Invasions of the Portuguese millipede, Ommatoiulus moreleti, in southern Australia

Invasions of the Portuguese millipede, Ommatoiulus moreleti, in southern Australia

Pedobiologia 56 (2013) 213–218 Contents lists available at ScienceDirect Pedobiologia - International Journal of Soil Biology journal homepage: www...

1MB Sizes 76 Downloads 105 Views

Pedobiologia 56 (2013) 213–218

Contents lists available at ScienceDirect

Pedobiologia - International Journal of Soil Biology journal homepage: www.elsevier.de/pedobi

Invasions of the Portuguese millipede, Ommatoiulus moreleti, in southern Australia Geoff H. Baker a,∗ , Linda Grevinga b , Natalie Banks a,c a b c

CSIRO Ecosystem Sciences, G.P.O. Box 1700, Canberra, ACT 2601, Australia J.F. Blumenbach Institute of Zoology and Anthropology, University of Goettingen, Berliner Str. 28, 37073 Goettingen, Germany Murdoch University, 90 South Street, Perth, WA 6150, Australia

a r t i c l e

i n f o

Article history: Received 13 May 2013 Received in revised form 21 July 2013 Accepted 26 August 2013 Keywords: Millipedes Ommatoiulus moreleti Southern Australia Invasions Predicted distribution

a b s t r a c t The Portuguese millipede, Ommatoiulus moreleti, was first recorded in Australia on Eyre Peninsula, South Australia in 1953. By the early 1980s, O. moreleti had spread to several sites across southern Australia where it had become a significant nuisance pest invading houses in large numbers in autumn and spring. At that time, it was predicted that O. moreleti would further expand its distribution to occupy most areas of southern Australia with mean annual rainfall of 300–2400 mm (excluding areas with summer rain maxima), mean daily minimum air temperatures in winter of 0–15 ◦ C and mean daily maximum temperatures in summer of 18–33 ◦ C, based on climatic matching of where it had already spread to. This paper reports the known distribution of O. moreleti in Australia in 2013 and confirms the earlier prediction. Whilst stadial age and body size (width) of O. moreleti now vary greatly between individual sites in Australia, geographic scale patterns in these traits are slight, with both age and size only weakly correlated with rainfall and temperature. Millipedes are smaller in Australia compared with Portugal. The importance of this size variation is discussed with reference to its likely influence on fecundity and the success of an introduced biological control agent (Pelidnoptera nigripennis: Diptera, Sciomyzidae). Crown Copyright © 2013 Published by Elsevier GmbH. All rights reserved.

Introduction The Portuguese millipede, Ommatoiulus moreleti (Lucas, 1860) (Diplopoda, Julidae) is an introduced pest in southern Australia (Baker 1978a, 1985a). It is primarily a nuisance pest, entering houses in large numbers especially when it is most active during autumn and, to a lesser extent, spring (Baker 1979a,b, 1988). The millipedes can be so abundant they can prevent trains from achieving traction on railway lines because of their squashed remains. In addition, O. moreleti can cause feeding damage to various horticultural and grain crops (e.g. strawberries, seedling canola). Dense populations of O. moreleti have been recorded in native woodlands and grasslands in southern Australia (Baker 1978b,c), but their impacts on native biota and soil functioning have been little studied (Griffin and Bull 1995). Current methods for controlling O. moreleti in southern Australia rely primarily on the use of pesticides or physical barriers to prevent entry to buildings (e.g. http://www.agric.wa.gov.au; http://www.pir.sa.gov.au/factsheets). Classical biological control has been tried (Baker 1984, 1985b,c; Bailey 1989), but with little success thus far. Most notably, much focus has been placed on

∗ Corresponding author. Tel.: +61 2 6246 4406; fax: +61 6246 4094. E-mail address: [email protected] (G.H. Baker).

