Journal of Chemical Ecology

, Volume 38, Issue 6, pp 604–614

Foraging in the Dark – Chemically Mediated Host Plant Location by Belowground Insect Herbivores

Authors

    • Hawkesbury Institute for the EnvironmentUniversity of Western Sydney
  • Uffe N. Nielsen
    • Hawkesbury Institute for the EnvironmentUniversity of Western Sydney
    • School of Science and HealthUniversity of Western Sydney
Article

DOI: 10.1007/s10886-012-0106-x

Cite this article as:
Johnson, S.N. & Nielsen, U.N. J Chem Ecol (2012) 38: 604. doi:10.1007/s10886-012-0106-x

Abstract

Root-feeding insects are key components in many terrestrial ecosystems. Like shoot-feeding insect herbivores, they exploit a range of chemical cues to locate host plants. Respiratory emissions of carbon dioxide (CO2) from the roots is widely reported as the main attractant, however, there is conflicting evidence about its exact role. CO2 may act as a ‘search trigger’ causing insects to search more intensively for more host specific signals, or the plant may ‘mask’ CO2 emissions with other root volatiles thus avoiding detection. At least 74 other compounds elicit behavioral responses in root-feeding insects, with the majority (>80 %) causing attraction. Low molecular weight compounds (e.g., alcohols, esters, and aldehydes) underpin attraction, whereas hydrocarbons tend to have repellent properties. A range of compounds act as phagostimulants (e.g., sugars) once insects feed on roots, whereas secondary metabolites often deter feeding. In contrast, some secondary metabolites usually regarded as plant defenses (e.g., dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA)), can be exploited by some root-feeding insects for host location. Insects share several host location cues with plant parasitic nematodes (CO2, DIMBOA, glutamic acid), but some compounds (e.g., cucurbitacin A) repel nematodes while acting as phagostimulants to insects. Moreover, insect and nematode herbivory can induce exudation of compounds that may be mutually beneficial, suggesting potentially significant interactions between the two groups of herbivores. While a range of plant-derived chemicals can affect the behavior of root-feeding insects, little attempt has been made to exploit these in pest management, though this may become a more viable option with diminishing control options.

Keywords

InsectNematodeRoot exudatesRoot-feedersSoilPests

Herbivores in the Soil

Soil-dwelling herbivores comprise mammals (e.g., rodents) and invertebrates (e.g., insects and nematodes), which feed on a wide range of plant species and belowground plant structures (Andersen, 1987; Hunter, 2001; Johnson and Murray, 2008). While root-feeding insects are generally less well-studied than shoot herbivores (Hunter, 2001), it is widely recognized that they play pivotal roles in terrestrial ecosystems (Johnson and Murray, 2008). In agricultural systems, damage caused by belowground herbivores can be profound (Blackshaw and Kerry, 2008), with yield reductions of up to 60 % reported for the vine weevil (Otiorhynchus sulcatus) for example (Clark et al., 2012). Similarly, the ecological significance of soil-dwelling herbivores is apparent, with many studies illustrating how they can influence the community dynamics of plants (e.g., De Deyn et al., 2003), soil micro-organisms (e.g., Grayston et al., 2001), and aboveground herbivores (e.g., Johnson et al., 2009). In the context of abundance, it has been shown that root-xylem feeding cicadas in eastern deciduous forest of North America have the greatest collective biomass of any terrestrial animal in terms of biomass per unit area (Karban, 1980). Even in low diversity Australian pastures, it is common for the weight of sheep per acre to be exceeded by the weight of root-feeding insects in the soil (Britton, 1978). In this review, we are principally concerned with insect herbivores (see Rasmann et al., 2012, this issue, for discussion of nematodes), although we consider potential similarities and interactions between the two types of invertebrate herbivores.

