Prediction of the Geographic Distribution of the Psyllid, Arytinnis

Prediction of the Geographic Distribution of the Psyllid, Arytinnis hakani (Homoptera: Psyllidae), a Prospective Biological Control Agent of Genista monspessulana, Based on the Effect of Temperature on Development, Fecundity, and Survival Author(s): Lincoln Smith Source: Environmental Entomology, 43(5):1389-1398. Published By: Entomological Society of America URL: http://www.bioone.org/doi/full/10.1603/EN14086

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. PHYSIOLOGICAL ECOLOGY Prediction of the Geographic Distribution of the Psyllid, Arytinnis hakani (Homoptera: Psyllidae), a Prospective Biological Control Agent of Genista monspessulana, Based on the Effect of Temperature on Development, Fecundity, and Survival

LINCOLN SMITH1

Exotic and Invasive Weeds Research Unit, USDA Agricultural Research Service, 800 Buchanan Street, Albany, CA 94710

Environ. Entomol. 43(5): 1389Ð1398 (2014); DOI: http://dx.doi.org/10.1603/EN14086 ABSTRACT The psyllid, Arytinnis hakani (Loginova), is a prospective biological control agent of Genista monspessulana (French broom), an invasive shrub originating from western Europe. It is a multivoltine species that is not known to diapause. The insect is established in Australia, where it appears to cause heavy defoliation and mortality of the target weed, except at warm sunny sites. This suggests that bright light or high temperatures may hamper the agent. We measured the effect of temperature on development rate, survival, and fecundity of the psyllid to determine its suitable temperature range. Intrinsic rate of increase was highest near 22ЊC, and there was no population growth at the extremes of 5ЊC and 26ЊC. Net reproductive rate was highest at 16.5ЊC. Fecundity was highest at 22ЊC, and decreased to half at 16ЊC and at 27ЊC. Adult female longevity decreased with increasing temperature over the range studied. Nymphal survivorship was highest at 16ЊC and dropped to 0% at 5ЊC and 26ЊC. Eggs were able to complete development in 83 d at 5ЊC, but with only 20% survivorship versus 78Ð95% survivorship at higher temperatures. For populations with a stable age distribution, only 2Ð3% of the population is in the adult stage. Climate modeling using CLIMEX indicated that the geographic distribution of the psyllid is constrained by high temperature stress in Australia. The psyllid is predicted to be suitable in coastal California but not in the Sierra foothills.

KEY WORDS population growth, demography, life history, temperature, climate matching

