See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/261722670

Genetic diversity in insulata and its implications for biological control of

Article in BioControl · December 2013 DOI: 10.1007/s10526-013-9558-3

CITATIONS READS 6 104

5 authors, including:

Nontembeko Dube Yoseph Assefa Agricultural Research Council, South Africa University of Swaziland

7 PUBLICATIONS 84 CITATIONS 65 PUBLICATIONS 375 CITATIONS

SEE PROFILE SEE PROFILE

Costas Zachariades Terence Olckers Agricultural Research Council, South Africa University of KwaZulu-Natal

56 PUBLICATIONS 1,083 CITATIONS 85 PUBLICATIONS 1,287 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Biological control of Chromolaena odorata View project

Eldana cold tolerance View project

All content following this page was uploaded by Terence Olckers on 05 August 2014.

The user has requested enhancement of the downloaded file. Genetic diversity in Pareuchaetes insulata and its implications for biological control of Chromolaena odorata

Nontembeko Dube, Yoseph Assefa, Costas Zachariades, Terence Olckers & Des Conlong

BioControl Journal of the International Organization for Biological Control

ISSN 1386-6141 Volume 59 Number 2

BioControl (2014) 59:253-262 DOI 10.1007/s10526-013-9558-3

1 23 Your article is protected by copyright and all rights are held exclusively by International Organization for Biological Control (IOBC). This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23 Author's personal copy

BioControl (2014) 59:253–262 DOI 10.1007/s10526-013-9558-3

Genetic diversity in Pareuchaetes insulata and its implications for biological control of Chromolaena odorata

Nontembeko Dube • Yoseph Assefa • Costas Zachariades • Terence Olckers • Des Conlong

Received: 12 October 2012 / Accepted: 9 December 2013 / Published online: 20 December 2013 Ó International Organization for Biological Control (IOBC) 2013

Abstract Chromolaena odorata (L.) King & H.E. analyses and cross-mating experiments were conducted. Robins. (Asteraceae) threatens several economic and Mitochondrial DNA sequences revealed significant environmental activities in the eastern subtropical genetic similarity between them. Cross-mating trials regions of South Africa and is a target for biological between Floridian and Jamaican populations of P. control. Three populations of Pareuchaetes insulata insulata revealed no significant pre- and post-zygotic (Walker) (: ) originating from isolation, with no demonstrable differences in measured

Florida (USA), Cuba and Jamaica were released at 30 parameters between the ‘parental’ populations, the F1 separate sites in KwaZulu-Natal, South Africa. Sub- ‘hybrids’ and the backcrosses. These results suggest that sequent overlapping of two of these populations in the P. insulata populations released in South Africa are part field and the likelihood of interbreeding posed a of the same ‘parental’ population. Genetic incompati- potential risk of establishment failure. To explore bility is, therefore, relatively unlikely to be the cause of the genetic diversity of the different P. insulata any failure in establishment of P. insulata at any of the populations and test for the existence of pre- and post- release sites. zygotic reproductive isolation between them, molecular Keywords Asteraceae Backcross Fertility Fitness Lepidoptera: Erebidae Mitochondrial Handling Editor: John Scott. DNA South Africa N. Dube (&) C. Zachariades Agricultural Research Council, Plant Protection Research Institute, Private Bag X6006, Hilton 3245, South Africa Introduction e-mail: [email protected]

N. Dube C. Zachariades T. Olckers D. Conlong Chromolaena odorata (L.) King & H.E. Robins. School of Life Sciences, University of KwaZulu-Natal, (Asteraceae: Eupatorieae), is a perennial pioneering Private Bag X01, Scottsville 3209, South Africa shrub native to the tropical and subtropical Americas. It is one of the worst weeds in the humid tropics and Y. Assefa Department of Zoology, Faculty of Science and subtropics of the Old World and is highly invasive in Agriculture, University of Fort Hare, Private Bag X1314, several parts of Africa, Asia and Oceania (Zachariades Alice 5700, South Africa et al. 2009). A biotype of C. odorata distinct from that invading other areas was first recorded as naturalized D. Conlong South African Sugarcane Research Institute, Private Bag in South Africa in the 1940s (Zachariades et al. 2011), X02, Mount Edgecombe 4300, South Africa and has since spread rapidly in the eastern subtropical 123 Author's personal copy