Aims and Scope

Despite the importance of soil-dwelling herbivores, our understanding of the chemical ecology underpinning their interactions with host plants is limited compared to those herbivores feeding on aerial parts of the plant. In the first attempt to systematically address how root-feeding insects locate and select host plants by using chemicals in the rhizosphere and in root tissues, Johnson and Gregory (2006) collated information from 78 studies. In the current review, we aim to update this synthesis but more importantly identify new issues in the chemical ecology underpinning root location by soil-dwelling insects. Like the article by Johnson and Gregory (2006), we place some terms in inverted commas (e.g., ‘attractant’) because not all of the studies use the strict definition of the word. We focus on insect herbivores, since the chemical ecology of nematodes is covered by Rasmann et al. (2012, this issue), and we are unaware of any studies that address the chemical ecology of root herbivory by mammals.

Mechanisms of Root Location

The earliest studies addressing insect orientation to roots suggested that they encountered roots at random, and there was little scope for chemical mediation (Lees, 1943; Thorpe et al., 1946). Even though wireworms were ‘attracted’ to components of plant roots, Thorpe et al. (1946) argued that any orientation to roots was in response to changes in soil architecture (e.g., cracks) in the rhizosphere caused by root growth. Subsequent studies (Klingler, 1957, 1958; Doane et al., 1975) showed that wireworms orientated towards roots using respiratory emissions of carbon dioxide (CO2), and there has been growing evidence that root exudates are used by a range of herbivores to locate suitable host plants ever since (Bais et al., 2006; Johnson and Gregory, 2006; Wenke et al., 2010). In some cases, it has been established that volatile cues mediate ‘attraction’ to roots without necessarily identifying the chemical compounds involved. These include location by wireworms (Calkins et al., 1967; Horton and Landolt, 2002), the clover root weevil Sitona hispidulus (Wolfson, 1987), the wheat bulb fly (Delia coarctata) (Stokes, 1956; Long, 1958; Scott, 1974), and the grass grub (Costelytra zealandica) (Sutherland, 1972; Sutherland and Hillier, 1974b).

Respiratory Emissions of Carbon Dioxide

Respiratory emissions of carbon dioxide (CO2) remains the most widely reported root exudate implicated in the ‘attraction’ of a number of root herbivores (Table 1). CO2 is the most abundant gaseous exudate from roots and diffuses relatively rapidly in soil (Payne and Gregory, 1988). Several studies have reported that soil-dwelling insects are sensitive to even very small increases in CO2 concentrations; 0.02 mmol mol−1 for the wireworm Ctenicera destructor (Doane et al., 1975) and 0.03 mmol mol−1 for the vine weevil, O. sulcatus (Klingler, 1958). Conversely, very high concentrations of CO2 can ‘repel’ (Klingler, 1958) or become toxic to insects (Bernklau and Bjostad, 1998a). The ubiquitous nature of CO2, the stronger vertical gradients (between the air and the upper soil), and the high density of roots (>1 cm cm−3) (Gregory, 2006), led Johnson and Gregory (2006) to question whether this was an effective means for root-feeding insects to locate roots, particularly in mixed plant communities and when the herbivore specializes on particular plant species. Instead, they proposed that in some systems CO2 might act as a ‘search trigger’ causing insects to forage more intensely within a potential resource patch. This has since been supported empirically for the clover root weevil, Sitona lepidus (Johnson et al., 2006) and the cabbage root fly, Delia radicum (pers. obs.). Similarly, the European cockchafer, Melolontha melolontha, orientated within CO2 gradients to the source of the CO2, but this orientation disappeared when other plant-derived signals were present (Reinecke et al., 2008). Beyond this example, we have little understanding of how other root exudates (see section below) interact with respiratory emissions of CO2, but it seems highly likely that insect behavior will be moderated by the interplay of different signals. In any case, these studies lend support to the idea that CO2 emissions are supplemented by other chemical signals that may ‘attract’, ‘deter’, or even mask (proposed by Reinecke et al., 2008), any attraction to CO2 sources (Johnson et al., 2006).
Table 1

Soil insect herbivores showing behavioral responses to CO2, adapted from Johnson and Gregory (2006) with superscript numbers referring to adjacent references

Insect order

Insect species

Plant specificity

Dose–response

References

Diptera

Carrot root fly

s

dr1

(von Städler, 19712; Jones and Coaker, 1977, 1979) 1

Psilae rosae

nm2

Cabbage root fly

s

na

Jones and James (unpublished) cited in Jones and Coaker (1978)