Several psyllids are known to be important crop pests, ture on life history, can help us predict population and some have been used as biological control agents growth rates and model potential geographic distri- of invasive plants. For example, the Asian citrus psyl- butions (Van Klinken et al. 2003, Chiarelli et al. 2011, lid, Diaphorina citri Kuwayama (Homoptera: Psylli- Munyaneza et al. 2012, Myint et al. 2012). dae), transmits a pathogen that causes huanglongbing The psyllid, Arytinnis hakani (Loginova) is pro- (citrus greening disease; Grafton-Cardwell et al. spective biological control agent of French broom 2013), and the potato psyllid, Bactericera cockerelli (Genista monspessulana (L.) L.A.S. Johnson; Fabaceae: (Sˇulc), transmits a pathogen that causes zebra chip Genisteae) in North America (Sheppard and Thom- disease (Munyaneza 2012). Psyllid weed biological ann 2004, Cook and Smith 2014). A. hakani is native to control agents includeÑAphalara itadori Shinji for the western Mediterranean region, occurring from Japanese knotweed (Fallopia japonica (Houttuyn) Portugal to Italy in Europe, and from Morocco to Ronse Decraene), Arytainilla spartiophila (Fo¨rster) Algeria in North Africa. The psyllid is known to feed for Scotch broom (Cytisus scoparius (L.)), Boreiogly- on only G. monspessulana under natural Þeld condi- caspis melaleucae Moore for paper bark tree tions (Hodkinson and Hollis 1987, Sheppard and (Melaleuca quinquenervia (Cav.) Blake), and Heterop- Thomann 2004). The psyllid is multivoltine and not sylla spinulosa Muddiman, Hodkinson & Hollis for known to have any diapause. Eggs are attached to sensitive plant (Mimosa invisa Martius; Wilson and leaves and young stems by a pedicel. Nymphs progress Garcia 1922, Center et al. 2007, Syrett et al. 2007, Shaw through Þve instars as they feed on phloem from et al. 2009). Understanding the quantitative effects of leaves, stem branch tips, and ßowers. A. hakani was important environmental factors such as tempera- evaluated for introduction to Australia, but was dis- covered to already be established before a release permit was issued (Sheppard and Henry 2012). It now Mention of trade names or commercial products in this publication occurs in several Australian states, and high densities is solely for the purpose of providing speciÞc information and does not imply recommendation or endorsement by the USDA. of the psyllid are killing plants in many locations 1 Corresponding author, e-mail: [email protected]. (Sheppard et al. 2014). However, the psyllid does not 1390 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 5 appear to perform well in hot weather in Australia (Ϸ10 cm in height) in an incubator at constant tem- (A.W. Sheppard, personal communication). In south- perature and a photoperiod of 12:12 (L:D) h, illumi- ern France, the number of psyllids was reported to nated by four 20-watt ßuorescent bulbs (F20T12 day- decrease progressively during summer as tempera- light emitting 76 Ϯ 3 ␮EmϪ2 sϪ1 PAR tures rose, suggesting a decrease in oviposition or [photosynthetically active radiation]). Incubator immature survivorship; however, no data were pre- temperatures were set at 5, 10, 18, 22, 26, and 29ЊC. sented (Commonwealth ScientiÞc and Industrial Re- Adults were transferred daily to fresh cuttings to mea- search Organization [CSIRO] Entomology 2002). sure daily fecundity and adult survivorship (n ϭ 30, 23, Thus, high temperatures may be limiting the effec- 23, 23, and 20 at 10, 18, 22, 26, and 29ЊC, respectively). tiveness of this agent; however, no data are available Eggs changed from translucent white to opaque yel- quantifying the effect of temperature on demographic low after Ϸ3 d and were then counted because they life history characters. were easier to see. Eggs were monitored daily to re- The geographic distribution of some arthropods in- cord developmental time for eclosion and survivorship troduced as biological control agents of invasive (n ϭ 384, 280, 106, 102, 110, and 79 at 5, 10, 18, 22, 26, Њ weeds has been limited by climate (Byrne et al. 2004, and 29 C, respectively). Neonate L1 nymphs were Robertson et al. 2008). For example, the northern transferred to a young potted plant (7-Ð13 cm in distributions of Aceria chondrillae (Canestrini) and height) in groups of 5 per plant using a Þne pin at- Mecinus janthiniformis Toevski & Caldara are limited tached to an applicator stick (n ϭ 50, 80, 90, 49, 50, and by exposure to low temperatures in North America 98 at 5, 10, 18, 22, 26, and 29ЊC, respectively). Nymphs (De Clerck-Floate and Miller 2002; Milan et al. 2006). were monitored to measure development time and To select agents that are likely to have the most impact survivorship up to adult emergence. Each plant was on the target weed, it would be useful to be able to enclosed in a clear plastic drinking cup that had the assess the potential geographic distribution of a can- bottom cut out and the top screened with organdy, to didate before going through the long expensive pro- prevent escape of nymphs. Two cups were taped to- cess of meeting regulatory permitting requirements, gether in tandem to enclose tall plants. Plants were mass rearing, release, and establishment (McClay and watered as needed with tap water and no fertilizer was Balciunas 2005, Zalucki and Van Klinken 2006). One applied. method is to use existing data on the known distribu- Calculations and Statistical Analysis. Generalized tion of the agent in one region or continent to predict linear models (GLM) and RyanÐEinotÐGabrielÐ its distribution in another (Dhileepan et al. 2012, Mc- Welch multiple range tests (REGWQ) were used to Carren and Scott 2013). However, the geographic analyze the effect of temperature on the quantitative distribution of the target plant in the land of origin may response variables (longevity and fecundity; SAS In- be more constrained than in the adventive region stitute 2003). Response variables that could not be Þt (McFadyen 1991, Broennimann et al. 2014). In this by a linear or theoretical model were Þt by Bezier case, using only geographic data of the agentÕs distri- smoothing curves generated by Microsoft Excel (Figs. bution in the land of origin is likely to underestimate 1 and 2). Logistic regression or ␹2 tests were used for its predicted range in the adventive region. An alter- proportional data (survivorship and sex ratio). The native approach is to measure the quantitative effects Wilson score interval was used to calculate 95% con- of temperature on critical life history characters (By- Þdence intervals for proportion data (Wilson 1927). rne et al. 2004; Manrique et al. 2008, 2012). ConÞdence intervals for the sum of two means (viz., The purpose of this study is to measure the effect of the combined development time of eggs and nymphs) temperature on survivorship, development, and re- were based on estimating the standard deviation ϭ 2 ϩ 2 production of A. hakani to provide a basis for deter- s(1 ϩ 2) sqrt(s1 s2 ) (Sokal and Rohlf 1981). mining its suitable temperature range and optimal Development rates (1/development time) of eggs conditions for population growth. and females (oviposition to adult emergence) were Þt by the nonlinear Sharpe DiMichele biophysical poikilotherm model (Wagner et al. 1984). Materials and Methods ͑ ͒ Organisms. Psyllids originally collected on G. r T monospessulana in southern France (Bormes le`s Mi- T HA 1 1 mosas, France; accession EIWRU-2011--1001) were D exp ͫ ͩ Ϫ ͪͬ T R T T maintained in a laboratory colony on potted G. mon- ϭ o o spessulana plants under ßuorescent grow lamps at HL 1 1 HH 1 1 Ϸ Њ Ϸ 1 ϩ expͫ ͩ Ϫ ͪͬ ϩ expͫ ͩ Ϫ ͪͬ 20 C, 50% relative humidity (RH), and a photo- R TL T R TH T period of 12:12 (L:D) h. Voucher specimens are held at U.S. Department of AgricultureÐAgricultural Re- where r(T) is development rate at temperature T (ЊK), search Service (USDA-ARS) Albany laboratory. Seed- D is development rate at the presumed optimal tem- Њ ϭ Њ lings collected in Marin County, CA, were trans- perature To (298.15 K 25 C) assuming no enzyme planted into 250-cc plastic pots (8 cm in diameter by inactivation, HA is the enthalpy of activation, R is the 5 cm in height) containing a 3:1:1 mix of Supersoil: universal gas constant (1.987 cal/degree/mol), HL is perlite: sand in a greenhouse. Pairs of newly emerged the change in enthalpy associated with low-temper- adult females and males were placed on plant cuttings ature inactivation, TL is the temperature of half-inac- October 2014 SMITH:TEMPERATURE-DEPENDENT DEMOGRAPHY OF Arytinnis hakani 1391