254 N. Dube et al. regions of South Africa, where it is considered one of due to ‘hybrid’ infertility and diminished fitness, in the the worst invasive alien plants (Goodall and Erasmus event that the two populations had significant genetic 1996). differentiation. In an attempt to biologically control C. odorata, Where species constitute geographically separated releases of three populations of Pareuchaetes insulata populations, as occurs with P. insulata, there is a high (Walker) (Lepidoptera: Erebidae) collected from Florida possibility that genetic differences have evolved (USA), Cuba and Jamaica were conducted at 30 sites in between these populations (e.g. Assefa et al. 2006a). KwaZulu-Natal (KZN) province between 2001 and 2009 Indeed, several hundred studies have reported the (Zachariades et al. 2011). The congeneric Pareuchaetes findings of gene 9 environment interactions in a variety pseudoinsulata Rego Barros, used previously as a of organisms (e.g. Futuyma and Peterson 1985). Genetic biological control agent on C. odorata in several regions variation in geographically isolated populations has in the paleotropics (Zachariades et al. 2009), failed to frequently been reported to cause the development of establish in South Africa, possibly due to poor climatic biotypes in both pests and natural enemies (e.g. Avise matching (Zachariades et al. 2011). A population of P. 2000;Assefaetal.2006a;Sezonlinetal.2006). Testing insulata from Florida was subsequently used in South for genetic divergence and/or pre- and post-mating Africa because of the closer climate match between the isolation in weed biological control agents from differ- two regions (Kluge and Caldwell 1993a). However, after ent provenances is either not practiced or practiced releases of about 781,000 at 17 sites around KZN infrequently, but see Mound et al. (2010). However, over a two-year period (Zachariades et al. 2011), initial given the high degree of specificity displayed by some monitoring indicated poor or no establishment. Having weed biocontrol agents to their host plants, and the eliminated all reasonable causes, apart from poor agent/ potential for disruption of this specificity through host-plant biotype matching—the C. odorata biotype interbreeding of different biotypes of an agent (Hoff- invading southern Africa differs morphologically from mann et al. 2002), it was considered important to take the C. odorata in Florida—two further P. insulata this possibility into consideration in the P. insulata cultures were imported, one each from Jamaica and release programme. Cuba, from where the southern African C. odorata The past two decades have witnessed an increasing biotype originates (Paterson and Zachariades 2013). interest in the process by which barriers to gene flow Insects from the Jamaican population were released at evolve between populations as a result of geographical seven sites in KZN between 2003 and 2008, and those and ecological divergent selections (Avise 2000). from the Cuban population at five sites between 2003 and Mitochondrial DNA (mtDNA) has become the most 2006, but these either disappeared or persisted only at low commonly employed marker and has proven powerful numbers (Zachariades et al. 2011). In late 2004, the site at for genealogical and evolutionary studies of which the largest number (335,000) of Floridian P. populations and for determining genetic relationships insulata were released, on the KZN coast near the town of among populations (Sperling et al. 1999; Avise 2000; Umkomaas (Cannonbrae plantation, 30°13.2360S, Scheffer 2000; Simmons and Scheffer 2004). Due to 30°46.3920E), was found to have indeed supported maternal inheritance and a relatively faster rate of establishment. During the summer of 2005–2006, a evolution, mtDNA has been used to provide insights population outbreak occurred, causing extensive defoli- into population genetic structure, gene flow, biogeog- ation of C. odorata close to the release site (Zachariades raphy and intraspecific relationships (Sperling et al. et al. 2011). By 2009, eight years after the first release, the 1999; Simmons and Scheffer 2004). Such data are also Floridian population of P. insulata had spread for more capable of revealing cryptic lineages representing than 50 km along the KZN coast in either direction from distinct species within geographically widespread and the release site as well as 10 km inland. This expansion apparently morphologically homogeneous organisms by the Floridian population of P. insulata soon resulted in like P. insulata (Scheffer 2000; Mound et al. 2010). an overlap between this established Floridian site and two However, results of mtDNA analysis are more mean- Jamaican population release sites (at Finningley Estates, ingful when additional data are available for compar- 30°14.5060S, 30°46.3920E and at Umdoni Trust, ison. Information on morphology, ecology, 30°23.0570S, 30°40.2480E), posing the likelihood of hybridization and differences in nuclear DNA are used interbreeding and a potential risk of establishment failure together with mtDNA markers in studying intraspecific 123 Author's personal copy

Lack of genetic diversity in Pareuchaetes insulata 255 and interspecific genetic relationships between organ- as recommended for DNA isolation from animal tissue, isms (Simmons and Scheffer 2004). and the extracted DNA was stored at -20 °C until This study examines genetic differentiation in three required for amplification. Polymerase chain reaction geographically isolated populations of P. insulata, one (PCR) amplification was performed in a 50 llvolume each from Cuba, Jamaica and southern Florida in the containing 19 PCR buffer, 1.5 mM MgCl2,200lM USA. It also investigates the effect of potential genetic dNTPs, 15 pmol of each PCR primer, one unit of differentiation on the fertility, longevity and fitness of SuperTherm Gold Taq DNA polymerase and 1 llof inter-population crosses of Floridian and Jamaican genomic DNA. PCR was performed using a Perkin Elmer populations of this biological control agent. GeneAmp PCR System 2400, using the reaction condi- tions detailed in Assefa et al. (2006a). Amplified DNA was purified using the Qiagen QIAquickTM PCR purifi- Materials and methods cation kit, following the manufacturer’s protocol. DNA sequencing reactions were performed using the ABI populations PRISMÒ BigDyeTM Terminator v3.0 Ready Reaction Cycle Sequencing Kit, cleaned using ethanol/EDTA Stock colonies of Floridian, Jamaican and Cuban precipitation with slight modification of the manufac- populations of P. insulata were established at the South turer’s protocol, and sequences were visualized on an African Sugarcane Research Institute (SASRI) for mass ABI 3100 Genetic Analyzer. Primers used in the study rearing and release against C. odorata in South Africa. were as mentioned in Assefa et al. (2007). Editing and Pareuchaetes insulata specimens used in this study assembling of DNA sequence chromatograms was done were obtained from the same SASRI laboratory colo- using the Staden package (Staden 1996). Sequences were nies. The Floridian colony was first imported in 2000, then aligned using ClustalX (Thompson et al. 1997)and while those from Jamaica and Cuba were first imported manually corrected using BioEdit sequence alignment in 2002 (Zachariades et al. 2011). Further insects were editor (Hall 1999). Phylogenetic analysis was performed imported from Florida in 2002 and from Jamaica in by maximum parsimony (MP) in MEGA4 (Tamura et al. 2003, and integrated into their respective SASRI 2007). colonies. The Floridian colony was terminated in 2003, while the Jamaican colony died out in 2004. A Inter-population mating new Jamaican colony was imported in late 2004 and supplemented with additional imported material in ‘Hybridization’ trials were conducted in 2007, by cross- 2005. Insects from the established Floridian population mating from the Floridian population with those at Cannonbrae were collected on four occasions in from the Jamaican population. For each cross, the five 2006–2007 to form a new colony at SASRI and then pairs of adults were kept together in one container on the supplement it. The study was conducted in two phases. first night to optimise mating. However, this meant that In the first phase, DNA sequence analysis of regions of any eggs laid on the same night could not be assigned to the mitochondrial cytochrome oxidase c subunit I gene individual females, and thus had to be excluded from the (COI) was performed on a total of 18 P. insulata analysis. Overall, 6.64 % of all eggs laid were deposited samples, six from each of the three regional populations. on this first night (Table 1). For the purpose of this paper, In the second phase, the fertility of ‘hybrids’ from the adult fecundity is taken as the sum of the number of eggs Floridian and Jamaican populations was assessed laid by a female over her lifetime and the number retained through crossbreeding and back-crossing trials. The in her ovarioles, while embryonic viability is indicated by Cuban population of P. insulata could not be included in the number of eggs hatched. Control matings were the second phase as the colony at SASRI was killed by a conducted using ‘parental’ stocks, and the ‘hybrid’ fungal disease before the study commenced. broods were reared simultaneously with control broods