Delia brassicae

Lepidoptera

Lesser cornstalk borer

s

dr

(Huang and Mack, 2001, 2002)

Elasmopalpus lignosellus

Coleoptera

Western corn rootworm

s

dr3

(Strnad et al., 1986) 3 (Strnad and Bergman, 1987a) 4 (Macdonald and Ellis, 1990) 4 (Macdonald and Ellis, 1990) 4 (Bernklau and Bjostad, 1998b, a) 2,3

Diabrotica virgifera virgifera

nm4

Grass grub

s

nm

(Galbreath, 1988)

Costelytra zealandica

Black vine weevil

g

nm

(Klingler, 1957, 1958, 1965, 1966)

Otiorhynchus sulcatus

Wireworms

g

nm

(Thorpe et al., 1946; Klingler, 1957, 1958, 1965, 1966)

Agriotes spp.

Ctenicera destructor

g

nm

(Doane et al., 1975)

Agriotes obscurus lineatus

g

nm

(Doane et al., 1975)

Limonius californicus

g

nm

(Doane et al., 1975)

Hypolithus bicolor

g

nm

(Doane et al., 1975)

Cockchafer

g

nm

(Klingler, 1957)

Melolontha vulgaris

European cockchafer

g

na

(Reinecke et al., 2008)

Melontha vulgaris

Southern corn rootworm

s

nm

(Jewett and Bjostad, 1996)

Diabrotica undecimpunctata

Ground beetle

g

dr

(Hamilton, 1917)

Evarthrus sodalis

Clover root weevil

s

dr

(Johnson et al., 2006)

Sitona lepidus

Plant specificity refers to the host-plant range of the insects; g = generalist feeders (polyphagous), s = specialist feeders (mono/oligophagous). Dose response refers to whether insects showed a dose-dependent response to CO2; dr = dose dependent, nm = not measured and na = information not available

Root Exudates Other than CO2

In their review, Johnson and Gregory (2006) listed around 60 compounds that potentially play a role in host plant location, revised and updated in Table 2. There are several additions, with approximately 74 different compounds now reported in the literature. The vast majority (>80 %) are regarded as ‘attractants,’ with the remaining compounds having either ‘repellent’ properties, or being both ‘attractive’ or ‘repellent’ depending on concentration. There is a trend for low molecular weight compounds (e.g., alcohols, esters, and aldehydes) to have ‘attractant’ properties, while hydrocarbons tend to be ‘repellent’. As noted by Johnson and Gregory (2006), only methyl eugenol and allyl-isothiocyanate were ‘attractive’ to more than two insect species (Table 2). Some studies (e.g., Finch and Skinner, 1974; Soni and Finch, 1979; Mochizuki et al., 1989; Weissteiner and Schutz, 2006) report chemical groups rather than specific compounds, so further breakdown of trends and patterns is difficult to achieve accurately.
Table 2

Chemical cues exuded by roots in the rhizosphere that enable root-feeding insects to locate host-plants, adapted from Johnson and Gregory (2006). Insect Orders are (D) Diptera and (C) Coleoptera. ‘Type’ describes the nature of the chemical cue; (es) esters, (ke) ketones, (ad) aldehydes, (ie) isothiocyanate, (ac) alcohols, (ca) carboxylic acids, (aa) amino acids, (mc) mercaptans, (hy) hydrocarbons, (is) isoflavonoids and (o) others. Effect refers to whether the chemical is an ‘attractant’ (+) or a ‘repellent’ (−)

Insect order

Insect species

Chemical compound

Type

Effect

Reference

D

Carrot root fly

methyl eugenol

o

+

(Jones and Coaker, 1977, 1979)

Psilae rosae

 

bornyl acetate

es

+

(Ryan and Guerin, 1982; Guerin and Ryan, 1984)

2,4-dimethyl styrene

hy

+

α-ionone

ke

+

β-ionone

ke

+

biphenyl

hy

+

falcarinol

o

+

(Maki et al., 1989; Maki and Ryan, 1989)

falcarindiol

o

+

falcarindiol monoacetate

o

+

trans-2-nonenal

ad

(Guerin and Ryan, 1984)