Fig. 2. Development time of eggs (oviposition to eclosion) (a), nymphs (L to adult) (b), and of immatures (ovi- Fig. 1. Survivorship of eggs from oviposition to eclosion 1 position to adult emergence) (c) as a function of constant (a), nymphs from L to adult emergence (b), and of imma- 1 temperature (mean Ϯ 95% CI). tures from oviposition to adult emergence (c) as a function of constant temperature (mean Ϯ 95% CI). ⍀ R ϭ ͸ l m , tivation at low temperature, HH is change in enthalpy 0 x x ϭ associated with high-temperature inactivation, TH is x 1 the temperature of half-inactivation at high temper- where ⍀ is age at death; ature. The equation was modiÞed by substituting 16.5ЊC for 25ЊC as the presumed optimal temperature ⍀ ϭ ϭ ͸ ͑Ϫ ͒ (To 298.15), because it was closer to the center of 1.000 lxmx exp rmx , the linear portion of the Arrhenius plot, which is xϭ1 where the optimal reaction rate should occur (Wag- solved for r by iteration; ner et al. 1984). PROC NLIN using the Marquardt m ␭ ϭ ͑ ͒ method was used to Þt the model to the data (SAS exp rm ; Institute 2003). ϭ ͑ ͒ To Þt a standard degree-day model, the lower de- T ln R0 /rm; velopment temperature threshold was calculated us- DT ϭ ln͑2͒/␭. ing the x-intercept of the linear equation Þtting development rate, and degree-day requirements were The lowest value of x was set based on the mean derived from the inverse of the slope (Campbell et al. developmental time from oviposition to adult emer- 1974). Only data points in the linear part of the de- gence for the given temperature (initial x ϭ develop- ϩ velopment rate curve were used for these calculations. ment time 0.5). The initial values for lx were cal- Net reproductive rate (R0), intrinsic rate of increase culated by multiplying egg and nymphal survivorship ␭ (rm), Þnite rate of increase ( ), mean generation time to use for day 0 of adult life, the initial x. Successive (T, d), and population doubling time (DT, d) were values of lx were based on daily adult survivorship (see calculated following Birch (1948) and Deevey (1947), Fig. 6). Values for mx were calculated by multiplying using age plus 0.5 as the pivotal age (x) and where lx the mean values for age-speciÞc fecundity (eggs per is the probability at birth of a female being alive at age female; see Fig. 6) by the proportion of emerging

x and mx is the mean number of female eggs produced adults that were female for a given temperature (sex at age x: ratio). 1392 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 5

Table 1. Effect of temperature on life history statistics of A. hakani

Nominal temp (ЊC)a 51018222629 Mean temp (ЊC) 5.3 10.0 16.5 21.7 25.5 27.3 SD temp (ЊC) 0.3 0.7 1.1 0.6 0.6 0.3 Mean RH (%) 76% 81% 55% 42% 37% 76% Egg DT (d)b 83.4 34.1 12.6 8.2 7.1 7.5 Female nymphal DT (d)c none 91.5 29.2 21.6 none none Female full DT (d)d none 125.6 41.8 29.8 none none Male nymphal DT (d)c none 88.0 30.2 21.5 none none Male full DT (d)d none 122.1 42.7 29.7 none none Fecunditye Ð 66.7 93.3 168.1 103.0 82.7 Longevity of all females (d)f Ð 26.5 17.4 14.5 10.0 13.4 Longevity of fertile females (d)f Ð 43.7 20.4 16.0 12.6 14.6 Male longevity (d)f 14.8 13.7 12.0 13.2 13.3 9.4 Mean daily fecundity (eggs/d)g Ð 4.5 10.4 16.3 17.2 10.8 Preoviposition period Ð 10.1 5.9 3.1 3.0 3.6 Oviposition period Ð 29.9 11.8 12.1 8.6 8.5 Postoviposition period Ð 3.7 2.3 1.5 1.2 2.7 Sex ratio (% female) Ð 28% 36% 62% Ð Ð Egg survivalh 18% 80% 91% 80% 95% 78% Nymphal survivali 0% 58% 68% 45% 0% 0% j Survival to adult 0% 46% 61% 36% 0% 0% Fig. 3. Development rate of eggs (oviposition to eclosion) (a) and nymphs (L to adult) (b) as a function of a Incubator setting. 1 b constant temperature. Curves are from nonlinear Þt of the Development time from oviposition to eclosion of L1 nymph. c 4-parameter Sharpe DiMichele biophysical poikilotherm Development time from neonate L1 to adult emergence. d Development time from oviposition to adult emergence. model that excludes the low temperature inhibition compo- e Mean fecundity includes that of females that never oviposited. nent. Open diamonds indicate estimates of hypothetical de- f Survival from adult emergence. velopment rate at temperatures in which no insects com- g Only during oviposition period of each female. pleted development (see text for explanation). h Survival from oviposition to eclosion of L1 nymph. i Survival from eclosion of L1 nymph to adult emergence. j Equals egg times nymphal survival. Fig. 1a; Table 1; ␹2 ϭ 286.6, df ϭ 1, P Ͻ 0.001). Nymphal survivorship was optimal at Ϸ14ЊC, and no adults were produced at 5, 25, or 27ЊC (Fig. 1b; Table 1). Immature Climate Modeling. The Match Locations function survival from oviposition to adult emergence was op- in the computer program CLIMEX (v. 1.1, Sutherst et timal (61%) at Ϸ16ЊC (Fig. 1c; Table 1). al. 1999) was used to predict the geographic distribu- Development. Development time of eggs and tion of both the host plant, G. monspessulana, and the nymphs are reported in Table 1 and Fig. 2. At 5.3ЊC, psyllid. Parameter values for the G. monspessulana nymphs survived as long as 250 d and developed as far model were chosen by starting with the CLIMEX as Þfth instar, which indicates that the minimum hy- Mediterranean climate template and adjusting cold pothetical developmental time at this temperature ϩ ϭ ϩ stress parameters to Þt the known geographic distri- was 333 d (egg [L1 through L5] 83.4 250 d). At bution of the plant in Europe. For the A. hakani model, 25.5ЊC, no nymphs were able to complete develop- values for hot and cold stress parameters were based ment from L1 to adult. However, in a separate exper- on the limits we observed for survival of nymphs, and iment starting with L2 nymphs, development time to development rate was based on our observed values adult emergence was 13.3 d, which suggests that hy- for developmental threshold temperature and degree- pothetical development time from oviposition to adult ϩ ϩ ϭ ϩ day requirements (egg to adult). Our analyses used is at least 23.4 d (egg L1 [L2 to adult] 7.1 climatic data from the 2493 weather stations provided 3 ϩ 13.3) at this temperature. in CLIMEX, plus 316 stations in California using 30- Wagner et al.Õs (1984) SAS computer program se- year averages from National Oceanic and Atmo- lected the four-parameter Sharpe DiMichele equation spheric Administration (NOAA) (Fox and Steinmaus that excludes the low temperature inhibition compo- 2001). nent to Þt to the development rate data for eggs and female nymphs (Fig. 3). Parameter estimates wereÑD (using 16.5ЊC) ϭ 0.0896145, HA ϭ 27434.5, Results TH ϭ 296.043, and HH ϭ 39106.9 for egg development The mean temperatures in the incubators differed rate, and D ϭ 0.0367651, HA ϭ 29607.1, TH ϭ 294.817, from the set points (Table 1), so all calculations and and HH ϭ 82394.9 for nymphal development rate. Þgures are based on the mean values. The mean RH The estimates of developmental threshold temper- values reported in Table 1 are of the air outside the atures derived from linear regression of developmen- plastic cups enclosing each plant, but the plants and tal rate were 6.4ЊC for eggs, 6.0ЊC for female nymphs, psyllids probably experienced higher RH because of and 6.1ЊC for egg to adult development. Degree days transpiration. for development were 126, 332, and 454 for eggs, Immature Survivorship. Survivorship of eggs was nymphs, and egg-to-adult, respectively. fairly constant between 10 and 27ЊC, ranging between Sex Ratio. Sex ratio (proportion female) of emerg- 78 and 95%, but was substantially lower at 5ЊC (18%, ing adults increased with temperature between 10 and October 2014 SMITH:TEMPERATURE-DEPENDENT DEMOGRAPHY OF Arytinnis hakani 1393