the biological control potential of the fly parasitoid, Pelidnoptera nigripennis (Diptera: Sciomyzidae), the females of which lay eggs externally on the anterior segments of mostly adult O. moreleti in Portugal. The larvae of P. nigripennis hatch from the egg capsule and burrow into the host millipede through anterior inter-segmental membranes, where they further feed and pupate. Releases of the fly have been made in Australia in attempts to control O. moreleti, but the species has not established. Bailey (1989) suggested that host specimens in Australia were unsuitably small compared with millipedes in Portugal and that parasitoid eggs might thus fail to adhere. Baker (1985a) showed that, having invaded a new area, O. moreleti erupts in abundance for several years, and then subsequently declines to more modest numbers. The mechanism(s) driving this temporal pattern are not properly understood (Baker 1985a; McKillup et al. 1988; Bailey 1997), but recognition of the phenomenon reduced the perceived urgency for applied research to halt the problems experienced by urban residents afflicted by the millipedes. O. moreleti however continues to expand its range in Australia, with newly invaded land-holders and the environment experiencing the “boom and bust” dynamics of this invasive pest. O. moreleti was first recorded in Australia at Coffin Bay near Port Lincoln, Eyre Peninsula (South Australia) in 1953 (Baker 1978a). It was then found at Bridgewater, near Adelaide in 1964, in suburban Melbourne in 1965, and by the mid 1980s at opposite ends of the

0031-4056/$ – see front matter. Crown Copyright © 2013 Published by Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.pedobi.2013.08.002

214

G.H. Baker et al. / Pedobiologia 56 (2013) 213–218

country, e.g. in Perth, Western Australia and Canberra, Australian Capital Territory (Baker 1985a). The millipede colonises a wide variety of habitats including sand dunes, pine forests, grasslands, Eucalyptus woodlands, urban gardens, and agricultural fields. Such a distribution suggests that occurrence per se is not acutely sensitive to soil and vegetation type across Australian landscapes, and indeed similar broad habitat distributions have been noted within its native Portugal (Baker 1984). Bailey and de Mendonc¸a (1990) however suggested that the geographic boundaries of O. moreleti’s distribution on the Iberian Peninsula may be related to changes in litter quantity. The abundance of O. moreleti can vary markedly between habitat types (Baker 1978c, 1984; Bailey and de Mendonc¸a 1990). Potential distributions of invasive animals and plants are often predicted and there are many tools to enable this (Elith et al. 2006, 2010; Gallien et al. 2012), but relatively few such predictions are revisited many years later to evaluate the invading species’ progress and the accuracy of expectations (Jeshke, 2011). Soil biota have been particularly scarce in developing our understanding of invasion biology (Hendrix 2006; Hill et al. 2012). However, the known distribution of O. moreleti in the 1980s prompted Baker (1985a) to predict that it would eventually invade much of southern Australia, based on the climatic envelope of sites occupied already in Australia, plus those sites reported as locations for the species in Portugal and as an introduced species in the Portuguese territories of Madeira and the Azores. This climatic envelope included sites with a mean annual rainfall of 300–2400 mm (excluding areas with a summer maximum), mean daily minimum air temperatures in winter of 0–15 ◦ C, and mean daily maximum air temperatures in summer of 18–33 ◦ C. The present work reports the current known distribution, making use of new data collected during a national survey involving school children who are members of the CSIRO Double Helix Science Club (http://www.csiro.au/thehelix). These data were then mapped using facilities provided via the “Atlas of Living Australia” website (http://www.ala.org.au). As well as testing the original prediction for the distribution of O. moreleti in Australia, we specify and revise the expected future pattern.