D

Cabbage root fly

isothiocyanatesa

ie

+/−

(Finch and Skinner, 1974)

Delia radicum

allyl isothiocyanate

ie

+

(Koštál, 1992; Ross and Anderson, 1992)

ethyl isothiocyanate

ie

+

n-dipropyl disulphide

o

+

(Ross and Anderson, 1992)

allyl alcohol

o

+

methyl eugenol

o

+

hexanol

ac

+

(Koštál, 1992)

hexanal

ad

+

 

cis-3-hexen-1-ol

ac

+

linalool

ac

+

hexylacetate

o

cis-3-hexenyl acetate

o

benzaldehyde

o

myrcene

hy

terpinene

hy

α-pinene

hy

limonene

hy

D

Onion root fly

n-propyl disulphide

o

+

(Matsumoto and Thorsteinson, 1968; Ross and Anderson, 1992)

Delia antiqua

methyl disulphide

o

+

n-propyl mercaptan

mc

+

ethyl acetate

es

+

(Ikeshoji et al., 1980)

tetramethylpyrazine

o

+

n-heptanal

ad

+

propanol

ac

+

(Mochizuki et al., 1989)

butanol

ac

+

pentanol

ac

+

hexanol

ac

+

heptanol

ac

+

pentanal

ad

+

hexanal

ad

+

heptanal

ad

+

valeric acid

ca

+

caproic acid

ca

+

enanthic acid

ca

+

21 estersb

es

+

allyl-isothiocyanate

ie

+

(Ross and Anderson, 1992)

2-phenyl ethanol

ac

n-dipropyl disulphide

o

+

sulphur compoundsc

o

+/−

(Soni and Finch, 1979)

ethyl sulphide

o

+

(Matsumoto, 1970)

n-butyl sulphide

o

+

iso-butyl sulphide

o

+

n-butyl methyl sulphide

o

+

n-butyl ethyl sulphide

o

+

iso-pentyl sulphide

o

+

allyl sulphide

o

+

n-propyl sulphide

o

+

D

Turnip root fly

allyl-isothiocyanate

ie

+

(Rygg and Sömme, 1972; Ross and Anderson, 1992)

Delia floralis

methyl eugenol

o

+

(Ross and Anderson, 1992)

n-dipropyl disulphide

o

+

allyl alcohol

o

+

phenylethyl-isothiocyanate

ie

(Rygg and Sömme, 1972)

C

Clover root borer

estragole

o

+

(Kamm and Buttery, 1984)

pentadecanal

ad

+

Hylastinus obscurus

hexadecanal

ad

+

hexanoic acid

ca

+

ethyl laurate

es

+

ethyl benzoate

es

+

  

E-2-hexenal

ad

+

(Tapia et al., 2007)

methyl benzoate

es

+

 

limonene

hy

C

Pine weevil

α-pinene

hy

+

(Nordenhem and Nordlander, 1994)

Hylobius abietis

ethanol

ac

+

C

Wireworms

ethyl acetate

es

+

(Morgan and Crumb, 1928)

Agriotes spp.

nitrobenzene

o

+

aspartic acid

aa

+

(Thorpe et al., 1946)

asparagine

aa

+

malic acid

ca

+

succinic acid

ca

+

glutamine

aa

+

glutamic acid

aa

+

C

Bark beetle

α-pinene

hy

+

(Rudinsky, 1966; Rudinsky and Zethner-Møller, 1967)

Hylastus nigrinus

β-pinene

hy

+

camphene

hy

+

C

Cockchafer

Monoterpenesd

hy

+

(Weissteiner and Schutz, 2006)

Melolontha hippocastani

C

Clover root weevil

Fomononetin

is

+

(Johnson et al., 2005)

Sitona lepidus

C

Western corn rootworm

1,4-benzoxazin-3-1 derivatives, including 2,4-dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA)

 

+

(Robert et al., 2012)