Fig. 4. Sex ratio of adults as a function of constant temperature (mean Ϯ 95% CI).

21.7ЊC ranging from 28% to 62% (Fig. 4; Table 1; Y ϭ 0.0281ϫX Ϫ 0.0302; logistic regression likelihood ratio P Ͻ 0.02). Sex ratio could not be measured at 5, 25, or 27ЊC because no nymphs completed development at these temperatures. Adult Longevity. Adult female longevity decreased with increasing temperature, and fertile females tended to live longer (LS mean ϭ 22.4 Ϯ 1.2 [SE] d) than those that died without ovipositing (LS mean ϭ 2.6 Ϯ 1.9 d; Fig. 5a; Table 1; GLM: temperature: ϭ Ͻ ϭ Ͻ F(1, 110) 71.8, P 0.0001; fertility: F(1, 110) 46.7, P Fig. 6. Daily fecundity (mean Ϯ SE) and survivorship of 0.0001). Male longevity did not signiÞcantly change female A. hakani at constant temperatures (note different horizontal and vertical scales at 10ЊC). (Online Þgure in color.)

with increasing temperature (LS mean ϭ 11.9 Ϯ 1.0 d; ϭϪ ϫ ϩ 2 ϭ ϭ Fig. 5b; Y 0.1385 X 15.13, R 0.31, F(1,129) 2.48, P Ͼ 0.1). Fecundity. Females were able to oviposit at all temperatures that were tested (10Ð27.3ЊC), and fecundity (eggs per lifespan) was highest at 21.7ЊC (Fig. ϭ Ͻ 5c, Table 1; GLM: F(4, 111) 4.02, P 0.005, REGWQ ␣ ϭ 0.05). Mean daily fecundity increased with temperature between 10 and 25.5ЊC, but decreased at 27.3ЊC (Table 1; Fig. 6). Preoviposition period was fairly constant from 21.7 to 27.3ЊC (3.1Ð3.6 d), but increased up to 10.1 d at 10ЊC. Oviposition period was fairly constant from 16.5 to 27.3ЊC, but increased up to 29.9dat10ЊC. Postoviposition period was fairly constant from 10 and 27.3ЊC, ranging between 1.2 and 3.7 d. Population Growth. The intrinsic rate of increase peaked near 21.7ЊC at 0.0677, with a corresponding Þnite rate of increase (1.095). These data indicate a maximum daily population growth rate of 9.5% per day, assuming a stable age distribution (Table 2; Fig. 7). This corresponds to a doubling time of 7.6 d and generation time of 38.5 d. The highest net reproductive rate (female progeny per female) was observed at 16.5ЊC. Population increase was not possible at temperatures above 25.5ЊC or below 5.3ЊC because of the Fig. 5. Longevity of females (a) and males (b) from adult failure of nymphs to complete development to the emergence and life-time fecundity (c) as functions of con- adult stage. The stable age distribution was calculated stant temperature (mean Ϯ 95% CI). at 10.0, 16.5, and 21.7ЊC (Table 2), and at all temper- 1394 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 5

