivariate significance with nor- MZ16 dissecting microscope, or with a digital micrometer and Lei- mal data distribution assumptions. The mean number of days from ca Application Suite software (Version 2.7.1) (Leica Microsystems, female parent wasp removal to collection of pupae from shoots New York, NY). Four to eighty larvae were measured per day of was used to infer the larval–pupal transition time. The number of development (average of 30 ± 4 larvae per day), and a total of days post-oviposition at which individual pupae were collected 707 larvae were measured. Two to sixty-four pupae were mea- as well as generation time (oviposition to new adult), were com- sured per day, beginning on the 17th day after oviposition (average pared between French and Spanish wasp populations with ANOVA. of 19 ± 3 pupae per day) and a total of 286 pupae were measured. Adult longevity was determined from tests with individual On separate shoots that were left undissected, the time from re- wasps in cages by counting the number of days between wasp lease of the parent wasp onto the shoot to emergence of new adults emergence and death, and was compared between French and was recorded to determine total generation time. The length of 34 Spanish wasps using ANOVA. Wasp survival over time was com- adult females with and without antennae were measured with the pared between populations with SAS PROC LIFETEST using an esti- digital micrometer, as were five eggs proximal to the oviducts that mated log-rank v2 test. The percentage of wasps that produced were dissected from each of 10 newly emerged adult females. The offspring was compared between French and Spanish wasps and total number of eggs in the ovaries was counted. among wasp ages with maximum-likelihood v2 tests (SAS PROC FREQ). The reproductive output of individual wasps was compared 2.4. Adult female longevity, reproductive output, and gall position between countries of origin using ANOVA. The intrinsic reproduc-
tive rate r was determined by the equation r =lnR0/T (Begon et al., Newly emerged adult female T. romana wasps were caged indi- 1990), where R0 is average reproduction per female (including all vidually as for wasp development tests and were transferred to females that failed to produce) and T is generation time. Population new A. donax shoots daily. Adult males (recognizable by their long, doubling time (Td) was determined by the equation setose antennae) were rarely collected from quarantine colonies; Td = T Â (log(2)/log(q2/q1)) where q1 is size of the parent wasp co- when available they were paired with individual females in cages. hort and q2 is total number of wasps produced. This equation as- A total of 229 Perpignan (French) females and 69 Granada (Span- sumes a constant net population growth rate. The distributions ish) females were caged. The longevity (in days) of each female of wasp exit holes on the stem, relative to the location of the shoot wasp was noted. The shoot tip of each A. donax plant exposed to tip at the time of female parent removal, were compared between a wasp was marked with a felt tip pen immediately after female re- French and Spanish wasps using a Kolmogorov–Smirnov test (SAS moval. Shoots were maintained in a growth chamber at 27 °C and PROC NPAR1WAY). were re-caged after 25 days. New adults were collected from galled Wasp behavior was compared between 23 French wasps that shoots daily. About 45–50 days after parent wasp removal, the were reproductively successful at least once in their lives and 21 shoots were dissected and larvae, pupae, and non-emerged live non-productive wasps. Most (92%) observations involved wasps adults were counted. Exit holes in the stem were counted, and their that were 3 days old or less. Differences in counts of probing events P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179 173 and in proportional ‘time budgets’ were assessed using PROC 5.25 ± 0.05 mm long (range 2.19–6.70 mm, 286 pupae measured) GLIMMIX in SAS with a Poisson (probe count) or binomial (time and were initially white but became black shortly before adult budget) data distribution specification, and a random residual term eclosion. Meconia containing larval cuticle and other wastes were (time budget data only). A similar approach was used to compare found inside gall cells containing pupae. lifetime wasp behavior between morning and afternoon observa- tion periods. A total of 37 wasps were observed in both periods, 3.1.2. Adults with seven additional wasps observed only in the afternoon. To The adult female of T. romana (Fig. 1A) was diagnosed as fol- present both the reproductively successful/non-productive and lows: The wasp has a black body, antennal flagella and legs and morning/afternoon comparisons (Table 1), the number of probes pale yellow wing venation. The ventral margin of the lower face, and minutes spent in the five behaviors were summed across all lateral panel of the pronotum, apex of the valvulae, apical portion wasps observed in each category, and time measurements were of the coxae, trochanters, knees, apices of the tibiae and tarsi are then divided by the total observation time for all wasps in that yellow-testaceous. The head is globose in dorsal view, with long category. temples. The distance between the lateral ocelli is small in compar- ison to other Palearctic members of this genus. The flagellum is fili- 3. Results form with elongate segments, quite longer than broad. The funicle is five-segmented. The head and pro- and mesonotum have long 3.1. Characteristics of T. romana suberect pilosity. The pronotal collar and midlobe of the mesoscu- tum are coarsely rugose-punctured. The pronotal collar is some- 3.1.1. Immature stages what depressed anteriorly in the middle. The pronotum is long Adult T. romana females (Fig. 1A) deposited eggs into the stem and shoulder-like anteriorly. The notauli, located on the mesoscu- wall or cavity by curling their abdomen (Fig. 1B) and inserting eggs tum, are deep, wide, and crenulate. The metasoma (gaster) is sub- with their needle-like ovipositor, leading to the formation of gall sessile, elongated and has an acute apex. In colonies, males were tissue and shoot distension (Fig. 1C) within 2 weeks. The eggs of readily distinguished from females by their approximately 2-fold T. romana were creamy-white, with a hook-like pedicel on