Diabrotica virgifera virgifera

aFinch and Skinner (1974) report several unspecified isothiocyanates that are either attractive or repellent to the cabbage root fly, Delia radicum

bMochizuki et al. (1989) list 21 closely related esters (not shown) that are attractive to the onion root fly, Delia antiqua. Esters with seven carbon atoms or less were attractive, those with eight or more were not

cSoni and Finch (1979) report 15 sulphur compounds that are either attractive or repellent to the onion root fly, Delia antiqua, depending on concentration. In addition to those described in Matsumoto and Thorsteinson (1968) that are present in onions (Allium cepa), Soni and Finch (1979) list sulphur compounds that are attractive to D. antiqua but are not known to be present in A. cepa

dWeissteiner and Schutz (2006) reported selective preference for carrot roots which primarily released monoterpenes, whereas less attractive oak roots emitted fatty acids. It was not established which compounds in the blend were attractive

The recent inclusion of 1,4-benzoxazin-3-one derivatives (Robert et al., 2012) and formononetin (Johnson et al., 2005) as being ‘attractive’ to western corn rootworm and the clover root weevil, respectively, is interesting as these compounds are usually regarded as having defensive properties in the plant. 2,4-Dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA), in particular, is regarded as having powerful insecticidal properties. Workers have found it challenging to identify chemicals underpinning host plant ‘attraction’ by the western corn rootworm; initial findings (Bjostad and Hibbard, 1992; Hibbard et al., 1994) were later revised (Bernklau and Bjostad, 1998b) or found to conflict with other studies (Xie et al., 1990, 1992). Nonetheless, given that many shoot-feeding insects have evolved the capacity to cope with plant defensive compounds and in some cases exploit them for host plant location (Bernays and Chapman, 1994), it seems likely that this occurs with soil-dwelling insects too. Moreover, Robert et al. (2012) also noted that root-feeding was positively correlated with concentrations of phenolic acids (e.g., chlorogenic acid), which also has been seen for the vine weevil (O. sulcatus) that feeds on the roots of raspberry (Clark et al., 2011) and blackcurrant (Johnson et al., 2011).

In addition to root exudates (usually volatiles) used by insect herbivores to locate and distinguish host plants from a distance, there is a whole range of contact chemosensory compounds that can either stimulate or deter feeding once initial herbivory has started (see Table 3 in Johnson and Gregory, 2006). To our knowledge, there have been no significant changes to this list, and the behavior of soil-dwelling herbivores seems generally similar to insect herbivores feeding aboveground (Chapman, 2003). For example, most of the ‘phagostimulants’ reported for root herbivores are sugars (50 %), whereas secondary compounds tend to ‘deter’ further feeding. The isoflavonoids, in particular, have been reported as having deterrent effects on generalist root herbivores (Sutherland et al., 1980; Russell et al., 1982; Lane et al., 1985; Gaynor et al., 1986). However, as discussed above, one isoflavonoid (formononetin) had ‘attractive’ properties for the specialist clover root weevil (Sitona lepidus) (Johnson et al., 2005). Given the strong affinity of this species with the root nodules of white clover (Trifolium repens) tissues (Gerard, 2001), which contain significant levels of formononetin (Mathesius, 2001), it seems likely that the clover root weevil may have overcome any deterrent effects. Moreover, formononetin may be relatively benign (and, therefore, easier to adapt to) given that Sutherland et al. (1980) reported that neither the grass grub (Costelytra zealandica) nor the African black beetle (Heteronychus arator F.) were negatively affected by this compound.

Comparing Chemical Cues Used by Insect and Nematode Herbivores

The chemical ecology of nematodes is reviewed by Rasmann et al. (2012, this issue), but here we consider whether insects and nematodes show any similarities in terms of chemically mediated orientation to roots. Like insect herbivores, plant parasitic nematodes have been found to use CO2 as a ‘long-distance’ kairomone for root location (Klingler, 1961; Prot, 1980). Theoretically this can act up to 1 m away for a single root and >2 m for a root mass (Dusenbery, 1987). Apart from CO2, our ability to compare insect and nematode responses to plant metabolites is limited to a few specific compounds. Specifically, DIMBOA and glutamic acid tend to ‘attract’ both nematodes (Riga et al., 1997; Friebe et al., 1998) and insects (see Table 2). Similarly, ascorbic acid ‘attracts’ nematodes (Bird, 1959, 1962) and acts as a ‘phagostimulant’ to the grass grub C. zealandica (Sutherland and Hillier, 1974a). By contrast, cucurbitacin A, a bitter triterpenoid compound from cucumber, ‘repels’ Meloidogyne incognita (Haynes and Jones, 1976; Chitwood, 2002) while it acts as a ‘phagostimulant’ to the curcurbit beetle (Eben et al., 1997). The response to these compounds is, however, rather idiosyncratic as highlighted by the study by Riga et al. (1997), which showed that male Globodera rostochiensis and G. pallida were ‘attracted’ to L-glutamic acid but not D-glutamic acid.