of the ‘parental’ stocks. The F1 ‘hybrid’ adults were then Mitochondrial DNA sequence divergence backcrossed to the ‘parental’ type control broods. The types of crossings carried out are indicated in Table 1. Genomic DNA was extracted from individual thoraces of The ‘parental’ and ‘hybrid’ colonies were maintained the 18 specimens using the Qiagen DNeasyTM Tissue Kit in the laboratory under standardised conditions (27 °C 123 Author's personal copy

256 N. Dube et al. and 78 % RH). Newly emerged adults were sexed and Neonates hatching from the leaves and the cage/jar walls five females were placed together with five males in a were recorded daily. 30 9 30 9 30 cm perspex cage for 24 h to allow Larval rearing methods differed slightly between optimal mating. The moths were then separated into five ‘parental’ broods and the ‘hybrid’ and back-crossed individual pairs which were placed into 1 l plastic mating broods, because the ‘parental’ broods were also being jars with gauze lids (one pair per jar) containing used to determine the development of individual instars. 5 9 5 9 5cmOasisÒ floral foam blocks with five leafy In total, 50 neonates from each of the ‘parental’ cuttings of C. odorata each measuring 10–12 cm in Floridian and Jamaican populations, subsampled hap- length. Small cotton-wool plugs (2 cm diameter), hazardly from among the five pairs, were placed into sprayed with a 50:50 honey:water solution, were pro- Petri dishes containing C. odorata leaves, at ten per dish. vided for nutrition. Eggs laid on the upper and lower Thereafter they were transferred to clean Petri dishes surfaces of the leaves and on the walls and lids of the with fresh leaves every two days. The third instars were perspex cage and the adult mating jars were counted. transferred to 2 l plastic trays (22 9 32 9 5cm)with Leafy cuttings were replaced with fresh C. odorata gauze-panelled lids containing 15 C. odorata bouquets bouquets daily and leaves on which eggs had been laid stalked in a 15 9 5 9 5cmOasisÒ block, to a maxi- were placed in 90 mm diameter plastic Petri dishes and mum of ten larvae per tray, and were relocated to clean labelled with the date and the pair number. Eggs laid on trays with fresh bouquets every two days until they the walls of the cage or in the plastic mating jars were pupated. Trays were kept upright to allow frass to fall to marked with a date, counted, and left in situ until they the bottom of the container, to minimize disease hatched. Adult longevity was recorded and, on death, development. For the ‘hybrid’ broods, 300 neonates females were dissected, the presence of a spermatophore from each of the two crosses were placed into Petri noted, and eggs remaining in the ovarioles were counted. dishes for two days, with a maximum of 25 larvae per

Table 1 Mean (± SE) adult fecundity (number of eggs laid and retained in ovarioles) and fertility (number of eggs hatched) of ‘parental’, F1 ‘hybrid’ and backcross broods of P. insulata populations (both ungrouped and grouped) originating in Jamaica (J) and Florida (F) Mating pair N No. eggs laid No. eggs Fecundity No. eggs hatched N with % retained in (day 2 onwards) some hatched Day 2 onwards Day 1 (% of ovarioles hatching total laid)

F# 9 F$ 5 437.5 ± 43.4a 96.4 (18.1) 5.2 ± 2.0 539.2 324.6 ± 94.4a 479 J# 9 J$ 5 266.0 ± 66.5bc 0 24.4 ± 15.1 290.4 157.4 ± 86.5abc 359 F# 9 J$ (=FJ) 5 229.0 ± 37.0bc 33.0 (12.2) 30.0 ± 25.0 280.8 90.2 ± 35.2bc 445 J# 9 F$ (=JF) 5 238.0 ± 66.1bc 0 18.8 ± 11.4 268.0 151.8 ± 67.7abc 464 FJ# 9 F$ 5 353.8 ± 32.8 0 9.2 ± 5.6 363.0 276.0 ± 31.5 5 78 FJ$ 9 F# 5 190.0 ± 26.4 16 (9.6) 2.6 ± 0.8 169.8 121.0 ± 34.7 3 59 JF# 9 F$ 5 312.4 ± 42.1 0 3.6 ± 1.0 316.2 243.4 ± 47.4 5 78 JF$ 9 F# 5 362.2 ± 69.4 0 28.8 ± 12.7 391.4 241.6 ± 98.9 4 67 Backcross F 20 305.0 ± 26.3b 220.5 ± 30.5ab FJ# 9 J$ 5 236.6 ± 56.1 60.8 (20.4) 1.2 ± 0.7 298.6 173.4 ± 47.7 5 79 FJ$ 9 J# 5 200.0 ± 74.8 0 26.0 ± 22.9 226.0 105.8 ± 83.9 2 53 JF# 9 J$ 5 145.4 ± 17.8 35.0 (19.4) 4.4 ± 2.2 184.8 104.0 ± 15.7 5 89 JF$ 9 J# 5 222.2 ± 58.7 0 3.6 ± 2.2 225.8 100.2 ± 60.3 4 78 Backcross J 20 201.0 ± 26.9c 120.8 ± 27.3c Power of analysis 0.72 0.58 between F and J In a column, means followed by a common letter were not statistically different (P \ 0.05). Only those statistical comparisons conducted using grouped backcrosses are shown