one longer setose antenna. Adult female wasps averaged end (Fig. 1D) and a short hair on the micropylar end. The eggs 5.38 ± 0.55 mm long without antennae and 6.77 ± 0.66 mm long without the pedicel averaged (±SE) 0.39 ± 0.07 mm long, and the with antennae (range 5.14–7.73 mm, 34 Spanish wasps measured). total egg length with pedicel was 1.26 ± 0.08 mm long (range The female:male sex ratio for wasps produced over several gener- 1.05–1.39 mm, 47 eggs measured). The larvae hatched 4–5 days ations from a colony of French wasps was 9:1, while for Spanish after oviposition, and gall tissue was already forming around the colony progeny the sex ratio was 26:1. Newly emerged and 1- eggs and larvae at that time. Gall tissue was found at the shoot api- day-old females, never exposed to A. donax, contained an average cal meristem, a white band of tissue located directly below the leaf (±SE) of 40 ± 3.1 eggs (range 1–58 eggs, 20 Spanish females primordia, or within 3 cm below the meristem. Young gall tissue dissected). was white, while older material (observed more than 2 weeks after oviposition) was yellow. Gall tissue filled the stem cavity, causing 3.2. Immature development exterior shoot distension within 2 weeks of oviposition (Fig 1C). However, on main shoots greater than ca. 1 cm in diameter, gall The development study combined data for larvae reared from formation often occurred with no visible shoot distension. The lar- French and Spanish wasps. The number of larval instars was in- vae (Fig. 1E) were white-yellow and partially translucent, averag- ferred by inflections (declines or plateaus) in daily mean head cap- ing 3.65 ± 0.06 mm long (range 0.4–8.6 mm, 707 larvae sule width, and the pattern of increase of larval length (Fig. 2A) measured) and had head capsules averaging 0.42 ± 0.004 mm in paralleled that of head capsule width (Fig. 2B). The first instar width (range 0.10–0.64 mm) with poorly developed bidentate lasted 5 days (days 4–8 inclusive post-oviposition), the second in- mandibles. The larvae were apodous and moved by undulations star 5 days (days 9–13), and the third instar 12 days (days 14–25). of the body wall. In dissections, the gut appeared blind, with no Larvae were classified into instars on the basis of these time ranges, connection between the hindgut and the presumptive anal area and length and head capsule width measurements differed signif- (A. Cohen, personal communication). As the larvae matured they icantly across the three instars in a multivariate ANOVA chewed individual ‘cells’ inside the gall, and each gall contained (W = 0.376, F = 221.7, df = 4, 1406, P < 0.0001) (Fig. 3A). The mea- an average of 13.8 ± 0.7 cells (range 1–42 cells, 198 galled A. donax surements for some of the larvae classified on the basis of time shoots examined). The gall tissue became dry and woody as the as third instars grouped closely with the measurements on first larvae reached the pupal stage. The pupae (Fig. 1F) averaged and second instars (Fig. 3A). In univariate analyses, mean larval
Table 1 Comparison of the number of probes and behavioral time budgets (% of time) of T. romana between wasps that were reproductively successful at least once in their lifetime and non-reproductive wasps, and between morning and afternoon observations.a
Categoryb Wasps observed Total time (h) Probes Probing (%) On stem (%) On leaf (%) On cage (%) On cotton (%) Reproductive success Successful 23 267.3 825a 16.5a 13.5a 28.0a 39.4a 2.6a Unsuccessful 21 167.2 306b 9.9a 6.1a 24.0a 55.3a 4.7a Time of day Morning 37 177.0 215b 6.4b 5.5b 39.1a 47.3a 1.6a Afternoon 44 266.5 916a 19.1a 14.2a 18.1b 44.1a 4.5a
a For presentation, the number of probes and time spent in each behavior were summed across all wasps in each category and percentage of time allocated to each behavior was determined by dividing by the total time of observation. Behavioral comparisons based on reproductive success involved two separate groups of wasps that were either reproductively successful at least once in their lifetime or non-reproductive. Comparisons of morning and afternoon behaviors involved the same wasps observed during both time periods, with seven wasps observed in the afternoon only. b Summed values under each comparison with different letters within each column represent significant differences in analyses of variance based on results for individual wasps (P 6 0.003). 174 P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179
Fig. 2. Daily averages (±SD) of Tetramesa romana (A) larval and pupal length and (B) larval head capsule width, in galls dissected 4–33 days after oviposition. Brackets indicate inferred time durations of instars. Fig. 3. (A) Association between larval length and head capsule width of T. romana, length varied significantly across instars (F = 282.2, df = 2, 704, with observations grouped into three instars on the basis of larval age at the time of P < 0.0001) (Fig. 3B), as did mean head capsule width (F = 534.4, dissection. (B) Average (±SE) length and head capsule width compared between df = 2, 704, P < 0.0001) (Fig. 3B). The average time to pupation instars. Means with different letters for each measurement are significantly different in analyses of variance (P < 0.0001). was 26.0 ± 0.2 days (median 27 days, 286 pupae observed), although pupae were observed as early as the 17th day after ovipo- sition (Fig. 2A). Spanish larvae reached pupation 23.6 ± 3.6 days 3 days, maximum of 10 days) (260 wasps observed). Wasps reared after oviposition (median 23 days), 4 days faster than French larva, from Spanish populations lived 1.3 days longer (4.7 ± 0.3 days, which required 27.3 ± 2.3 days (median 27 days) (F = 107.2, df =1, median 4 days, 69 wasps observed) than French wasps (3.4 ± 0.1 284, P < 0.0001). The pupal development time for wasps from both days, median 3 days, 191 wasps observed) (F = 19.9, df = 1, 259, countries was estimated to be 7 days, on the basis of a total larval P < 0.0001). French wasps had only a 13% probability of being alive development time of 26 days and a total generation time (see be- 6 days after emergence, while Spanish wasps had a 32% chance of low) of 33 days. surviving 6 days. The survival function for Spanish wasps was sig- nificantly different from that of French wasps (log-rank test, 3.3. Generation time v2 = 16.4, df =1,P < 0.0001) (Fig. 4).