Materials and methods Large-scale surveys involving school children, members of the CSIRO Double Helix Science Club, have been used previously in Australia to demonstrate the distributions of invertebrates (e.g. “Earthworms Downunder” reported by Baker et al. 1997). “Millipede Mayhem” was established along similar lines in 2007, in this case enabling collections at 454 sites. Through the club’s regular magazine (The Helix), children were asked to volunteer to collect a few millipedes, during autumn, from habitats near where they lived and send them (live by post, amongst loose, moist paper tissue in sealed plastic tubes and protected by padded bag, with brief habitat description and site address) to a central CSIRO collection point, where the specimens were identified. Collections were made between February and May, 2007. Sites were spread widely throughout the country (from Perth in the west to Sydney in the east; from Hobart in the south to Kuranda in the north), but understandably were concentrated mostly in the south-east of the country and along the eastern sea-board, where the majority of people live. Collectors were provided with a photograph of O. moreleti, and with instructions to look especially for it. Because of the pest’s notoriety in the media, especially in invaded areas, it is widely recognised by the public. Several native millipedes were however also collected, but data for such are not reported here. The sex, stadium and level of maturity of each of the millipedes collected during the “Millipede Mayhem” survey were determined. Stadium was measured using the ocular field method (Vachon

1947; Saudray 1952; Baker 1978b). Females (≥9th stadium) were dissected and the presence of mature eggs in the ovitubes was recorded (mature eggs are rarely found in younger stadia, Baker 1978b). Males (all stadia) were classified as juvenile, mature adult or intercalary adult, according to gonopod morphology (Baker 1978b). The width of the body at the 20th segment was measured, using an eye-piece graticule, for all individuals in ≥8th stadium. Climate data were accessed from the Australian Bureau of Meteorology web site, for stations as near as possible to the collection sites. Annual rainfall and summer (January) and winter (July) average daily maximum and minimum temperatures for the 2 years preceding autumn 2007 were selected as climatic specifications to reflect the prevailing growth conditions for millipedes in the 8th and older stadia collected live during the survey. A small number of location records for O. moreleti were already lodged in the “Atlas of Living Australia” (ALA) database prior to the present study. The distribution of these sites, along with those recorded in “Millipede Mayhem”, plus other published records (Baker 1978a, 1985a) and unpublished data for more recent years (G. Baker and colleagues: see acknowledgements) were visualised in ALA by importing a comma-separated values (csv) input file which comprised location, longitude, latitude, state and source of the records. Besides these site records, an environmental envelope was added to the map to show the predicted distribution of O. moreleti in Australia, based on climatic criteria (after Baker 1985a). The criteria were: annual mean rainfall of 300–2400 mm (ALA layer: bioclim bio12), excluding areas with summer rain maximum (ALA layer: srain1 ≤ 0.5), mean minimum temperature in the coldest period of the year of 0–15 ◦ C (ALA-layer: bioclim bio06) and mean maximum temperature of the warmest period of 18–33 ◦ C (ALA-layer: bioclim bio05). The map thus created (Fig. 1) closely matched that drawn by hand by Baker (1985a). In addition, the Predict option under the ALA Tools Menu with the best 5 independent terrestrial environmental layers [precipitation driest quarter [Bio17], precipitation seasonality [Bio15], radiation warmest quarter [Bio26], radiation seasonality [Bio23] and moisture index highest quarter mean [Bio32] was used to identify where O. moreleti could now occur based on location records for the millipede as known in January 2013]. The Predict Tool uses a method called “MaxEnt” for Maximum Entropy, which considers where species occur in terms of location and environment, and also where the species does not seem to occur. A model is built using the available species and environmental data and the result is essentially a probability surface showing where the species could occur (http://www.ala.org.au/spatial-portal-help/predict/). “MaxEnt” has been shown to have good predictive capability across a broad set of applications (Elith et al. 2010; Kearney et al. 2010; Hill et al. 2012). In some relatively small regions, such as near the millipede’s origins on southern Eyre Peninsula and in suburban Adelaide, there are many location data within short distances of each other, some of which are already published (e.g. see Figs. 2 and 4 in Baker 1978a). To facilitate resolution at the continental level (i.e. Australia as a whole) only a single record was logged in ALA for each country town or city suburb, irrespective of the number of actual occurrences recorded in the same locality. In total, 325 locations for O. moreleti were plotted using the ALA mapping tools. Various statistical tests (2 , Sign test, Mann–Whitney U Statistic, Pearson and Spearman Rank Correlations) were applied to the data, following methodologies and assumptions as per Statistix® (Analytical Software, 2000). Results and discussion O. moreleti has now been recorded throughout southern Australia, generally within the region predicted by Baker (1985a)