Is Insect Orientation Affected by Nematode Herbivory?

The role of micro-organisms in allelopathy is reviewed by Cipollini et al. (2012, this issue). However, considering the overlap in resources utilized by insect and nematode herbivores, the question arises: ‘Do plant parasitic nematodes influence the behavior of insect herbivores and vice versa?’ In particular, will feeding on roots by one group enhance or reduce the attractiveness of a host to the other group?

Despite the potentially destructive impact that arises from simultaneous attack by plant parasitic nematodes and insect herbivores (Blackshaw and Kerry, 2008), very few studies have investigated whether these two groups interact through chemical cues from plants. One of the few illustrative examples of an interaction between herbivore induced plant chemical responses and the two groups of herbivores is outlined in the paper by Ali et al. (2011). They showed how Citrus aurantium and Citrus paradisi × Ponsirus trifoliata, induce the production of Pregeijerene and Geijerene (a terpene and associated breakdown product) in response to herbivory by the root weevil Diaprepes abbreviatus. These compounds are known to ‘attract’ entomopathogenic nematodes, antagonists of the root weevil, but the authors found that they also ‘attract’ the plant parasitic nematode Tylenchulus semipenetrans, which is a major pest on citrus plants (Ali et al., 2011). In short, while the release of herbivore induced plant volatiles may help ameliorate the attack of one type of herbivore by signaling for its pathogens, it may induce the attack of another type of herbivore.

There is further evidence that suggests that the two groups of herbivores can potentially interact through plant responses to herbivory. In particular, juveniles of some nematode species, such as G. rostochiensis, show complete dependence on the presence of plant root exudates for hatching (Perry, 1997), and Heterodera schachtii second stage juveniles show oriented searching in the presence of root exudates (Clemens et al., 1994; Perry and Aumann, 1998). Hence, higher concentrations of exudates in the soil caused by insect herbivores may trigger the hatching of nematodes and increase the pathogen load on the plant. Similarly, plant parasitic nematodes may influence the release of plant derived organic compounds into the soil solution, and through this influence the orientation of insect herbivores. It has, for example, been shown that feeding on white clover roots by H. trifolii and other nematodes increases the amount of photosynthetically derived C in the microbial biomass (Yeates et al., 1999), and feeding by M. incognita increases concentrations of non-volatile water soluble 14C and several metal ions (Ca, Mg, Na, K, Fe, Cu) in root exudates from tomato plants (Van Gundy et al., 1977). This indicates that plant parasitic nematodes can increase the concentration of plant derived C in the soil solution, and it seems likely that insect herbivores would respond to the increased concentrations of root exudates caused by nematode herbivory. Whether an insect herbivore will be attracted or repelled through this will, however, depend on what other plant metabolites are being released. For instance, a study that investigated the influence of nematodes (not limited to plant parasitic nematodes) on Plantago lanceolata found that when nematodes were present in the soil P. lanceolata increased the concentration of aucubin and catalpol (iridoid glucosides) in the root exudates (Wurst et al., 2010). These two compounds are known to be broadly toxic to, or at least deter, generalist herbivores aboveground (Dobler et al., 2011), and it seems likely that a similar effect would be found for insect herbivores belowground. However, some specialist herbivores aboveground are known to sequester iridoid glucosides, thus reducing their palatability to predators (Dobler et al., 2011), and we cannot rule out a similar effect in belowground specialist insect herbivores.