The average time from the day of parent wasp oviposition to 3.4.2. Frequency of reproductive success emergence of adults from galls was 33.3 ± 0.1 days (2133 wasps Across all 298 T. romana wasps isolated individually on A. donax emerged, median time 33 days). However, generation time varied beginning on the day of emergence, 43% produced offspring at least between a minimum of 25 and a maximum of 50 days. Spanish once. A greater proportion of Spanish wasps (65%, n = 69) were wasps had a marginally longer generation time (33.9 ± 0.1 days, reproductive than were French wasps (35.8%, n = 229) (v2 = 18.7, median 33 days; 878 wasps reared from 71 plants) than did French df =1, P < 0.0001). Twenty-four of the 82 reproductive French wasps (32.8 ± 0.1 days, median 32 days; 1255 wasps reared from wasps produced offspring on more than one shoot, 10.5% of all 90 plants) (F = 8.28, df = 1, 159, P = 0.005). French wasps tested. Out of 45 reproductive Spanish wasps, 25 produced offspring on more than one shoot, 36.2% of all Spanish 3.4. Adult longevity, reproduction, and gall position wasps tested. On separate A. donax shoots, two to four French or Spanish females were isolated inside each cage. The reproductive 3.4.1. Longevity success rate differed significantly between 41 shoots caged with Adult wasps confined in cylinder cages on A. donax shoots and two wasps, of which 34% developed galls, 148 shoots caged with transferred daily lived an average (±SE) of 3.7 ± 0.1 days (median three wasps, of which 65.5% developed galls, and 90 shoots caged P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179 175
French wasps (n = 82) (18.7 ± 1.2, maximum = 55) (F = 9.13, df =1, 125, P = 0.003) and the same was true when only adult offspring were considered (Spanish wasps, 23.4 ± 2.4; French wasps, 16.6 ± 1.2; F = 7.67, df = 1, 125, P = 0.007). The number of exit holes did not correspond exactly to the number of adult offspring col- lected (Spanish wasps, 20.2 ± 2.1 holes; French wasps, 15.0 ± 1.2 holes), suggesting that more than one wasp used the same exit hole in a few cases. Considering both reproductively successful and unsuccessful wasps, lifetime output was 6.7 ± 0.7 offspring per French wasp and 17.4 ± 2.2 offspring per Spanish wasp. The intrinsic rate of increase r for the entire cohort of 229 French wasps was 0.054 and the population doubling time (DT) was 13.4 days. For the cohort of 69 Spanish wasps, r = 0.081 and the doubling time was 9.2 days. Eighty percent of the offspring of French and Spanish wasps combined were produced within the first 4 days, beginning on the day of emergence (Fig. 5), and the number of offspring per female (both successful and non-productive wasps included) var- ied significantly according to wasp age (F = 3.81, df = 8, 711, Fig. 4. Survival of newly emerged French and Spanish adult female T. romana caged P = 0.0002) (Fig. 5). However, three Spanish wasps produced a total on A. donax. of 50 offspring when they were 9 days old. with four wasps, of which 48.9% developed galls (v2 = 15.3, df =2, 3.4.4. Position of galls on shoots P = 0.0005). Wasp age also influenced reproduction (v2 = 46.6, Across 186 A. donax shoots that had been exposed to French df = 10, P < 0.0001). Of 573 A. donax shoots exposed to French or (104 shoots) or Spanish (82 shoots) wasps and were then dissected Spanish wasps that were 4 days old or less, 182 (31.8%) developed after progeny emerged, 82% of the 2075 exit holes indicative of gall galls and yielded wasp progeny, while only 16 of 153 shoots formation were found at the apical node marked at the time of par- (10.5%) exposed to wasps that were 5 days old or older yielded ent wasp removal, or within two nodes above the tip. The distribu- progeny. The total offspring produced by the colonies declined by tion of exit holes across nodes differed between French and 40% or more each day beginning with 3-day-old wasps (Fig. 5). Spanish wasps (Kolmogorov–Smirnov test, adjusted KS = 4.27, P < 0.0001) (Fig. 6). Spanish T. romana wasp offspring produced exit 3.4.3. Reproduction by individual wasps holes 19% more frequently at the shoot position located one or The mean (±SE) lifetime reproductive output of reproductively more new nodes above the apical node at oviposition than did successful wasps was 21.2 ± 1.2 offspring, including both adult French wasps (Fig. 6). wasps and larvae/pupae still in the stem at the time of dissection. A typical French wasp was exposed to three shoots in its lifetime 3.5. Oviposition behavior and a typical Spanish wasp to four shoots, beginning on the day of emergence. The average reproductive output per shoot in a total Abdomen-curling behavior, suggestive of ovipositor insertion of 182 tests involving successful female wasps from both countries into the stem (Fig. 1B), was frequently observed in T. romana fe- exposed to new shoots daily within the first 4 days of their life- males confined on A. donax stems in cages. Wasps spent the times was 14.1 ± 0.7 offspring (maximum = 42). Reproductively remainder of their time resting on the stem, leaf blade, cage mate- successful Spanish wasps (n = 45) produced more adult and larval rial, or cotton plug containing honey, with short (1–2 s) bouts of offspring over their lifetimes (26.0 ± 2.4, maximum = 66) than did