G.H. Baker et al. / Pedobiologia 56 (2013) 213–218

215

Fig. 1. Known locations of O. moreleti in Australia (䊉), as of January 2013, and the species’ predicted distribution (in red) as per Baker (1985a). The latter is plotted here as an environmental envelope which includes areas with annual mean rainfalls of 300–2400 mm (ALA layer: bioclim bio12) excluding areas with summer maximum (ALA layer: srain1 ≤ 0.5), mean minimum temperature in the coldest period of the year of 0–15 ◦ C (ALA layer: bioclim bio06) and mean maximum temperature in the warmest period of 18–33 ◦ C (ALA layer: bioclim bio05). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

(Fig. 1). There are, however, several areas, most notably in southwestern Western Australia, western Victoria and southern New South Wales, where O. moreleti has yet to be found, but where it was expected to occur according to Baker (1985a). Some parts of these latter areas remain predicted regions for the millipede’s eventual invasion, when the current locations of O. moreleti in Australia are used via the ALA Predict option to indicate climatic suitability (Fig. 2). The sites where O. moreleti was found during “Millipede Mayhem” had 176–1434 mm annual rainfall, −1.5 to 10.5 ◦ C mean daily minimum air temperature in winter and 18.7–36.5 ◦ C mean daily maximum air temperature, across both 2005 and 2006 (i.e. the 2 years prior to sampling, when the millipedes that were collected were alive). 2006 was unusually dry. All but one of the site records with an annual rainfall <300 mm occurred in 2006. The “Millipede Mayhem” survey recorded O. moreleti at 358 sites, but not at a further 96 sites (where only other species were found). The majority of sites with nil records for O. moreleti were in New South Wales and Queensland (84.4%). The furthest north on the east coast of Australia that O. moreleti was found was at Mount Victoria in the Blue Mountains to the west of Sydney. The millipede was recorded only twice in Sydney, in the suburbs of Oatley, near Botany Bay, and Lidcombe, near Paramatta. There were 77 nil records for sites in and north of Sydney, up to Kuranda in northern Queensland. A total of 4215 O. moreleti individuals were collected across all sites. Most (99.3%) were in stadium 8 and older. There were some small, geographic (state-based) shifts in the stadial age and size (width at the 20th segment) of O. moreleti, with for example, stadia 9 and 10 predominating in the south-east of Australia and

stadia 10 and 11 in the south and west (Table 1) (2 (based on the numbers of individuals in each stadium) = 454.7, P < 0.0001). There was no evidence of geographical differences in male sexual maturity. Virtually all males (96.6%, N = 1694) in the 8th or older stadia were mature adults; only very few were either juvenile (2.6%) or intercalary (0.8%). The majority of the females in the 9th and older stadia were gravid (overall 59.0%), with a slightly higher proportion of gravid females (71.4%) in Tasmania than in the other regions (53.1–66.5%). Whilst it is well accepted that the activity and breeding seasons of millipedes are influenced by prevailing weather (Hopkin and Read, 1992), information relating growth and length of life cycle in individual species to weather is scarce. Shifts in life cycle traits of some millipedes, including O. moreleti, with altitude (and hence climate) have been reported (Pedroli-Christen 1978; Read 1985; Baker and Baez 1989). Baker (1978c) argued that O. moreleti has a flexible life cycle, supported by features such as periodomorphosis (Baker 1978b), which enable extended adult life and the postponement of reproduction during unfavourable breeding seasons. But the ages (stadia) and body sizes described here for O. moreleti throughout southern Australia during autumn 2007 showed only limited evidence of variability across a wide range of rainfall and temperature scenarios. Millipedes tended to be smaller (body width, measured at the 20th segment) where rainfall was higher and where temperatures were warmer (Table 2), but such differences were only slight, where they were detected. In contrast, stadial age was slightly greater where temperatures were warmer, but again such a trend was weak (Table 3). Links between stadial age and rainfall were confusing with both negative and positive