Table 2. Effect of temperature on demographic statistics fecundity also decreased, but had less effect. Eggs and nymphs survived well down to at least 10ЊC, but not to Њ a Nominal temp ( C) 10 18 22 5ЊC, which limited population growth at cold temper- Mean temp (ЊC) 10.0 16.5 21.7 atures. The ability of females to oviposit was not mea- rm 0.0141 0.0677 0.0906 sured at 5ЊC, but we suspect that the rate may be very ␭ 1.014 1.070 1.095 T 153.3 54.1 38.5 low at this temperature. Thus, independent of other Ro 8.67 38.8 32.7 stresses to the host plant, such as drought or shading, Doubling time (d) 49.2 10.2 7.6 we would expect this psyllid to have the most impact Stable age distribution: at Ϸ22ЊC, and to have little impact above 25ЊC. Adults Eggs 52.3% 64.7% 62.9% Ϫ Њ Nymphs 44.4% 32.8% 35.0% and nymphs did not survive freezing at 4 C, and we Adults 3.3% 2.6% 2.0% saw no signs of diapause in any developmental stages; however, we did not study the possible effect of pho- ␭ rm, intrinsic rate of increase; , Þnite rate of increase; T, generation toperiod on diapause. Occurrence of diapause would time (d); Ro, net reproductive rate (female/female/generation). a presumably help the insect persist during climatically Incubator setting. adverse seasons and increase its predicted geographic range; however, we are not aware of any reports of atures eggs comprised the greatest proportion of the diapause. three principal life stages, and adults comprised Ͻ4%. Developmental time previously measured at diurnal Њ Climate Modeling. Parameter values for the temperature alternating between 20 and 15 C (pho- CLIMEX models for G. monspessulana and A. hakani toperiod unreported) was 9.3 d for eggs and 26.3 d for are listed in Table 3. In Australia, the region predicted nymphs (CSIRO Entomology 2002; their Fig. 2). for the plant (Fig. 8a) is similar to its reported distri- These values lie close to the curves in Fig. 2A and B, Њ bution (Sheppard and Henry 2012). The region pre- using the mean value of 17.5 C. Generation time, ap- dicted to be suitable for the psyllid is smaller than parently measured as the time from oviposition until that for the plant (Fig. 8b), being restricted on its the Þrst egg of the next generation, was 32-Ð47 d northern limit by excessive heat stress (not shown). (CSIRO Entomology 2002; their Table 7), which is Њ In California, the region predicted for the plant (Fig. lower than our prediction would be (51.1 d at 17.5 C, 9a) is similar to its reported distribution (Jepson In- by linear interpolation); however, the method used by terchange, http://ucjeps.berkeley.edu/interchange. these authors normally would underestimate the value html). The region predicted to be suitable for the compared with that calculated by the standard for- psyllid is smaller than that for the plant (Fig. 9b), being mula for generation time (Birch 1948). restricted on its eastern limit by excessive cold, heat Herrera et al. (2011) observed developmental times Ϸ stress, or both (not shown). The model predicted ranging from 9.2 to 9.4 d for eggs and 32.0Ð33.9 d for completion of between four to eight generations per nymphs (their Fig. 1) measured on a variety of plant year over the latitudinal range in coastal California, accessions at unregulated ßuctuating diurnal temper- Њ four to nine generations in the Mediterranean, and ature in the laboratory (18Ð27 C and a photoperiod of Þve to nine in Australia. 16:8 [L:D] h). Lack of precise data records prevents directly comparing their results to ours. Egg survivorship ranged from 91 to 99%, and nymphal survivorship Discussion was Ϸ70Ð88% (their Fig. 2), which are comparable to Effect of Temperature on A. hakani. The results our results. Preoviposition period was 3.9Ð5.6 d, ovi- Ϸ indicate that A. hakani is best adapted to temperatures position period was 17Ð20 d, female longevity was Ϸ Ϸ between 10 and 22ЊC. Failure of nymphs to complete 20Ð25 d (from their Fig. 3). Females laid 13Ð18 Ϸ development above Ϸ25ЊC was the principal con- eggs per day and 230Ð370 eggs per lifetime (from straint to population growth at warm temperatures, their Fig. 4). They reported that most females laid Ͻ despite high survivorship of eggs. Adult longevity and 101Ð618 eggs, but a few laid 100. These values are comparable to ours except that the oviposition period appears to have been longer. Stable Age Distribution. The fact that adults com- prise only Յ3% of A. hakani population that has a stable age distribution, suggests that sampling for adults at Þeld sites, such as by beating branches, would usually give a very low indication of the total number of psyllids present. For example, Sheppard and Thom- ann (2004) surveyed sites in Europe by beating trees and observed an average of 12 A. hakani per plant. Although they did not report what life stages were collected, in our experience nymphs are very resistant to falling off branches whereas adults readily jump. Thus, 12 adult psyllids would normally represent a Fig. 7. Finite rate of increase as a function of constant population of Ϸ400 psyllids of all stages (ϭ12/0.03). temperature. Although mortality caused by predators, heavy pre- October 2014 SMITH:TEMPERATURE-DEPENDENT DEMOGRAPHY OF Arytinnis hakani 1395

Table 3. Parameter values for the CLIMEX compare locations models for G. monspessulana (Gemo) and A. hakani (Arha) compared with the template for Mediterranean climate