Fig. 5. Average (±SE) and total reproduction of larvae and adults by French and Spanish T. romana wasps exposed to new A. donax shoots daily beginning on the day Fig. 6. Distribution of exit holes made by French and Spanish wasps at nodes below of emergence (day 1) until death. Numbers above bars indicate the number of (negative) or above (positive) the shoot tip of A. donax plants at the time of wasps included in calculations for each day. oviposition. 176 P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179
flight. Wasps that reproduced at least once in their lifetime made pest Tetramesa in North America (Davis, 1918; USDA, 1954; Salt, 2.5-fold more probes than did non-productive wasps (F = 181.1, 1971), and the third instar may serve the same role in T. romana df = 1, 42, P < 0.0001), but did not differ from non-reproductive in temperate regions such as southern France. wasps in proportion of time budgeted to probing or other behav- iors (P > 0.38) (Table 1). The mean length of individual probing 4.3. Generation time of T. romana events did not differ on the basis of reproductive success (success- ful wasps, 3.3 ± 0.4 min, non-productive wasps, 3.3 ± 0.2 min, The 33-day generation time of T. romana wasps maintained at F = 0.04, df = 1, 1129, P = 0.834). Wasps observed in the afternoon 27 °C suggests that the number of generations per year in the field made 4.3-fold more probes (F = 283.6, df = 1, 79, P < 0.0001) and after release are likely to be more than the one to two generations spent 13% more time probing (F = 17.2, df = 1, 79, P < 0.0001) and per year reported for other Tetramesa spp. in Britain (Claridge, 9% more time resting on the stem (F = 14.0, df = 1, 79, P = 0.0003), 1961; Dawah et al., 1995) and temperate North America (Davis, than did the same wasps in the morning (Table 1). Wasps spent 1918; Knowlton and Janes, 1933; USDA, 1954; Shanower and 21% less time resting on leaf blades in the afternoon (F = 9.35, Waters, 2006). In the LRGB, average daily temperatures are likely df = 1, 19, P = 0.003) while time spent resting on the cage or on a to be near or above 27 °C for at least 4–5 months of the year, sug- honey plug did not differ between observation periods (P > 0.15). gesting that T. romana will complete at least five generations per year under field conditions. The variable generation time in lab- 4. Discussion reared wasps (ranging from 17 and 50 days after oviposition) likely reflected variable development rates of third-instar larvae. In the 4.1. Characteristics of T. romana field, variation in development times may buffer T. romana against environmental extremes. Galled A. donax shoots hosting an adven- Our diagnosis of the adult stage of T. romana is consistent with tive population of T. romana in Austin, Texas cut under winter con- but builds upon past descriptions of this species (Steffan, 1956; ditions (average daily temperature 5–15 °C) and maintained at Claridge, 1961; Zerova, 1976). Little prior information was avail- 25 °C yielded adult wasps up to 2 months after cutting (J. Goolsby able on immature stages of T. romana. The egg pedicel, a structure and P. Moran, unpublished data), suggesting that reduced develop- 2- to 3-fold longer than the egg itself (Holmes and Blakeley, 1971), ment rates of the final instar will characterize the life cycle of T. could serve as a respiratory conduit between the stem surface and romana in the mild winter climate of the LRGB. egg during egg development, as in cynipid galling wasps (Askew, 1984) or may facilitate oviposition (Vårdal et al., 2003). The larvae 4.4. Adult longevity of T. romana of Tetramesa spp. have few features to separate species (Roskam, 1982), but high host specificity is expressed in larval feeding (Cla- The survival duration of French and Spanish T. romana reared ridge, 1961; Dawah, 1987). The large size of the adult female of T. under growth chamber and greenhouse conditions (25–30 °C) is romana (well over 5 mm) relative to other Palearctic Tetramesa spp. short relative to those reported from field studies on other Tetra- was noted by Steffan (1956). The female-biased sex ratios and fully mesa species (USDA, 1954; Claridge, 1961). However, survival developed eggs we observed in newly emerged French and Spanish times for these species were based on observations of populations, T. romana agree with past findings (Phillips, 1936; Claridge, 1961) rather than individuals. Daily laboratory handling may have short- that reproduction by T. romana occurs via thelytokous partheno- ened adult female wasp survival relative to what could be expected genesis (Askew, 1984), with diploid females developing from in the field. Resource provisioning may have also influenced sur- unfertilized eggs. The addition of occasional T. romana males to vival. The dietary requirements of adult Tetramesa are unknown cages containing females did not alter female-biased sex ratios in (Claridge, 1961), but could involve floral nectar as a carbohydrate progeny, suggesting that males are sterile. source. Honey-water was provided in this study. For mass-rearing purposes, additional resources such as yeast lysate may extend 4.2. Development of T. romana adult lifetimes (Cohen, 2004) and thereby increase reproduction, as well-developed eggs were found inside female T. romana that The identification of three larval instars in T. romana is the sim- had died within 4 days of emergence. The longer survival time of plest explanation for the observed patterns of increase in larval Spanish as compared to French wasps suggests that Spanish popu- head capsule width and body length. No prior information of com- lations will be more suitable for mass-rearing. parable detail is available to validate our conclusions about the number and duration of the instars and of the