Amynthas agrestis
(crazy worm)

Amynthas agrestis is an epigeic (litter-dwelling) Asian earthworm. It is native from Japan and the Korean Peninsula and has been introduced to the eastern United States, where it has spread widely, predominantly in forests. It has also been recorded from one location in Canada, near the USA border. Introduction and long-distance spread are thought to be associated with either transport of horticultural materials (for greenhouses, nurseries, and planting locations) or release of earthworms used as fishing bait.

The native range of A. agrestis includes the four main islands of Japan and the Korean Peninsula. Following introduction, it is now widespread in eastern North America, from Alabama and Georgia to New Hampshire and Vermont. Additional locations include northern Louisiana, eastern Texas, southeastern Oklahoma (Tandy, 1969; Gates, 1982; Damoff and Reynolds, 2009), one location in Wisconsin (Qiu and Turner, 2017), one in Ontario (Reynolds, 2014) and a greenhouse in Maine (Gates, 1966). While many of these localities consist of natural or managed habitats, such as arboreta (e.g., Gates, 1953), A. agrestis has been recorded in many human-created environments (e.g., culture beds, nurseries, compost heaps; Gates, 1954; 1958; 1963; Reynolds, 1977).

The routes of introduction/spread mentioned in the 'History of Introduction and Spread' section (horticultural materials and bait) continue to be relevant today and may unknowingly result in new invasions in remote or protected areas (e.g., Callaham et al., 2003; Snyder et al., 2011). The risk of introduction from horticultural materials may extend to any organic matter product associated with horticulture or landscaping, including soil, leaf litter, compost and mulch. Mulch (pieces of shredded wood up to 5 cm) is not an ideal habitat for this species, but survival in/under mulch has been documented and is linked to a high risk for transport and establishment of this species in gardens (Bellitürk et al., 2015).

In addition to local purchase for bait or composting, earthworms purported to be A. agrestis can also be ordered online (Ziemba et al., 2016). Earth-moving activities (e.g., road building) have been shown to act as pathways of introduction and spread for other invasive earthworms, being a potential route of transport for A. agrestis.

Genetics

A. agrestis genetics has been little studied, although interest in using molecular techniques to aid identification has been increasing. Schult et al. (2016) used two loci to examine the genetic divergence of Amynthas spp. and suggest that there may be cryptic speciation. However, morphological identification of some individuals by an expert was able to refer specimens in each lineage to a valid species, only by using species morphology.

Reproductive Biology

All earthworms are hermaphrodites and many, including A. agrestis, have developed parthenogenesis (Gates, 1958; Chang et al., 2016). Thus, most individuals encountered in the native or introduced range lack male pores and/or other parts of the male sexual system. Unlike many earthworms, A. agrestis has an annual life cycle (Tandy, 1969; Reynolds, 1978; Callaham et al., 2003; Chang et al., 2016). They overwinter as cocoons, hatch in spring, reach adulthood in summer, lay cocoons and all adults die by winter. A. agrestis lays spherical or subspherical cocoons of slightly under 2 mm diameter (Snyder et al., 2013). Cocoon colour is variable, from yellow to brown to red. Most cocoons contain one individual, but ~2% contain two individuals (Ikeda et al., 2015). Cocoon viability after 4 months has been shown to be between 60 and 70% (Ikeda et al., 2015). External sexual structures do not develop until near sexual maturity.

Physiology and Phenology

In Japan (Kanagawa Prefecture), A. agrestis juveniles were found in April-May and adults beginning in June (Uchida and Kaneko, 2004). In Louisiana, USA, juveniles were found April-June and adults were present in May-September (Tandy, 1969). In Tennessee, USA, juveniles or aclitellate adults were found May-June and adults were collected in June-December (Reynolds, 1978). Adults collected in September in Tennessee and North Carolina laid cocoons between September and November (Blackmon, 2009). At a high elevation site in north Georgia, USA (~1450m above sea level), juveniles were collected July-September, while adults were collected August-November (Callaham et al., 2003). Similarly, in Vermont, USA adults were observed July-November (Görres et al., 2014).

A. agrestis is known to bioaccumulate heavy metals, particularly Se, Zn, Cd and Hg (Richardson et al., 2015).

Longevity

A. agrestis has a life cycle lasting ~1 year. Little research has monitored the cocoon stage. Development from hatchling to adult is estimated to be a minimum of 90 days, based on frost-free periods of invaded locations (Görres and Melnichuk, 2012). Although, in the field, adults die in late fall or early winter, they have been kept alive in the laboratory through February (Snyder et al., 2013; Ikeda et al., 2015).