Parameter values for each model Parameter Description Mediterranean Gemo Arha Temperature DV0 Limiting low temp 10ЊC10ЊC6ЊC DV1 Lower optimal temp 16ЊC16ЊC21ЊC DV2 Upper optimal temp 24ЊC26ЊC22ЊC DV3 Limiting high temp 28ЊC30ЊC26ЊC PDD Minimum degree days above DV0 600 600 454 Moisture SM0 Limiting low moisture 0.1 0.2 0.1 SM1 Lower optimal moisture 0.4 0.5 0.4 SM2 Upper optimal moisture 0.7 1.0 1.5 SM3 Limiting high moisture 1.5 2.0 2.5 Cold stress TTCS Cold stress temp threshold 0 0 2 THCS Cold stress accumulation rate 0.005 0.005 0.5 DTCS Cold stress degree-day threshold 15 10 10 DHCS Cold stress degree-day rate 0.001 0.001 0.01 Heat stress TTHS Heat stress temp threshold 30 30 28 THHS Heat stress accumulation rate 0.002 0.002 0.01 DTHS Heat stress degree-day threshold 0 0 26 DHHS Heat stress degree-day accumulation rate 0 0 0.1 Dry stress SMDS Dry stress threshold 0.02 0.02 0.2 HDS Dry stress accumulation rate 0.05 0.05 0.05 Wet stress SMWS Wet stress threshold 1.6 1.6 1.6 HWS Wet stress accumulation rate 0.0015 0.0015 0.0015 Hot-dry stress TTHD Hot-dry temp threshold 0 30 0 MTHD Hot-dry moisture threshold 0 0.3 0 PHD Hot-dry stress accumulation rate 0 0.05 0 Hot-wet stress TTHW Hot-wet temp threshold 23 23 23 MTHW Hot-wet moisture threshold 0.5 0.5 0.5 PHW Hot-wet stress accumulation rate 0.075 0.075 0.075

All other parameters were 0. See text for explanation. cipitation, or short-term temperature extremes can (Ferris and Klyver), had highest population growth affect the age distribution encountered in the Þeld, the rate at 25ЊC, but failed to complete development at stable age distribution is representative of populations 30ЊC, the next highest temperature studied (Madubu- with overlapping generations that are growing fast. nyi and Koehler 1974). Comparison With Other Psyllids. Of the psyllids The psyllid Boreioglycaspis melaleucae Moore, a for which the effect of temperature on demographic biological control agent of the subtropical tree, characters has been studied, A. hakani appears to be Melaleuca quinquenervia, could develop and oviposit adapted to the lowest temperatures. The developmen- between 15Ð30ЊC (Chiarelli et al. 2011). No nymphs tal threshold of the Asian citrus psyllid, D. citri, was survived, and fecundity was severely reduced at 10ЊC. Ϸ10Ð11ЊC (Liu and Tsai 2000), whereas it was 6ЊC for These authors did not calculate intrinsic rate of in- A. hakani. D. citri failed to develop at the extreme crease, but it probably was highest near 25ЊC. The temperatures of 10 and 33ЊC, versus 5 and 26ЊC for A. climate in south Florida appears to be suitable for this hakani, and its intrinsic rate of increase peaked at 28ЊC, insect, which has become established and is helping to versus 22ЊC. D. citri adults lived longer (35 vs. 16 d for suppress growth and reproduction of M. quinquenervia fertile females at 28 and 21.7ЊC, peak temperatures for (Tipping et al. 2008). The psyllid, Prosopidopsylla ﬂava D. citri and A. hakani population growth, respectively) Burkhardt, was introduced to control mesquite and were much more fecund than A. hakani (748 vs. (Prosopis spp.) in Australia, but has established only at 168 eggs). Although optimal conditions for the potato the two coolest sites, which have climate most similar psyllid, Bactericera nigricornis (Forster), have not to their region of origin (Van Klinken et al. 2003). been determined, at 25ЊC female longevity was 24 d, Comparison to climate of the region of origin in Ar- fecundity was 63 eggs, and Þnite rate of increase was gentina suggests that most Australian sites were too 1.060 on the most suitable potato variety tested (Fathi warm. At diurnally ßuctuating temperature 27/23ЊC, 2011). These numbers are more comparable with adult longevity was 20 d, fecundity was 232 eggs. The those of A. hakani, although at a lower temperature Japanese knotweed psyllid, Aphalara itadori Shinji, such as 18ЊC. The Albizzia psyllid, Psylla uncatoides could develop between 15Ð30ЊC, although lower tem- 1396 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 5

Fig. 8. Predicted geographic distribution of G. monspessulana (a) and A. hakani (b) in Australia based on the Eco- climatic Index calculated by the Match Locations function of CLIMEX (parameter values are in Table 3). (Online Þgure in color.) peratures were not tested (Myint et al. 2012). The Fig. 9. Predicted geographic distribution of G. monspes- intrinsic rate of natural increase was highest at 25ЊC, sulana (a) and A. hakani (b) in California based on the and the population growth could occur between 15 Ecoclimatic Index calculated by the Match Locations func- Њ tion of CLIMEX (parameter values are in Table 3). (Online and 30 C. So, these three species also appear to be Þgure in color.) adapted to warmer climates than A. hakani. Predicting Climatic Envelope. Our results can be used to predict population growth as a function of temperature and can be used in climatological geo- of another psyllid, B. melaleucae (Chiarelli et al. 2011), graphic information system (GIS) models as a basis for which is not surprising given that the insects have predicting where the psyllid should perform well (Van access to healthy plants which supply them water. Klinken et al. 2003, Chiarelli et al. 2011, Myint et al. 2012, May and Coetzee 2013). The predicted distri- Acknowledgments bution of A. hakani in Australia matches well with locations where it has been observed to be abundant I thank Rene´ Sforza, USDA-ARS, European Biological (New South Wales, Victoria, South Australia, and Tas- Control Laboratory, for collecting and shipping the psyllids to California, and M. Irene Wibawa and Caroline Nunn for mania; R.T. Roush and A.W. Sheppard, personal com- growing test plants, maintaining the A. hakani colony, and munication) and indicates that the psyllid is restricted conducting the experiment. I am grateful to Fritzi S. on its northern limit by heat stress. The psyllid is Grevstad, Oregon State University, and Paul D. Pratt, USDA- predicted to be suitable throughout the coastal distri- ARS, for comments on a previous draft. bution of G. monspessulana in California; however, it is predicted to not do well in the Sierra foothills because of both cold stress in winter and heat stress in References Cited summer. Although relative humidity was not included Birch, L. 1948. The intrinsic rate of natural increase of an in our models, it had no effect on developmental rate insect population. J. Anim. Ecol. 17: 15Ð26. October 2014 SMITH:TEMPERATURE-DEPENDENT DEMOGRAPHY OF Arytinnis hakani 1397