pupal stage (Phillips, 4.5. Reproduction of T. romana 1936; Claridge, 1961; Al-Barrak et al., 2004). The wasp Eurytoma tumoris Bugbee galling Scots pine (Pinus sylvestris L.) was reported 4.5.1. Frequency of reproduction to have four instars (Stark and Koehler, 1964), and larvae of a Eury- Despite the fact that reproduction was parthenogenetic, fe- toma sp. leafminer evaluated for control of mother-of-millions males varied in their ability to reproduce. Almost twice as many weed (Bryophyllum delagoense (Ecklon and Zeyher)) completed five Spanish females reproduced as French females, again suggesting instars (Witt et al., 2004). Some T. romana larvae classified as third that the Spanish population is better-suited to large-scale rearing. instars, based solely on the number of days between oviposition The cause of reproductive failure is uncertain. Almost all newly and dissection, may have in fact been second instars developing emerged females contained eggs, though some eggs may have con- more slowly than normal, due perhaps to inferior tissue quality tained a non-viable zygote. Variation in stem diameter, which can or quantity at the edges of galls. The more than 2-fold longer dura- influence oviposition (Dubbert et al., 1998) may have affected suc- tion of the third as compared to the first and second instar of T. cess, if non-reproductive wasps, which made fewer probes than romana may facilitate developmental plasticity in responses to reproductive wasps, were confined on thinner stems. Alternatively, environmental cues. Under field conditions, final instar larvae non-reproductive females may have deposited eggs into stem loca- may develop rapidly during the long warm season of Mediterra- tions that were unsuitable for gall formation. The fact that galling nean Europe and the LRGB, where summer daily temperatures rates differed but were not consistently lower on shoots exposed to average 25–35 °C, and where winter temperatures may be high en- groups of 2–4 Spanish wasps compared to a single wasp suggests ough to permit year-round wasp activity. In contrast, the penulti- that inundative releases will not lead to competitive interference. mate instar or the pupa is the overwintering stage in temperate About one-third of the Spanish wasps demonstrated an ability to P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179 177 induce gall formation on more than one shoot during their short 5. Conclusions lifetimes, an important attribute since both primary and axillary shoot production by giant reed is profuse. The declines in galling The rapid immature development, short generation time and ability in long-lived wasps (more than 5-days old) could reflect prolific asexual reproductive capacity of the stem-galling wasp T. egg depletion, insufficient egg provisioning due to poor adult nutri- romana support its candidacy for field release as a biological con- tion, or age-related reduced muscular capacity to penetrate the trol agent against giant reed (McClay and Balciunas, 2005; van shoots of A. donax. Klinken and Raghu, 2006), in combination with the wasp’s high de- gree of host specificity (Goolsby and Moran, 2009), its abundance 4.5.2. Reproductive output and wide distribution in its native range (Kirk et al., 2003) and A single reproductive T. romana wasp produced 20–30 offspring its ability to increase shoot mortality in quarantine tests (J.A. Goo- parthenogenetically, showing the potential to increase population lsby, unpublished data). Establishment and efficacy predictions for size 20-fold or more, even if a female galls only one shoot in its life- weed biological control agents require knowledge of plant demog- time. Even taking unsuccessful wasps into account, the intrinsic raphy (Müller-Schärer and Schaffner, 2008). The life history traits rate of increase and doubling time of the Spanish T. romana popu- of T. romana, as revealed here on young shoot tips, in combination lation are comparable to those of the saltcedar leaf beetle (D. elong- with the perennial vegetative development of giant reed popula- ata deserticola)(Lewis et al., 2003) which has been released tions in California (Spencer and Ksander, 2006) and the LRGB, sug- successfully in the arid western US to control saltcedar (Tamarix gest that conditions for wasp establishment will be favorable. spp.; Tamaricaceae), exotic invasive trees which often co-occur Plasticity in the life cycle of T. romana, mediated in part by variable with A. donax in the LRGB. The reproductive output of T. romana immature development time, may allow the wasp to establish un- is lower than the fecundity, expressed as production of 70 eggs der variable temperature regimes, as shown by the existence of per female, reported for wheat jointworm (T. tritici)(USDA, adventive T. romana populations in both southern California (Dud- 1954), but production of adults is a more reliable measure since ley et al., 2006) and in Austin and Laredo, Texas (J.A. Goolsby, it is likely that not all eggs survive. When provided a new A. donax unpublished data). The longer adult longevity and greater repro- shoot daily, the average Spanish wasp produced 40% more adults ductive output of wasps from Spain as compared to France sug- than did the average French wasp, in contrast to host range tests gests that Spanish wasps are more suitable for the mass-rearing (Goolsby and Moran, 2009), in which wasps were confined on that will be necessary for field releases in the Lower Rio Grande Ba- one shoot from emergence until death. sin of the US and Mexico.