216

G.H. Baker et al. / Pedobiologia 56 (2013) 213–218

Fig. 2. Predicted distribution of O. moreleti, generated in the “Atlas of Living Australia”, using the Predict option in the ALA Tools Menu, with the best 5 independent terrestrial environmental layers (precipitation driest quarter [Bio17], precipitation seasonality [Bio15], radiation warmest quarter [Bio26], radiation seasonality [Bio23] and moisture index highest quarter mean [Bio32], and based on location records for the millipede as known at January 2013). Regions with darkest colour represent highest likelihood of occurrence; yellow has least. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

Table 1 Proportions (%) of different (a) stadia (males and females combined) and (b) size (body widths at 20th segment; 8th stadium and older only) groups of O. moreleti collected regionally during “Millipede Mayhem”. Tas (Tasmania), Vic (Victoria), NSW (New South Wales), ACT (Australian Capital Territory), SA (South Australia), WA (West Australia). Stadia

(a) Region Tas. Vic. NSW and ACT SA WA

5

6

7

8

9

10

11

12

13

0.1

0.3

1.3 0.9

4.7 4.6 2.3 0.6 1.0

34.9 28.5 33.8 8.9 12.2

38.7 33.9 33.6 40.3 34.4

14.7 19.4 22.1 26.2 29.6

5.2 9.9 5.9 16.1 14.5

0.5 2.3 2.3 7.1 7.9

0.1 0.1

14

N

556 1700 390 845 724

0.1 0.7 0.3

Body width (mm) 1–1.50 (b) Females Tas. Vic. NSW and ACT SA WA Males Tas. Vic. NSW and ACT SA WA

1.51–2.00

2.01–2.50

2.51–3.00

3.01–3.50

3.51–4.00

0.3 1.1 1.3 0.2 –

5.2 11.3 10.5 3.6 3.6

43.3 35.9 38.0 22.2 34.9

35.2 34.8 30.8 46.1 40.3

14.4 15.0 18.6 25.4 19.3

1.6 1.9 0.8 2.5 1.9

3.2 6.1 14.2 2.4 3.8

68.9 62.4 68.9 32.5 65.4

26.3 30.9 16.9 60.8 30.1

1.6 0.6 – 4.3 0.7

– – – – –

– – – – –

G.H. Baker et al. / Pedobiologia 56 (2013) 213–218

217

Table 2 Significant Pearson Correlation Coefficients (r) between body sizes of O. moreleti (width in mm at 20th segment) and annual rainfall, average daily maximum air temperature (January), and average daily minimum air temperature (July) recorded within previous two years at closest meteorological station to site of millipede collection. Dash = not significant (P > 0.05). Data highlighted in bold had P < 0.01. Stadia

Average daily temperatures (◦ C)

Annual rainfall (mm)

January maxima

July minima

Year

2005

2006

2005

2006

2007

2005

2006

Females 9 10 11 12 Total

−0.134 −0.071 −0.139 −0.126 −0.053

– – −0.145 −0.123 –

−0.206 – −0.133 – –

−0.212 – – – –

−0.184 – −0.109 – –

0.158 0.091 – – 0.108

0.141 0.088 – – 0.123

Males 9 10 11 12 Total

– −0.139 −0.136 −0.254 −0.091

– −0.121 −0.174 – −0.082

−0.129 −0.104 – – –

−0.108 – – – –

−0.106 −0.107 – – –

0.159 0.140 0.173 – 0.209

0.134 0.144 0.174 – 0.218

Table 3 Significant Spearman Rank Correlation Coefficients (rho) between stadial age of O. moreleti and annual rainfall, average daily maximum air temperature (January), and average daily minimum air temperature (July) recorded within previous two years at closest meteorological station to site of millipede collection. Dash = not significant (P > 0.05). Data highlighted in bold had P < 0.01. Non-parametric correlation statistic calculated because of ordinal nature of stadial age data. Stadia