Broennimann, O., B. Petitpierre, A. Guisan, P. Mra´z, and H. Hodkinson, I. D., and D. Hollis. 1987. The legume-feeding Mu¨ ller-Scha¨rer. 2014. Contrasting spatio-temporal cli- psyllids (Homoptera) of the west Palaearctic Region. matic niche dynamics during the eastern and western Bull. Br. Mus. Nat. Hist. Entomol. 56: 1Ð86. invasions of spotted knapweed in North America. J. Bio- Liu, Y. H., and J. H. Tsai. 2000. Effects of temperature on geogr. 41: 1126Ð1136. biology and life table parameters of the Asian citrus psyl- Byrne, M. J., J. Coetzee, A. J. McConnachie, W. Parasram, lid, Diaphorina citri Kuwayama (Homoptera: Psyllidae). M. P. Hill, J. M. Cullen, D. T. Briese, D. J. Kriticos, W. M. Ann. Appl. Biol. 137: 201Ð206. Lonsdale, L. Morin, et al. 2004. Predicting climate com- Madubunyi, L. C., and C. S. Koehler. 1974. Development, patibility of biological control agents in their region of survival and capacity for increase of the Albizzia psyllid introduction, pp. 28Ð35. In Proceedings of the XI Inter- at various constant temperatures. Environ. Entomol. 3: national Symposium on Biological Control of Weeds, 27 1013Ð1016. AprilÐ2 May 2003, Canberra, Australia. Manrique, V., W. A. Overholt, R. Diaz, and J. P. Cuda. 2008. Campbell, A., B. D., Frazer, N. Gilbert, A. P. Gutierrez, and Temperature-dependent development and potential dis- M. MacKauer. 1974. Temperature requirements of some tribution of Episimus utilis (Lepidoptera: Tortricidae), a aphids and their parasites. J. Appl. Ecol. 11: 431Ð438. candidate biological control agent of Brazilian pepper- Center, T. D., P. D. Pratt, P. W. Tipping, M. B. Rayamajhi, T. K. tree (Sapindales: Anacardiaceae) in Florida. Environ. En- Van, S. A. Wineriter, and F. A. Dray. 2007. Initial impacts tomol. 37: 862Ð870. and Þeld validation of host range for Boreioglycaspis Manrique, V., R. Diaz, W. A. Overholt, M. G. Pogue, and melaleucae Moore (Hemiptera: Psyllidae), a biological M. D. Vitorino. 2012. Description and biology of Paectes control agent of the invasive tree Melaleuca quinquenervia longiformis (Lepidoptera: Euteliidae), a new species (Cav.) Blake (Myrtales: Myrtaceae: Leptospermoideae). from Brazil and potential biological control agent of Bra- Environ. Entomol. 36: 569Ð576. zilian peppertree in Florida. Biocontrol Sci. Technol. 22: Chiarelli, R. N., P. D. Pratt, C. S. Silvers, J. S. Blackwood, and 163Ð185. T. D. Center. 2011. Inßuence of temperature, humidity, May, B., and J. Coetzee. 2013. Comparisons of the thermal and plant terpenoid proÞle on life history characteristics physiology of water hyacinth biological control agents: of Boreioglycaspis melaleucae (Hemiptera: Psyllidae), a Predicting establishment and distribution pre- and post- biological control agent of the invasive tree Melaleuca release. Entomol. Exp. Appl. 147: 241Ð250. quinquenervia. Ann. Entomol. Soc. Am. 104: 488Ð497. McCarren, K. L., and J. K. Scott. 2013. Host range and po- Cook, B. S., and L. Smith. 2014. Prerelease efÞcacy test of tential distribution of Aceria thalgi (Acari: Eriophyidae): the psyllid, Arytinnis hakani, a prospective biological con- a biological control agent for Sonchus species. Aust. J. trol agent of the invasive weed Genista monspessulana. Entomol. 52: 393Ð402. Biocontrol Sci. Technol. 24: 641Ð651. (http://dx.doi.org/ McClay, A. S., and J. K. Balciunas. 2005. The role of pre- 10.1080/09583157.2014.884993). release efÞcacy assessment in selecting classical biologi- (CSIRO) Commonwealth Scientiﬁc and Industrial Research cal control agents for weeds Ñ applying the Anna Kar- Organization Entomology. 2002. Foreign exploration enina principle. Biol. Control 35: 197Ð207. and host speciÞcity testing of biological control agents of McFadyen, R.E.C. 1991. Climate modelling and the biolog- French broom in California. International Broom Initia- ical control of weeds: one view. Plant Prot. Q. 6: 14Ð15. tive 2001Ð2002. Contracted Research Report. (http:// www.cal-ipc.org/Þle_library/4859.pdf). Milan, J. D., B. L. Harmon, T. S. Prather, and M. Schwar- De Clerck-Floate, R., and V. Miller. 2002. Overwintering zlander. 2006. Winter mortality of Aceria chondrillae, a mortality of and host attack by the stem-boring weevil, biological control agent released to control rush skel- Mecinus janthinus Germar, on Dalmatian toadßax (Li- etonweed (Chondrilla juncea) in the western United naria dalmatica (L.) Mill.) in western Canada. Biol. Con- States. J. Appl. Entomol. 130: 473Ð479. trol 24: 65Ð74. Munyaneza, J. E. 2012. Zebra chip disease of potato: Biol- Deevey, E. S. 1947. Life tables for natural populations of ogy, Epidemiology, and Management. Am. J. Potato Res. animals. Q. Rev. Biol. 22: 283Ð314. 89: 329Ð350. Dhileepan, K., M. Trevino, K.A.D. W. Senaratne, R. E. Mc- Munyaneza, J. E., V. G. Sengoda, J. L. Buchman, and T. W. Fadyen, M. P. Vitelli, and A. S. McClay. 2012. Introduc- Fisher. 2012. Effects of temperature on ‘Candidatus Li- tion, establishment, and potential geographic range of beribacter solanacearumÕ and zebra chip potato disease Carmenta sp. nr ithacae (Lepidoptera: Sesiidae), a bio- symptom development. Plant Dis. 96: 18Ð23. logical control agent for Parthenium hysterophorus (As- Myint, Y. Y., K. Nakahira, M. Takagi, N. Furuya, R. H. Shaw. teraceae) in Australia. Environ. Entomol. 41: 317Ð325. 2012. Using life-history parameters and a degree-day Fathi, S.A.A. 2011. Population density and life-history pa- model to predict climate suitability in England for the rameters of the psyllid Bactericera nigricornis (Forster) Japanese knotweed psyllid Aphalara itadori Shinji on four commercial cultivars of potato. Crop Prot. 30: (Hemiptera: Psyllidae). Biol. Control 63: 129Ð134. 844Ð848. Robertson, M. P., D. J. Kriticos, and C. Zachariades. 2008. Fox, J. C., and S. Steinmaus. 2001. Climatic prediction of an Climate matching techniques to narrow the search for invasive plant in California: Ulex europaeus (gorse), pp. biological control agents. Biol. Control 46: 442Ð452. 34Ð37. In Proceedings of the 53rd Annual California SAS Institute. 2003. SAS for Windows 9.1.3 Service Pack 4. Weed Science Society, California Weed Science Society, SAS Institute, Cary, NC. Salinas, CA. Shaw, R. H., S. Bryner, and R. Tanner. 2009. The life history Grafton-Cardwell, E. E., L. L. Stelinski, and P. A. Stansly. and host range of the Japanese knotweed psyllid, 2013. Biology and management of asian citrus psyllid, Aphalara itadori Shinji: Potentially the Þrst classical bi- vector of the huanglongbing pathogens. Annu. Rev. En- ological weed control agent for the Eur. Union Biol. tomol. 58: 413Ð432. Control 49: 105Ð113. Herrera, A. M., R. I. Carruthers, and N. J. Mills. 2011. No Sheppard, A. W., and K. Henry. 2012. Genista monspessu- evidence for increased performance of a specialist psyllid lana (L.) L. JohnsonÐCape broom, pp. 267Ð273. In J. on invasive French broom. Acta Oecol. 37: 79Ð86. Cullen, M. Julien, R. McFadyen (eds.), Biological Control 1398 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 5