Nutrition

Soil and litter both play a large role in the biology of A. agrestis. Soil is a requirement for survival, weight maintenance and reproduction: in mesocosms with no soil (leaf litter only), A. agrestis lost weight and died, producing no cocoons (Ikeda et al., 2015). In this same study, A. agrestis with only soil survived, lost more weight than treatments with both litter and soil, and produced a limited number of cocoons. A. agrestis performs better on partially-decomposed leaf litter over fresh leaves, and appears to derive some nutrition from soil (Snyder et al., 2013; Ikeda et al., 2015). A combination of all three substrates provided the best nutrition (Ikeda et al., 2015). Zhang et al. (2010) found that A. agrestis feeds more on soil than Lumbricus rubellus (an epigeic European species also introduced in the USA) and, when needed, can shift its feeding strategy to more available soil or litter microbes.

A. agrestis can also survive in soil with mulch on top. Bellitürk et al. (2015) found that, in addition to survival and maturation over a 7 week incubation period with spruce, cedar, or pine mulch, A. agrestis increased lignin-digesting enzyme activity (phenoloxidase and peroxidase) in these soils. Mortality rates in this experiment were 10-30%, but there was no control against which to measure the effect of mulch on this variable.

Environmental Requirements

In the laboratory, adult A. agrestis did not survive at soil temperatures of -5, 5 or 35°C, but did survive at 12 and 25°C; moisture played a role, with no survival at 8% moisture and 25°C, but higher survival in wetter soils at the same temperature (Richardson et al., 2009). Field studies in Vermont indicate that 5°C may not be a minimum threshold temperature and that specimen size may play a role (Görres et al., 2014). Görres and Melnichuk (2012) suggest that the frost-free period (air temperature) may be an important predictor of A. agrestis tolerance of a site, because of time required to mature.

Temperature may play an important role for the development and hatching of cocoons. Blackmon (2009) examined whether cocoons would hatch at temperatures ranging from 5 to 30°C. Hatching rate was highest at 10°C and varied between 33, 60, and 100%, depending on the site where earthworms were collected. No hatching occurred at 5°C, and hatching occurred sooner at higher temperatures. Blackmon (2009) also tested whether changes in temperature (e.g., a winter cooling followed by spring warming) would affect hatching. Cocoons were lowered to 5, 10 or 15°C and incubated for 15, 30 or 60 days. Hatching occurred at all incubation lengths, but only at 10ºC, with the exception of one cocoon at 5°C. Cocoons also predominantly hatched 5-6 weeks after warming, regardless of incubation length.

The centipede Scolopocryptops sexspinosus and centipedes from the family Cryptopidae, rusty crayfish (Faxonius rusticus), wandering broadhead planarian (Bipalium adventitium), ribbon leech (Nephelopsis obscura) and salamanders Desmognathus monticola, Plethodon metcalfi and P. teyahalee have been documented to prey on A. agrestis (Craft, 2009; Gorsuch and Owen, 2014; Gao et al., 2017).

Animal feed, fodder, forage

• Bait/attractant

General

• Sport (hunting, shooting, fishing, racing)

A. agrestis is surface-dwelling and is best found by hand-searching through the leaf litter, at the soil surface, and under logs/rocks (Snyder et al., 2011). Decreased levels of partially-decomposed leaf litter and a granular surface soil may indicate an earthworm invasion, but not necessarily of this species. Wetter areas (e.g., along stream banks, valleys, deeper pockets of leaf litter) and areas impacted by human activity (e.g., roadsides, developed edges of forests) are more likely to support populations of A. agrestis.

A. agrestis can be easily confused with other pheretimoid earthworms. Amynthas alone includes hundreds of species. Sixteen species in Amynthas and closely related genera have been found in North America; Chang et al. (2016) provide a detailed key for these species. Most species can be easily differentiated from A. agrestis by size, colouration in life, and location of easily observable pores and markings. A. tokioensis and Metaphire hilgendorfi are quite similar to A. agrestis and the three species have been shown to coexist in some locations (Chang et al., 2016). Chang et al. (2016) provide a comparison table to help differentiate these species.

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Control

Physical/mechanical control

Blackmon (2009) and Ikeda et al. (2015) examined the role of fire in the control of A. agrestis invasion. Neither study found direct mortality, but both appear to have found indirect impacts: indirect/delayed mortality and weight loss, potentially due to starvation (Blackmon, 2009), and decreased cocoon viability (Ikeda et al., 2015).

Despite its wide dispersal across eastern North America over the last century, the biology and ecology of A. agrestis are very poorly known. The same is true for other invasive pheretimoid species, for which often no information exists beyond their taxonomic description (Burtelow et al., 1998). Understanding traits which aid A. agrestis’ dispersal, establishment and spread (e.g., life history traits, Callaham et al., 2003) is necessary, in order to fully understand and combat these invasions. Tracing pathways of invasion and quantifying their contribution to A. agrestis spread would also be very useful in slowing or stopping the spread of this species.

Earthworm invasions may be nearly impossible to eliminate without wholesale ecosystem destruction, but prescribed fire is one avenue that has shown promise and which needs further exploration.

More research is needed on the diversity of parthenogenetic morphs within the pheretimoids and the distribution of each invading species, as well as effective techniques to accurately identify species.

11/07/17 Original text by:

Bruce A. Snyder, Georgia College & State University, Milledgeville, GA, USA