4.5.3. Position of galls on shoots Acknowledgments Tetramesa romana females, especially those from Spain, strongly preferred shoot tips for oviposition, suggesting that these tissues We thank Alan and Gerhild Kirk for conducting surveys in over are uniquely susceptible to gall induction. Other Tetramesa species 15 countries for T. romana and for collection of the wasp from show a similar preference, including T. petiolata (Al-Barrak, 2006) France and Spain. We thank Crystal Salinas, Ann Vacek and Sandra on Deschampia cespitosa (L.) P. Beauv., as well as the wheat joint- Espinoza for technical assistance, and Joe Balciunas, Michael Gates, worm T. tritici (USDA, 1954) and the rye jointworm T. secale and two anonymous reviewers for critical reviews. This work was (Holmes and Blakeley, 1971). Attack on shoot meristematic tissues supported in part by a grant from the US Department of Homeland could enhance the biological control impact of T. romana (Tracy Security, Science and Technology Directorate, and by the Lower Rio and DeLoach, 1999), as in the case of a cecidomyiid gall-making Grande Valley Development Council. fly (Rhopalomyia n. sp.) targeting scentless chamomile (Tripleuro- spernum perforatum (Merat) Lainz) (Hinz, 1998). References 4.6. Oviposition behavior Al-Barrak, A., 2006. Host choice in Tetramesa petiolata (Walker) (Hymenoptera: Eurytomidae). Journal of Entomology 3, 55–60. Reproductive wasps made more probes of A. donax stems than Al-Barrak, M., Loxdale, H.D., Brookes, C.P., Dawah, H.A., Biron, D.G., Alsagair, O., non-reproductive wasps, but did not differ in overall time alloca- 2004. Molecular evidence using enzyme and RAPD markers for sympatric evolution in British species of Tetramesa (Hymenoptera: Eurytomidae). tion to probing or in length of probing events. Reproductive wasps Biological Journal of the Linnean Society 83, 509–525. may have been more sensitive to plant-derived physical or chem- Askew, R.R., 1984. The biology of gall wasps. In: Ananthakrishnan, T.N. (Ed.), The ical cues that stimulated at least some of the elements of probing. Biology of Gall Insects. Oxford and IBH Press, New Delhi, India, pp. 223–271. Awmack, C.S., Leather, S.R., 2002. Host plant quality and fecundity in herbivorous Some probes may have served to sample host tissues rather than to insects. Annual Review of Entomology 47, 817–844. deposit eggs. The mean length of probing events and proportion of Bean, D.W., Dudley, T.L., Keller, J.C., 2007. Seasonal timing of diapause induction time budgeted to probing did not differ on the basis of reproduc- limits the effective range of Diorhabda elongata deserticola (Coleoptera: Chrysomelidae) as a biological control agent for tamarisk (Tamarix spp.). tive success, suggesting that the extra probes made by successful Environmental Entomology 36, 15–25. wasps were not costly in terms of time. Variation in probing behav- Begon, M., Harper, J.L., Townsend, C.R., 1990. Ecology, Individuals, Populations, and ior tied to reproductive success could reflect variation in reproduc- Communities, second ed. Blackwell, Boston, MA. Bell, G., 1997. Ecology and management of Arundo donax, and approaches to tive structures or host perception, perhaps tied to the quality/ riparian habitat restoration in southern California. In: Brock, J.H., Wade, M., quantity of gall tissues available to the adults when they were lar- Pysek, P., Green, D. (Eds.), Plant Invasions: Studies from North America and vae (Awmack and Leather, 2002). Relatively large females reared Europe. Blackhuys Publishers, Ledien, The Netherlands, pp. 103–113. on large galls had the highest fecundity in the cynipid wasp Aphel- Boland, J.M., 2006. The importance of layering in the rapid spread of Arundo donax (giant reed). Madroño 53, 303–312. onyx grandulifera (Monzen) (Ito and Hijii, 2004), and T. romana fe- Briese, D.T., 2004. Weed biological control: applying science to solve seemingly males varied in size by about 35% around the mean value. Adult intractable problems. Australian Journal of Entomology 43, 304–317. nutrition was not likely involved, as reproductive and non-repro- Cantero-Martínez, C., Angas, P., Lampurlanés, J., 2003. Growth, yield and water productivity of barley (Hordeum vulgare L.) affected by tillage and N fertilization ductive females did not differ in time spent on honey-provisioned in Mediterranean semiarid, rainfed conditions of Spain. Field Crops Research 84, cotton. Wasps spent more time in probing and stem-resting behav- 341–357. iors in afternoon than in morning observations, perhaps as a sen- Chapman, R.F., 1982. The Insects: Structure and Function, third ed. Harvard University Press, Cambridge, MA. sory response to variable ambient light, temperature, or moisture Claridge, M.F., 1961. A contribution to the biology and taxonomy of some Palearctic levels. species of Tetramesa Walker (=Isosoma Walk.; =Harmolita Motsch.) 178 P.J. Moran, J.A. Goolsby / Biological Control 49 (2009) 169–179