123 Author's personal copy

Lack of genetic diversity in Pareuchaetes insulata 257

Petri dish. On the 3rd day after hatching, they were from 556 to 669 base pairs (bp). The ends of the transferred to 2 l plastic trays, at 25 larvae per tray, and alignment were trimmed to exclude those sites for when they reached the 4th instar each of these trays was which some specimens were missing data, yielding subdivided into two, with 12 and 13 larvae per tray. The 556 bp for which all specimens had data. Neighbour- eight back-crosses were treated similarly, but using 125 joining analysis using Kimura two-parameter on 18 neonates from each back-cross. sequences from the three regions yielded a phyloge- On the day of pupation, the pupae were collected, netic tree with two separate groups (Fig. 1). Sequences sexed and placed in 250 ml paper cups, capped to of specimens from Florida and Jamaica fell in group I, prevent adult escape, with 10 g of autoclaved vermic- and group II contained all sequences from Cuba. ulite to provide a substrate for the enclosing adults to Uncorrected pairwise sequence distances between the free themselves from the puparium. Cups were P. insulata populations ranged from 0 to 0.2 %. numbered and labelled with the population, date hatched, date pupated and tray number. The pupal Inter-population mating stadium and the percentage of pupae producing adults were recorded. Comparative fitness of the different Fecundity of adults and embryonic viability of eggs broods was determined by the following formula, derived from Maw’s (1976) host suitability index: The probability of detecting a difference that was statistically significant between the ‘parental’ popula- Population fitness Index tions of Florida and Jamaica was estimated, post-hoc, Mean mass of pupae % pupation on number of eggs laid (0.74), eggs found in ovarioles ¼ Mean development time to pupation (0.09) and embryonic viability (eggs hatched) (0.57). It was concluded that because of the small sample size Data were analysed using one-way analysis of and inherent variability, the null hypothesis could have variance (ANOVA) (Payne 2007). Where the probability been accepted in error. Consequently, the data from all from the overall F-test was declared significant four ‘hybrid’ broods backcrossed with the F ‘parent’ (P B 0.05), Fisher’s protected least significant differ- were grouped for ANOVA, as were the data from the ences were used to separate means and determine which four ‘hybrid’ broods backcrossed with the J ‘parent’ mean differences were statistically significant. Bartlett’s (Table 1). homogeneity of variance tests were performed prior to Although all females in this study laid eggs, only analysis. Where the underlying distribution of the variate 80 % of the batches hatched (Table 1). Eclosion in question was not normally distributed, and variances failure correlated closely with either a lack of a were heterogeneous, data were subjected to (x ? 1)-0.5 spermatophore, indicating that no mating had and arcsine transformations in an attempt to homogenize occurred, or the presence of a loaded spermatophore, variances. As the probability of the analysis to detect indicating incomplete fertilization. The number of significant differences seemed to be low in some of the eggs laid by the Floridian ‘parental’ brood was adult fecundity variates, and the null hypothesis (that statistically higher than both the Jamaican ‘parental’ there are no differences between any of the broods) could brood and the two ‘hybrid’ matings (F = 8.55, therefore be incorrectly accepted, the power of detecting 3,16 P = 0.010), but there were no significant differences differences between the two ‘parental’ populations was between the latter three crosses (F = 2.92, determined post-hoc. In order to increase sample size, 3,16 P = 0.07) (Table 1). The number of eggs laid (from where the probability of detection was low, data from day 2 on) among ‘parental’ populations, ‘hybrids’ and some broods were combined for ANOVA. backcrosses differed statistically (F11,48 = 2.59, P = 0.011). None of the ungrouped backcrosses Results displayed statistically different amounts of egg-laying to that of the Jamaican brood, although some were Molecular analysis significantly lower than that of the Florida brood (Fischer’s protected least significant difference test). DNA sequencing of the PCR products of P. insulata For the grouped backcrosses, the number of eggs laid specimens yielded 18 sequences of length varying (from day 2 on) among ‘parental’ populations, 123 Author's personal copy

258 N. Dube et al.

Fig. 1 Neighbour-joining tree showing the genetic relationship between 18 sequences of P. insulata from Florida, Cuba and Jamaica. The number above the node is a bootstrap support value (percentage), while each number following a specimen locality name indicates the individual insect specimen sequenced. The number in the legend is the mutation rate and the scale line above the number represents the equivalent distance in the tree to the mutation rate

‘hybrids’ and backcrosses differed statistically still only 0.58, but the null hypothesis was rejected,