of Weeds in Australia. CSIRO Publishing, Collingwood, Tipping, P. W., M. R. Martin, P. D. Pratt, T. D. Center, M. B. VIC, Australia. Rayamajhi. 2008. Suppression of growth and reproduc- Sheppard A. W., and T. Thomann. 2004. Quantitative Þeld tion of an exotic invasive tree by two introduced insects. surveys for the selection of biological control agents for Biol. Control 44: 235Ð241. Genista monspessulana based on host range and efÞcacy Van Klinken, R. D., G. Fichera, and H. Cordo. 2003. Tar- assessment, pp. 162Ð174. In J. M. Cullen, D. T. Briese, D. J. geting biological control across diverse landscapes: the Kriticos, W. M. Lonsdale, L. Morin, and J. K. Scott (eds.), release, establishment, and early success of two insects on Proceedings of the XI International Symposium on Bio- mesquite (Prosopis spp.) insects in Australian rangelands. logical Control of Weeds. CSIRO Entomology, Canberra, Biol. Control 26: 8Ð20. Australia. Wagner, T. L., H. Wu, P.J.H. Sharpe, R. M. Schoolﬁeld, and Sheppard A., T. Thomann, S. Ivory, and R. Sforza. 2014. R. N. Coulson. 1984. Modeling insect development Successful biological control of Montpellier broom rates: a literature review and application of a biophysical model. Ann. Entomol. Soc. Am. 77: 208Ð225. (Genista monspessulana (L.) L.A.S. Johnson) in Australia. Wilson, B. W., and C. A. Garcia. 1922. Host speciÞcity and In Proceedings of the XIV International Symposium on biology of Heteropsylla spinulosa (Hom.: Psyllidae) in- Biological Control of Weeds, 2Ð7 March 2014, Kruger troduced into Australia and Western Samoa for the bio- National Park, South Africa. logical control of Mimosa invisa. Entomophaga 37: 293Ð Sutherst, R. W., G. F. Maywald, T. Yonow, and P. M. Stevens. 299. 1999. CLIMEX: Predicting the effects of climate on plant Wilson, E. B. 1927. Probable inference, the law of succes- and animals. CSIRO Publishing, Collingwood, Australia. sion, and statistical inference. J. Am. Stat. Assoc. 22: 209Ð Sokal, R. R., and F. J. Rohlf. 1981. Biometry, 2nd edn. W. H. 212. Freeman, San Francisco, CA. Zalucki, M. P., and R. D. Van Klinken. 2006. Predicting Syrett, P., S. V. Fowler, H. M. Harman, L. M. Hayes, J. population dynamics of weed biological control agents: Memmott, and J. J. Sheat. 2007. Establishment of Ar- Science or gazing into crystal balls? Aust. J. Entomol. 45: ytainilla spartiophila Forster (Hemiptera: Psyllidae), a 331Ð344. new biological control agent for broom, Cytisus scoparius, in New Zealand. N. Z. Entomol. 30: 53Ð62. Received 31 March 2014; accepted 1 July 2014.