(Hymenoptera: Eurytomidae), with particular reference to the British fauna. in the USA. In: Porosko, C. (Ed.), Proceedings of the California Invasive Plant Transactions of the Royal Entomological Society of London 113, 175–216. Council Symposium, 2–4 October, 2003, Kings Beach, California, USA. California Cohen, A.C., 2004. Insect Diets: Science and Technology. CRC Press, Boca Raton, FL. Invasive Plant Council, Berkeley, CA, pp. 62–68. Coulson, J.R., Vail, P.V., Dix, P.E., Nordlund, D.A., Kauffman, W.C., 2000. 110 Years of Kleinjan, C.A., Edwards, P.B., 2006. Asparagus weeds in Australia—a South African Biological Control Research and Development in the United States Department perspective with emphasis on biological control prospects. Plant Protection of Agriculture, 1883–1993. US Department of Agriculture, Agricultural Research Quarterly 21, 63–68. Service, Beltsville, MD. Knowlton, G.F., Janes, M.J., 1933. Distribution and Damage by Jointworm Flies in Davis, J.J., 1918. The control of three important wheat pests in Indiana. Circular No. Utah. Bulletin No. 243, Utah Agricultural Experiment Station, Logan, UT. 82, Purdue University Agricultural Experiment Station, Lafayette, IN. Lewis, P.A., DeLoach, C.J., Knutson, A.E., Tracy, J.L., Robbins, T.O., 2003. Biology of Dawah, H.A., 1987. Biological species problems in some Tetramesa (Hymenoptera, Diorhabda elongata deserticola (Coleoptera: Chrysomelidae), an Asian leaf beetle Eurytomidae). Biological Journal of the Linnean Society 32, 237–245. for biological control of saltcedars (Tamarix spp.) in the United States. Biological Dawah, H.A., Hawkins, B.A., Claridge, M.F., 1995. Structure of the parasitoid Control 27, 101–116. communities of grass-feeding chalcid wasps. Journal of Animal Ecology 64, Lotfalizadeh, H., Delvare, G., Rasplus, J.-Y., 2007. Phylogenetic analysis of 708–720. Eurytominae (Chalcidoidea: Eurytomidae) based on morphological characters. Dawah, H.A., Al-Haddad, F.H., Jervis, M.A., 2002. Morphological and biological Zoological Journal of the Linnean Society 151, 441–510. characterization of three closely related species of Pediobius Walker Martin, T.J., Harvey, T.L., 1982. Cultivar response to wheat strawworm. Crop Science (Hymenoptera: Eulophidae). Journal of Natural History 36, 423–433. 22, 1233–1235. Decruyenaere, J.G., Holt, J.S., 2001. Seasonality of clonal propagation in giant reed. Matsumoto, R., Saigusa, T., 2001. The biology and immature stages of Thrybius Weed Science 49, 760–767. togashii Kusigemati (Hymenoptera: Ichneumonidae: Cryptinae), with a DiTomaso, J.M., Healy, E.A., 2003. Aquatic and Riparian Weeds of the West. ANR Pub. description of the male. Journal of Natural History 35, 1507–1516. 3241. University of California Press, Berkeley, CA. McClay, A.S., Balciunas, J.K., 2005. The role of pre-release efficacy assessment in Dubbert, M., Tscharntke, T., Vidal, S., 1998. Stem-boring insects of fragmented selecting classical biological control agents for weeds—applying the Anna Calamagrostis habitats: herbivore-parasitoid community structure and the Karenina principle. Biological Control 35, 197–207. unpredictability of grass shoot abundance. Ecological Entomology 23, 271–280. McFadyen, R.E.C., 1998. Biological control of weeds. Annual Review of Entomology Dudley, T.L., 2000. Arundo donax. In: Bossard, C.C., Randal, J.M., Hosovsky, M.C. 43, 369–393. (Eds.), Invasive Plants of California Wetlands. University of California Press, McGaugh, S., Hendrickson, D., Bell, G., Cabral, H., Lyons, K., McEachron, L., Munoz, O., Berkeley, CA, pp. 53–58. 2006. Fighting an aggressive wetlands invader: a case study of giant reed, Dudley, T.L., Lambert, A., Kirk, A., 2006. Augmentation biological control of Arundo (Arundo donax) and its threat to Cuatro Ciénegas, Coahuila, Mexico. In: Lozano- donax. In: Hoddle, M.S., Johnson, M.W., (Eds.), Proceedings of the Fifth California Vilano, L., Contreas-Balderas, A.J. (Eds.), Studies of North American Desert Fishes Conference on Biological Control, July 25–27, 2006, Riverside, California, USA. in honor of E.P. (Phil) Pister, Conservationist. Universidad Autónoma de Nuevo University of California-Riverside, Riverside, CA, pp. 141–144. León, Facultad de Ciencas Biológicas, Monterrey, Mexico, pp. 100–115 (in Dunmire, W.W., 2004. Gardens of New Spain: How Mediterranean Plants and Foods Spanish). Changed America. University of Texas Press, Austin, TX. Müller-Schärer, H., Schaffner, U., 2008. Classical biological control: exploiting Evans, H.C., 1991. Biological control of tropical grassy weeds. In: Barker, F.W.G., enemy escape to manage plant invasions. Biological Invasions 10, 859–874. Terry, P.J. (Eds.), Tropical Grassy Weeds. CAB International, Wallingford, UK, pp. Myers, J.H., 1987. Population outbreaks of introduced insects: lessons from the 52–72. biological control of weeds. In: Barbosa, P., Schultz, J.C. (Eds.), Insect Outbreaks. Everitt, J.H., Yang, C., DeLoach, C.J., 2005. Remote sensing of giant reed with Academic Press, New York, NY, pp. 173–193. QuickBird satellite imagery. Journal of Aquatic Plant Management 43, 81–84. Nechols, J.R., Andres, L.A., Beardsley, J.W., Goeden, R.D., Jackson, C.G. (Eds.), 