(F5,54 = 3.78, P = 0.005). The grouped J backcross with significant differences between mean egg hatch- contained the lowest number of eggs laid, and this was ing in some broods (F5,54 = 2.52, P = 0.040). The significantly lower than numbers of eggs laid in the number of eggs hatched in the J backcross was grouped F backcross and the F ‘parent’ (Table 1). significantly lower than that for the F backcross and Although differences between brood means of the the F ‘parent’, even though the F ‘hybrid’ had the number of eggs retained in the females’ ovarioles [data lowest mean hatching rate (Table 1). transformed using (x ? 1)-0.5] seemed not to be statistically significant, the power analysis indicated Adult longevity that the probability of detection of a difference was only 0.3, even after the backcrosses were grouped. Under the assumption that the mating process did not Consequently, no valid statistical analysis was possi- affect longevity, data on adult longevity within each of ble (Table 1). the four broods that were used in the different mating Despite the significant differences in numbers of trials, viz. F, J, FJ and JF were pooled for analysis, eggs laid, no statistically significant differences in giving sample sizes of 20 for each gender of each embryonic viability were detected between any of the ‘parental’ brood and ten for each gender of each Floridian and Jamaican ‘parental’ broods, their ‘hybrid’ brood. Longevity of adults of the backcrossed ‘hybrid’ crosses or the ungrouped backcrosses broods was not determined. F males lived (F11,48 = 1.50, P = 0.164). The data were not signif- 6.60 ± 0.41 days (± SE), while the females lived icantly skewed (skewness 0.578, standard error of 8.05 ± 0.39 days. J males lived 7.65 ± 0.41 days and skewness 0.309). Bartlett’s test for homogeneity of females for 7.65 ± 0.39 days. FJ males lived for variance between broods was marginally not of 7.20 ± 0.58 days, while the females lived for 2 statistical significance (v11 ¼ 18:94; P ¼ 0:062). 7.20 ± 0.55 days, and finally JF males lived for After grouping the backcross data, the probability of 6.30 ± 0.58 days and females for 7.20 ± 0.55 days. detecting the difference of 167.2 between the ‘paren- No significant differences were observed between tal’ populations as being statistically significant was broods, within a sex of (Kruskal–Wallis test, 123 Author's personal copy

Lack of genetic diversity in Pareuchaetes insulata 259

2 2 v3 ¼ 5:13; P ¼ 0:162 for males; v3 ¼ 2:40; P ¼ Florida and Jamaica) have matching mtDNA frag- 0:493 for females). ments and are part of one ‘parental’ population. The observed genetic variation in this study is far less than that recorded between West African and East African Survival rates and fitness scores clades of the agricultural pest Busseola fusca (Fuller) (2 % divergence in the mitochondrial gene cyto- The Jamaican ‘parental’ brood and the backcross FJ$ chrome b) which is thought to be a main recent 9 F# showed significantly lower percentage pupation fragmentation event (Sezonlin et al. 2006). (28 and 44 % respectively) and eclosion rates (18 and 21 %) than most other broods, and also had the lowest Inter-population mating overall fitness (Table 2). Substantial variability was observed in the percentage pupation, percentage The ‘hybridization’ experiment in this study was eclosion and/or fitness within the ‘parental’ broods, conducted to provide information on the ability of the F ‘hybrids’ and some backcrosses, but no consistent 1 genetic material of the Floridian and Jamaican pop- trends of difference were detected between broods ulations of P. insulata to become integrated into a new (Table 2). With one exception (FJ$ 9 F#), the fitness fit ‘hybrid’ genotype and of the ‘hybrid’ genotype to scores in the F ‘hybrids’ and backcrosses lay between 1 be integrated back into the ‘parental’ types. Results of the fitness scores of the ‘parental’ broods. Arcsine the mating experiments demonstrated that full inte- transformation of these variates did not improve gration of the genetic materials from the two popula- variance homogeneity. It was concluded that ANOVA tions produced a genotype as fit as that of either of the was sufficiently robust to deal with these departures parents. There was no consistent reduction in embry- from normality. For fitness scores, although the onic viability in the backcrosses, as there was between overall F probability was significant (F = 9.47, 11,56 Eldana saccharina Walker (Lepidoptera: Pyralidae) P B 0.001), the heterogeneity of variance in the populations (Assefa et al. 2006b) or increase in Jamaica ‘parental’ population was too high for Fish- embryonic inviability, as there was between Pararge er’s protected least significant test to be performed. mergera L. (Lepidoptera: Nymphalidae) populations (Oliver 1972). However, there was a significant difference between the two ‘parental’ broods in the Discussion number of eggs laid. This difference was also reflected in the ‘hybrids’ and the backcrosses. However, the Molecular analysis observed variation has no demonstrable trend associ- ated with hybrid inviability, which is usually accom- This study shows that Floridian and Jamaican popu- panied by sterility and loss of ecological fitness lations of P. insulata have similar mitochondrial gene (Oliver 1979). The patterns of post-mating reproduc- sequences. The Cuban population of this moth differs tive isolation generally follow Haldane’s rule, where from these two populations only by one mutational the heterogametic sex is much more likely to become step (0.2 %). This variation observed between the inviable or sterile (Wu and Davis 1993; Assefa et al. Cuban and the other two populations is within the level 2006b). In genetically differentiated populations, of variation that is expected in a population of a there will be a highly significant difference in the lepidopteran species (Assefa et al. 2007). In a molec- survival of larvae between the pure and hybrid ular study on Busseola species (Noctuidae), Assefa populations (Oliver 1972). In this experiment, the et al. (2007) reported a within-species divergence in survival, fecundity and fitness of F1 ‘hybrids’ and COI gene to be between 0.0 and 0.77 %, whereas backcrosses were not statistically different from one of inter-species divergence was between 5.5 and 7.4 %. the parents or were intermediate between the two