Average daily temperatures (◦ C)

Annual rainfall (mm)

January maxima

Females Males

2006

2005

2006

2007

2005

2006

– −0.062

0.053 –

0.135 0.111

0.055 –

0.062 0.054

0.152 0.229

0.185 0.247

correlations (very weak) recorded for males and females respectively (Table 3). The body sizes of the millipedes collected during “Millipede Mayhem” were smaller (when compared within individual stadia) than those collected by Baker (1984) from 220 sites throughout Portugal (Table 4) (using females as an example, Sign test using the 5 pairs of values for widths as per Table 4, P = 0.031; males showed similar trends – data for them not shown here). Baker (1984) also noted that the body widths of gravid females (stadia 9–13), collected at one site (Bridgewater, South Australia) during an autumn – early winter breeding season, were smaller than those collected similarly in Portugal (5 sites near Lisbon) (Australia: mean = 2.81 mm, n = 45; Portugal: mean = 3.74 mm, n = 71). Whilst the average body width of all female (≥8th stadium) O. moreleti collected in “Millipede Mayhem” was 2.66 mm (Fig. 3), the sizes at individual sites often varied greatly. For example, for the 20 adult females collected near Melrose, South Australia, the mean width was 3.24 ± 0.06 mm (¯x ± S.E.), whilst the mean width of 12 females at a site in Canberra, Australian Capital Territory, was much less (2.13 ± 0.13 mm) (Mann–Whitney U Statistic [Normal Approximation with Continuity Correction] = 4.51, P < 0.001). At the same sites, stadial age also differed (stadial age = 11.65 ± 0.18 and 9.58 ± 0.29 respectively; Mann–Whitney U = 4.04, P < 0.001). Such variation is not surprising, given the very different habitats sampled in

Table 4 Average body widths (mm) at the 20th segment for separate stadia of 2481 female O. moreleti collected from 358 sites across southern Australia (“Millipede Mayhem”) and 1162 females from 220 sites throughout Portugal (the latter data sourced from Baker, 1984). Stadia

Southern Australia Portugal

July minima

2005

8

9

10

11

12

13

1.84 2.15

2.20 2.68

2.60 3.22

2.86 3.80

3.08 4.25

3.27 N/A

Fig. 3. Frequency of female O. moreleti (stadia 8–14) with particular body widths (mm) at the 20th segment (N = 2491). Specimens were collected as part of the “Millipede Mayhem” citizen science project, conducted in locations across Australia.

“Millipede Mayhem”. Similar variability in body size between sites has been reported for collections of O. moreleti elsewhere, e.g. on Tenerife, Canary Islands (Baker and Baez 1989). In addition, Baker (1985a) demonstrated that the sizes of O. moreleti decreased with time since invasion at sites in South Australia, as population densities peaked and declined. Some of these observed variations in body size, although seemingly small, do however have relevance to the dynamics of O. moreleti’s abundance, and hence to pest control. The fecundity of O. moreleti increases with body size (Baker 1978b, 1984) and can be related as y = 240.5x − 535.6, where y = the estimated number of mature eggs carried by a female and x = its width in mm at the 20th segment. Fecundity can thus more than double with a shift of just 1 mm in body width, i.e. well within the broad range of sizes observed in the field. The rates of population