1995. Frandsen, P., Jackson, N., 1994. The impact of Arundo donax on flood control and Biological Control in the Western United States: Accomplishments and Benefits endangered species. In: Jackson, N.E., Frandsen, P., Duthoit, S. (Compilers), of Regional Research Project W-84, 1964–1989. Publication No. 3361, Arundo donax Workshop Proceedings, 19 November 1993, Ontario, California, University of California, Division of Agriculture and Natural Resources, USA. Team Arundo and California Exotic Pest Plant Council, Pismo Beach, CA, pp. Oakland, CA. 13–16. Neser, O.C., 2008. Eurytoma bryophylli sp. n., (Hymenoptera: Eurytomidae), a leaf Goolsby, J.A., Moran, P.J., 2009. Host range of Tetramesa romana Walker borer of Bryophyllum delagoense (Crassulaceae) from Madagascar and a (Hymenoptera: Eurytomidae), a potential biological control of giant reed, candidate for the biocontrol of the plant in Australia. African Entomology 16, Arundo donax L. in North America. Biological Control 49, 160–168. 60–67. Goolsby, J.A., van Klinken, R.D., Palmer, W.A., 2006. Maximizing the contribution of Noyes, J., 2004. Universal Chalcidoidea Database. Natural History Museum, London, native range studies towards the identification and prioritization of weed UK. Available from:
Shanower, T.G., Waters, D.K., 2006. A survey of five stem-feeding insect pests of Tewksbury, L., Casagrande, R., Blossey, B., Häfliger, P., Schwarzländer, M., 2002. wheat in the Northern Great Plains. Journal of Entomological Science 41, 40–48. Potential for biological control of Phragmites australis in North America. Sheppard, A.W., Raghu, S., 2005. Working at the interface of art and science: how Biological Control 23, 191–212. best to select an agent for classical biological control? Biological Control 34, Thornby, D., Spencer, D., Hanan, J., Sher, A., 2007. L-DONAX, a growth model of the 233–235. invasive weed species, Arundo donax L. Aquatic Botany 87, 275–284. Sheppard, A.W., van Klinken, R.D., Heard, T.A., 2005. Scientific advances in the Tracy, J.L., DeLoach, C.J., 1999. Suitability of classical biological control for giant reed analysis of direct risks of weed biological control agents to nontarget plants. (Arundo donax) in the United States. In: Bell, C.E. (Ed.), Arundo and Saltcedar Biological Control 35, 215–226. Management Workshop Proceedings, 17 June 1998, Ontario, California. Simmonds, F.J., 1980. Biological control of Cordia curassavica (Boraginaceae) in University of California Cooperative Extension, Holtville, CA, pp. 73–109. Malaysia. Entomophaga 25, 363–364. Tscharntke, T., Greiler, H.J., 1995. Insect communities, grasses, and grasslands. Spencer, D.F., Ksander, G.G., 2006. Estimating Arundo donax ramet recruitment Annual Review of Entomology 40, 535–558. using degree day-based equations. Aquatic Botany 85, 282–288. USDA, 1954. The Wheat Jointworm: How to Fight It. Leaflet No. 380, US Department Spencer, D.F., Liow, P.S., Chan, W.K., Ksander, G.G., Getsinger, K.D., 2006. Estimating of Agriculture, Washington, DC. Arundo donax shoot biomass. Aquatic Botany 84, 272–276. van Klinken, R.D., Raghu, S., 2006. A scientific approach to agent selection. Spencer, D.F., Stocker, R.K., Liow, P.-S., Whitehand, L.C., Ksander, G.G., Fox, A.M., Australian Journal of Entomology 45, 253–258. Everitt, J.H., Quinn, L.D., 2008. Comparative growth of giant reed (Arundo donax Vårdal, H., Shalén, G., Ronquist, F., 2003. Morphology and evolution of the cynipoid L.) from Florida, Texas, and California. Journal of Aquatic Plant Management 46, egg (Hymenoptera). Biological Journal of the Linnean Society 139, 247–260. 89–96. Wijte, A.H.B.M., Mizutani, T., Motamed, E.R., Merryfield, M.L., Miller, D.E., Alexander, Stark, R.W., Koehler, C.S., 1964. Biology of the gall wasp, Eurytoma tumoris, on Scots D.E., 2005. Temperature and endogenous factors cause seasonal patterns in pine (Hymenoptera: Eurytomidae). Pan-Pacific Entomologist 40, 41–46. rooting by stem fragments of the invasive giant reed, Arundo donax L. (Poaceae). Steffan, J.R., 1956. Notes sur la biologie d’Harmolita romana (Walk.) (Hym. International Journal of Plant Sciences 166, 507–517. Eurytomidae). Bulletin de la Societé Entomologique de France 61, 34–35. Witt, A.B.R., McConnachie, A.J., Docherty, S., 2004. Distribution and aspects of the Sterling, J.D.E., 1976. Resistance to barley jointworm. Canada Agriculture 21, 13–14. biology and host range of Eurytoma sp., (Hymenoptera: Eurytomidae), a Strong, D.R., Lawton, H., Southwood, R., 1984. Insects on Plants. Blackwell Scientific candidate agent for the biological control of Bryophyllum delagoense (Ecklon Publications, Oxford, UK. and Zeyher) Schinz (Crassulaceae) in Australia. African Entomology 12, 201– Szabo, P., Varhegyi, G., Till, F., Faix, O., 1996. Thermogravimetric mass spectrometric 207. characterization of two energy crops, Arundo donax and Miscanthus sinensis. Zerova, M.D., 1976. Hymenoptera 7. Part 6: Family Eurytomidae; subfamilies Journal of Analytical and Applied Pyrolysis 36, 179–190. Rileyinae and Harmolitinae. Fauna USSR 110, 1–230 (in Russian).