In similar studies conducted by Cognato (2004) and parents. None of the backcrosses and/or F1 ‘hybrid’ Assefa et al. (2006a), the within-species variation in broods was significantly lower in fitness than both the COI region of the mitochondrial DNA ranged from ‘parental’ broods. Hence, the observed difference does 0 to 4.9 %. This study, therefore, shows that the P. not reflect the presence of hybrid disadvantage (Oliver insulata populations from the three regions (Cuba, 1972, 1979). The absence of hybrid inviability and 123 Author's personal copy

260 N. Dube et al.

Table 2 Pupal mass, developmental duration, survival rates and fitness scores of ‘parental’, F1 ‘hybrid’ and backcross broods of Pareuchaetes insulata populations originating in Jamaica (J) and Florida (F) Mating pair Pupal mass (g) Oviposition– % pupation % eclosion Fitness score eclosion (days) Female Male

F# 9 F$ 0.19 ± 0.02 (23)a 0.14 ± 0.02 (18)a 35 ± 2.08 (26)a 84 ± 4.00ab 61 ± 3.40abc 0.79 ± 0.02a J# 9 J$ 0.19 ± 0.05 (4)a 0.15 ± 0.02 (10)a 35 ± 1.53 (7)ab 28 ± 18.55d 18 ± 11.14e 0.35 ± 0.22f F# 9 J$ (=FJ) 0.20 ± 0.03 (72)a 0.15 ± 0.02 (73)a 42 ± 4.72 (93)b 55 ± 10.00c 49 ± 10.00bcd 0.53 ± 0.03def J# 9 F$ (=JF) 0.21 ± 0.03 (127)a 0.15 ± 0.02 (114)a 41 ± 5.65 (196)b 94 ± 3.00a 76 ± 5.00a 0.76 ± 0.02ab FJ# 9 F$ 0.19 ± 0.03 (42)a 0.14 ± 0.03 (46)a 40 ± 3.89 (73)b 83 ± 7.94ab 78 ± 5.40a 0.60 ± 0.06cd FJ$ 9 F# 0.18 ± 0.05 (19)a 0.16 ± 0.02 (17)a 46 ± 3.88 (11)b 44 ± 8.20cd 21 ± 9.68e 0.27 ± 0.05f JF# 9 F$ 0.19 ± 0.03 (58)a 0.13 ± 0.02 (34)a 38 ± 3.32 (66)ab 81 ± 3.31ab 85 ± 6.69a 0.67 ± 0.04bc JF$ 9 F# 0.21 ± 0.04 (59)a 0.16 ± 0.02 (40)a 37 ± 3.32 (79)ab 81 ± 5.16ab 68 ± 5.82ab 0.60 ± 0.11cd FJ# 9 J$ 0.19 ± 0.05 (59)a 0.15 ± 0.03 (58)a 37 ± 3.56 (90)b 78 ± 8.82ab 29 ± 9.44de 0.45 ± 0.05ef FJ$ 9 J# 0.17 ± 0.03 (35)a 0.13 ± 0.03 (37)a 38 ± 3.36 (24)b 100 ± 0.00a 69 ± 8.12ab 0.76 ± 0.06a JF# 9 J$ 0.19 ± 0.03 (28)a 0.14 ± 0.03 (29)a 43 ± 5.83 (28)ab 69 ± 8.82abc 36 ± 2.14cde 0.42 ± 0.05ef JF$ 9 J# 0.18 ± 0.04 (39)a 0.13 ± 0.03 (47)a 40 ± 3.48 (28)ab 83 ± 7.53ab 27 ± 5.87de 0.58 ± 0.06cde Sample size in parentheses. In a column, means followed by a common letter were not statistically different (P \ 0.05). Values presented are means ± SE

significant similarity in fecundity and fitness between et al. (1988), Lyla et al. (1998) and Orapa et al. (2002). the ‘parental’, ‘hybrid’ and backcross broods in this In many Erebidae, the female moths are selective with experiment is compatible with the mitochondrial DNA regards to mate choice (Conner 2009), a characteristic result. In the only other reported Pareuchaetes cross- that may go some way to explaining variability in breeding trial, Cock and Holloway (1982) reported mating in the laboratory. that cross-breeding trials of P. insulata material from Biological studies on the Floridian and Jamaican Yucatan, Mexico with that from Guanacaste, Costa populations were also conducted in 2006, prior to Rica, proceeded normally, whereas similar trials these trials (Dube 2008). The results indicated that between these two populations and P. pseudoinsulata several parameters, including the percentage pupation from Trinidad provided evidence for reproductive and eclosion as well as the population fitness index, barriers and thus support for their status as separate were higher for the Jamaican than the Floridian species. population, i.e. the converse of this study’s results. Variability in fecundity and fertility levels between However, both colonies were infected with a fungus pairs of moths within each mating category was high late in 2006. While the Floridian colony was recol- in our study, and non-hatching was closely correlated lected from the established South African field site, the with lack of mating or complete fertilization. High Jamaican colony was bred back from a single labo- variability in egg viability (11–100 %) was also ratory egg batch. The fitness of the latter colony could recorded in a laboratory colony of Pareuchaetes therefore be expected to have decreased in 2007 as a aurata aurata (Butler) (Lepidoptera: Erebidae) (Kluge result of inbreeding, which may explain the lower and Caldwell 1993b), while Torres et al. (1991) found pupation, eclosion and population fitness achieved by that infertility of egg masses of P. pseudoinsulata the Jamaican ‘parental’ colony in the current study. varied from 19 to 100 % in the laboratory, which they attributed to a skewed adult sex ratio. High variability Consequences for field populations in the numbers of eggs laid by P. pseudoinsulata in the laboratory, and in hatching rates, have also been The observations from the mating experiments are reported by Dharmadhikari et al. (1977), Napompeth generally similar to the results of the mitochondrial 123 Author's personal copy