218

G.H. Baker et al. / Pedobiologia 56 (2013) 213–218

increase could thus be very variable between populations in Australia. Only the 8th and older stadia of O. moreleti are parasitised by P. nigripennis in Portugal (Baker 1985b; Bailey 1989) and in particular, <10% in the 10th and earlier stadia. This equates to parasitism of those millipedes that are <3.2 mm (approx.) in body width at the 20th segment. Bailey (1997) suggested that the failure of P. nigripennis to establish when released in South Australia to help control O. moreleti may have been due to the relatively small sizes of the local millipedes. The size distribution of O. moreleti collected during “Millipede Mayhem” lends support to this assertion (note male O. moreleti are smaller than females, thus exacerbating the issue). But the sizes of the millipedes where P. nigripennis was released in South Australia have not been published (nor the numbers of flies or the sites). Perhaps if sites had been selected where O. moreleti individuals were relatively large, establishment might have been achieved? Acknowledgements We especially thank the many members of the CSIRO Double Helix Science Club for their collecting efforts and staff in CSIRO Education, in particular Ross Kingsland, Liz Yuncken and Graham Walker, who contributed much to the running of “Millipede Mayhem”. Kim Perry and Bob Mesibov kindly provided several site records for O. moreleti in South Australia and Tasmania respectively that were hitherto unknown to the authors. Thanks also to Sarina Macfadyen and Dean Paini for comments on a draft manuscript. References Analytical Software, 2000. Statistix 7. User’s Manual. Analytical Software, Tallahassee, FL. Bailey, P.T., 1989. The millipede parasitoid Pelidnoptera nigripennis (F.) (Diptera: Sciomyzidae) for the biological control of the millipede Ommatoiulus moreleti (Lucas) (Diplopoda: Julida: Julidae) in Australia. Bull. Ent. Res. 79, 381–391. Bailey, P.T., 1997. Decline of an invading millipede, Ommatoiulus moreleti in South Australia: the need for a better understanding of the mechanism (Diplopoda, Julida: Julidae). Ent. Scand. Suppl. 51, 241–244. Bailey, P.T., de Mendonc¸a, T.R., 1990. The distribution of the millipede Ommatoiulus moreleti (Diplopoda, Julida: Julidae) in relation to other Ommatoiulus species on the south-western Iberian Peninsula. J. Zool., Lond. 221, 99–111. Baker, G.H., 1978a. The distribution and dispersal of the introduced millipede, Ommatoiulus moreletii (Diplopoda: Iulidae), in Australia. J. Zool., Lond. 185, 1–11. Baker, G.H., 1978b. The post-embryonic development and life history of the millipede, Ommatoiulus moreletii (Diplopoda: Iulidae), introduced in south-eastern Australia. J. Zool., Lond. 186, 209–228. Baker, G.H., 1978c. The population dynamics of the millipede, Ommatoiulus moreletii (Diplopoda: Iulidae). J. Zool., Lond. 186, 229–242. Baker, G.H., 1979a. The activity patterns of Ommatoiulus moreletii (Diplopoda: Iulidae) in South Australia. J. Zool., Lond. 187, 173–183.