Lack of genetic diversity in Pareuchaetes insulata 261

DNA analysis on the Jamaican and Floridian popula- Assefa Y, Conlong DE, Mitchell A (2007) DNA identification of tions of P. insulata. The low level of genetic Busseola (Lepidoptera: Noctuidae) larvae in Ethiopian sugarcane. Afr Entomol 15:375–379 differentiation among populations from the three Avise JC (2000) Phylogeography: the history and formation of regions and lack of reproductive isolation between species. Harvard University Press, Cambridge, USA Floridian and Jamaican populations thus suggests that Cock MJW, Holloway JD (1982) The history of, and prospects P. insulata populations released in South Africa are for, the biological control of Chromolaena odorata (Compositae) by Pareuchaetes pseudoinsulata Rego Bar- still parts of the same ‘parental’ population. Genetic ros and allies (Lepidoptera: Arctiidae). Bull Entomol Res incompatibility was, therefore, relatively unlikely to 72:193–205 be the cause of any failure in establishment of P. Cognato AI (2004) Standard percent sequence difference for insulata in the event that interbreeding of populations insects will not predict species boundaries. J Econ Entomol 97:1–17 occurred between or at any of the release sites. Conner WE (ed) (2009) Tiger moths and woolly bears: behavior, Ultimately, no establishment was recorded at any of ecology, and evolution of the Arctiidae. Oxford University the Cuban-population release sites, and the only way Press, New York, USA that Jamaican-population genes may have persisted is Dharmadhikari PR, Perera PACR, Hassen TMF (1977) The introduction of Ammalo insulata for the control of Eupa- through interbreeding between the Jamaican and the torium odoratum in Sri Lanka. Commonw Inst Biol Con- spreading, established Floridian population, at those trol Tech Bull 18:129–135 sites where Jamaican-population insects were still Dube N (2008) Investigation of the biology and cross breeding being released and to which the Floridian population of populations of Pareuchaetes insulata (Lepidoptera: Arctiidae) and the implications for the biological control of had already spread. Chromolaena odorata (Asteraceae) in South Africa. MSc On one hand, the results of this study caution that dissertation, University of KwaZulu-Natal, South Africa inbreeding of a laboratory culture may decrease its Futuyma DJ, Peterson SC (1985) Genetic variation in the use of fitness and that such a decrease may be transmitted to a resources by insects. Annu Rev Entomol 30:217–238 Goodall JM, Erasmus DJ (1996) Review of the status and inte- stronger culture, significantly weakening it too, but on grated control of the invasive alien weed, Chromolaena the other they suggest that crossing resulted in ‘genetic odorata, in South Africa. Agric Ecosyst Environ 6:151–164 rescue’ relative to the Jamaican population for some Hall TA (1999) BioEdit: a user-friendly biological sequence measures of fitness. However, since the majority of alignment editor and analysis program for Windows 95/98/ NT. Nucl Acid Symp Ser 41:95–98 Jamaican insects were released prior to 2007 (Zachari- Hoffmann JH, Impson FAC, Volchansky CR (2002) Biological ades et al. 2011), the low fitness recorded in this study control of cactus weeds: implications of hybridization did not play a major role in post-release population between control agent biotypes. J Appl Ecol 39:900–908 dynamics. Kluge RL, Caldwell PM (1993a) Host specificity of Pareu- chaetes insulata (Lep.: Arctiidae), a biological control agent for Chromolaena odorata (Compositae). Entom- Acknowledgments The molecular study was funded by the ophaga 38:451–457 Agricultural Research Council and Working for Water Kluge RL, Caldwell PM (1993b) The biology and host speci- Programme (South Africa), while the biology and ficity of Pareuchaetes aurata aurata (Lepidoptera: Arc- crossbreeding studies were conducted as part of an M.Sc. tiidae), a new association biological control agent for project by the senior author under the Professional Development Chromolaena odorata (Compositae). Bull Entomol Res Programme of the ARC, funded by the Department of Science 83:87–94 and Technology (South Africa). Mrs M. Whitwell (KZN Lyla KR, Abraham CC, Joy PJ (1998) Studies on fecundity and Department of Agriculture and Environmental Affairs) is fertility of Pareuchaetes pseudoinsulata Rego Barros thanked for assistance with statistical analyses. (Lepidoptera: Arctiidae). In: Ferrar P, Muniappan R, Jay- anth KP (eds) Proceedings of the fourth international workshop on the biological control and management of References Chromolaena odorata, vol 216. Agricultural Experiment Station, University of Guam, USA, pp 110–114 Assefa Y, Conlong DE, Mitchell A (2006a) Differences in Maw MG (1976) Biology of the tortoise beetle, Cassida mitochondrial DNA and fertility of crosses between pop- hemisphaerica (Coleoptera: Chrysomelidae), a possible ulations of Eldana saccharina (Lepidoptera: Pyralidae) biological control agent for bladder campion, Silene cu- from Kenya and South Africa: possible evidence for cubalus (Caryophyllaceae), in Canada. Can Entomol cryptic species? Sugar Cane Int 24(11–14):19 108:945–954 Assefa Y, Mitchell A, Conlong DE (2006b) Phylogeography of Mound LA, Wheeler GS, Williams DA (2010) Resolving Eldana saccharina Walker (Lepidoptera: Pyralidae). Ann cryptic species with morphology and DNA; thrips as a Soc Entomol France 42:331–338 potential biocontrol agent of Brazilian peppertree, with a

123 Author's personal copy

262 N. Dube et al.

new species and overview of Pseudophilothrips (Thysa- Wu CI, Davis AW (1993) Evolution of postmating reproductive noptera). Zootaxa 2432:59–68 isolation: the composite nature of Haldane’s rule and its Napompeth B, Thi Hai N, Winotai A (1988) Attempts on bio- genetic bases. Am Nat 142:187–212 logical control of Siam weed, Chromolaena odorata in Zachariades C, Day M, Muniappan R, Reddy GVP (2009) Thailand. In: Muniappan R (ed) Proceedings of the first Chromolaena odorata (L.) King and Robinson (Asteraceae). international workshop on biological control of Chromo- In: Muniappan R, Reddy GVP, Raman A (eds) Biological laena odorata. Agricultural Experiment Station, Univer- control of tropical weeds using . Cambridge sity of Guam, Mangilao, GU, USA, pp 57–62 University Press, Cambridge, USA, pp 130–162 Oliver CG (1972) Genetic differentiation between English and Zachariades C, Strathie L, Retief E, Dube N (2011) Progress French populations of the satyrid butterfly Pararge me- towards the biological control of Chromolaena odorata gera. Heredity 29:307–313 (L.) R.M.King & H.Rob. (Asteraceae) in South Africa. Afr Oliver CG (1979) Genetic differentiation and hybrid viability Entomol 19:282–302 within and between some Lepidoptera species. Am Nat 114:681–694 Author Biographies Orapa W, Bofeng I, Donnelly G (2002) Management of Chro- molaena odorata in Papua New Guinea: status of a bio- logical control programme. In: Zachariades C, Muniappan Nontembeko Dube joined ARC-PPRI as a student under the R, Strathie LW (eds) Proceedings of the fifth international Professional Development programme in 2006, and obtained workshop on biological control and management of an M.Sc. in Entomology from the University of KwaZulu- Chromolaena odorata. ARC-PPRI, Pretoria, South Africa, Natal in 2009. Prior to this she studied at the University of Fort pp 40–45 Hare where she obtained a B.Sc. in Entomology and Paterson ID, Zachariades C (2013) ISSRs indicate that Chro- Biochemistry in 2005, and a B.Sc. Hons in Entomology in molaena odorata invading southern Africa originates in 2006. Since 2009 she has been employed at ARC-PPRI as a Jamaica or Cuba. Biol Control 66:132–139 researcher on the biological control of Chromolaena odorata. Payne RW (ed) (2007) GenStat release 10 reference manual, part 3 procedure library PL18. VSN International, Hemel Yoseph Assefa is an Entomologist with a Ph.D. in molecular Hempstead, UK entomology, M.Sc. in crop protection and B.Sc. in crop Scheffer SJ (2000) Molecular evidence of cryptic species within production. He has research experience in pest management, Liriomyza huidobrensis (Diptera: Agromyzidae). J Econ insect ecology, insect biocontrol and insect molecular biology. Entomol 93:1146–1151 He taught entomology courses at universities in three African Sezonlin M, Dupas S, Le Ru¨ B, Gall P, Moyal P, Calatayud PA, countries for 12 years and he is currently serving University of Giffard I, Faure N, Silvain JF (2006) Phylogeography and Fort Hare in South Africa as an associate professor of population genetics of the maize stalk borer Busseola fusca Entomology. (Lepidoptera, Noctuidae) in sub-Saharan Africa. Mol Ecol 15:407–420 Costas Zachariades is a Senior Researcher at the Agricultural Simmons RB, Scheffer SJ (2004) Evidence of cryptic species Research Council, South Africa. He has been involved in within the pest Copitarsia decolora (Guene´e) (Lepidop- research on the biological control of Chromolaena odorata for tera: Noctuidae). Ann Entomol Soc Am 97:675–680 18 years, and is convenor of the IOBC working group on the Sperling FAH, Raske AG, Otvos IS (1999) Mitochondrial DNA biocontrol of this weed. sequence variation among populations and host races of Lambdina fiscellaria (Gn.) (Lepidoptera: Geometridae). Terence Olckers is a Senior Lecturer and entomologist at the Insect Mol Biol 8:97–106 University of KwaZulu-Natal, South Africa. His research Staden R (1996) The Staden sequence analysis package. Mol focuses on various aspects pertaining to the biological control Biotech 5:233–241 of invasive alien plants. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: molecular evolutionary genetics analysis (MEGA) soft- Des Conlong is senior entomologist at the South African ware version 4.0. Mol Biol Evol 24:1596–1599 Sugarcane Research Institute, where he has developed inte- Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins grated pest management strategies for sugarcane pests, based DG (1997) The ClustalX windows interface: flexible on biocontrol and habitat management options. He also heads strategies for multiple sequence alignment aided by quality the insect rearing programmes for sugarcane insect pests and analysis tools. Nucl Acids Res 24:4876–4882 their parasitoids, as well as for selected alien invasive weeds Torres DO, Vargas DG, Ritual SM, Alforja EM (1991) Studies and their biological control agents. He holds an honorary senior on causes of infertility of eggs of Pareuchaetes pseudoin- lecturer position in the School of Life Sciences at the sulata Rego Barros. In: Muniappan R, Ferrar P (eds) Pro- University of KwaZulu-Natal, and is a Professor Extraordinary ceedings of the second international workshop on in the Department of Conservation Ecology and Entomology, biological control of Chromolaena odorata. BIOTROP Stellenbosch University, both in South Africa. Special Publication 44, pp 113–119

123

View publication stats