Baker, G.H., 1979b. Eruptions of the introduced millipede, Ommatoiulus moreletii (Diplopoda, Iulidae), in Australia, with notes on the native Australiosoma castaneum (Diplopoda, Paradoxosomatidae). Sth Aust. Nat. 53, 36–41. Baker, G.H., 1984. Distribution, morphology and life history of the millipede Ommatoiulus moreletii (Diplopoda: Iulidae) in Portugal and comparisons with Australian populations. Aust. J. Zool. 32, 811–822. Baker, G.H., 1985a. The distribution and abundance of the Portuguese millipede Ommatoiulus moreletii (Diplopoda: Iulidae) in Australia. Aust. J. Ecol. 10, 249–259. Baker, G.H., 1985b. Parasites of the millipede Ommatoiulus moreletii (Lucas) (Diplopoda: Iulidae) in Portugal and their potential as biological control agents in Australia. Aust. J. Zool. 33, 23–32. Baker, G.H., 1985c. Predators of Ommatoiulus moreletii (Lucas) (Diplopoda: Iulidae) in Portugal and Australia. J. Aust. Ent. Soc. 24, 247–252. Baker, G.H., 1988. Patterns in the activity of the millipede Ommatoiulus moreleti (Diplopoda: Julidae). Aust. Ent. Mag. 15, 127–139. Baker, G.H., Baez, M., 1989. Life history and body size of the introduced millipede, Ommatoiulus moreleti (Lucas, 1860) (Diplopoda: Julidae), at different altitudes on Tenerife (Canary Islands). Rev. Ecol. Biol. Sol. 26, 473–489. Baker, G.H., Thumlert, T.A., Meisel, L.S., Carter, P.J., Kilpin, G.P., 1997. “Earthworms Downunder”: a survey of the earthworm fauna of urban and agricultural soils in Australia. Soil Biol. Biochem. 29, 589–597. Elith, J., Graham, C.H., Anderson, R.P., Dudi’k, M., et al., 2006. Novel methods improve prediction of species’ distributions from occurrence data. Ecography 29, 129–151. Elith, J., Kearney, M., Phillips, S., 2010. The art of modelling range-shifting species. Meth. Ecol. Evol. 1, 330–342. Gallien, L., Douzet, R., Pratte, S., Zimmermann, N.E., Thuiller, W., 2012. Invasive species distribution models – how violating the equilibrium assumption can create new insights. Glob. Ecol. Biogeogr. 21, 1126–1136. Griffin, T.T., Bull, C.M., 1995. Interactions between introduced and native millipede species in South Australia. Aust. J. Zool. 43, 129–140. Hendrix, P.F., 2006. Biological invasions belowground – earthworms as invasive species. Biol. Invas. 8, 1201–1204. Hill, M.P., Hoffman, A.A., Macfadyen, S., Umina, P.A., Elith, J., 2012. Understanding niche shifts: using current and historical data to model the invasive redlegged earth mite, Halotydeus destructor. Diversity Distrib. 18, 191–203. Hopkin, S.P., Read, H.J., 1992. The Biology of Millipedes. Oxford University Press, Oxford. Jeshke, J.M., 2011. Range modelling. In: Simberloff, D., Rejmánek, M. (Eds.), Encyclopedia of Biological Invasions. University of California Press, Berkeley, pp. 568–571. Kearney, M.R., Wintle, B.A., Porter, W.P., 2010. Correlative and mechanistic models of species distribution provide congruent forecasts under climate change. Conserv. Lett. 3, 203–213. McKillup, S.C., Allen, P.G., Skewes, M.A., 1988. The natural decline of an introduced species following its initial increase in abundance: an explanation for Ommatoiulus moreletii in Australia. Oecologia 77, 339–342. Pedroli-Christen, A., 1978. Contribution à la connaissance du développement postembryonnaire de Craspedosoma alemannicum (Verhoeff) et de Xylophageuma zschokkei Bigler (Diplopoda, Nematophora) dans une tourbière de Haut-Jura Suisse. Rev. Suisse Zool. 85, 673–679. Read, H.J., 1985. Stadial distributions of Ommatoiulus moreleti at different altitudes in Madeira with reference to life history phenomena (Diplopoda; Julidae). Bijdragen tot de Dierkunde 55, 177–180. Saudray, Y., 1952. Développement post-embryonnaire d’un Iulide indigene Cylindroiulus (Aneuloboiulus) silvarum Meinert. Archs Zool. Exp. Gen. 89, 1–14. Vachon, M., 1947. Contribution à l’ étude de développement post-embryonnaire de Pachybolus ligulatus Voges. Les étapes de la croissance. Annls Sci. Nat. (Zool.) 9, 109–121.