The Rhizosphere and Chemical Signals

As discussed elsewhere in this issue (Effmert et al., 2012; Hartmann and Schikora, 2012; Hiltpold and Turlings, 2012; Jung et al., 2012) the rhizosphere represents a very different medium for chemical signaling than aerial parts of the plant. The condition of the soil in terms of porosity, moisture content, and bulk density will affect both the diffusion of chemicals and the behavior of insects. As Johnson and Gregory (2006) point out, a gaseous molecule can diffuse through 1 m of air more rapidly than through 1 mm film of water within a soil pore (Payne and Gregory, 1988), and increasing soil bulk density from 1.1 Mg m−3 to 1.5 Mg m−3, reduces mobility of western corn root worms by 90 % (Strnad and Bergman, 1987b). Based on mathematical models of belowground insect herbivore orientation to host plants (Zhang et al., 2006) and later (Zhang et al., 2007), Johnson and Gregory (2006) suggested a conceptual model for host plant location by such herbivores. Essentially, in the absence of relevant semiochemicals, insects move in a random manner (Zhang et al., 2006), but in the presence of generic signals (e.g., CO2) begin to search localized patches more intensively (e.g., Johnson et al., 2006). In the case of specialist feeders, attraction and orientated movement becomes evident with more specific chemical signals followed by feeding stimulation or deterrence at the root interface, determined by contact chemosensory signals.

Future Challenges and Conclusions

The recent observation that root-herbivores might be positively affected by some secondary compounds, and phenolics in particular (Clark et al., 2011; Johnson et al., 2011; Robert et al., 2012), highlights how we still know relatively little about the role of root defenses against root herbivory (comprehensively discussed by van Dam, 2009). Given that there are numerous examples of shoot herbivores adapting to, and even exploiting defensive chemicals for host plant selection (Bernays and Chapman, 1994), it seems intuitive that root herbivores should do the same. While phenolic compounds seldom have positive effects on shoot herbivores (for exceptions, see Bernays and Woodhead, 1982; Bernays et al., 1983), it remains possible that feeding on plant tissues with low nitrogen concentrations (usually the case in roots) causes herbivores to adaptively exploit phenolics for physiological development. For example, the grasshopper, Anacridium melanorhodon, conserves available nitrogen by using phenolics for cuticle sclerotization (Bernays and Woodhead, 1982). In attempting to better characterize how root herbivores respond to root chemistry, it may become clearer whether compounds regarded as having defensive roles against shoot herbivores have the same function belowground.

There also is a gap in our knowledge about the interactive effects between insect and nematode herbivores belowground, and a clear need for further investigation into their potential interactions. To promote this we need more studies that quantify the release of specific herbivore induced plant compounds in response to specific organisms, and whether the release of such compounds increase or decrease the attractiveness of a plant host to another type of herbivore. Such knowledge will perhaps enhance our capability to manage populations of soil-dwelling herbivore pests and thus secure optimal output.

In terms of the future challenge of global climate change, elevated air temperature and atmospheric CO2 concentrations are unlikely to have direct effects on signaling and root-feeding insects (Staley and Johnson, 2008). Predicted increases in the concentration of atmospheric CO2 will still be well below current concentrations in the soil (Payne and Gregory, 1988), so it seems unlikely that this will interfere with CO2 attractants. Likewise, the buffering effects of soil will minimize the effects of elevated air temperatures (Staley and Johnson, 2008), although it may exacerbate the effects of lower precipitation patterns which would increase soil porosity. Greater porosity significantly increases diffusion rates of gaseous root exudates, but impairs diffusion of chemicals in solution (Payne and Gregory, 1988), which may, therefore, alter chemical signaling between roots and herbivores. Indirect (i.e., plant-mediated) effects of climate change on root signaling with belowground herbivores may be envisaged. For example, it generally is thought that elevated CO2 promotes root biomass relative to shoot biomass (Rogers et al., 1994, 1996), so it might reasonably be expected that some phagostimulants and/or deterrents may become diluted in root tissues (but see Staley and Johnson, 2008 for exceptions) resulting in altered rates of root herbivory. In terms of the effects of elevated CO2 on semio-chemicals in the rhizosphere, this is likely to be highly system specific. For example, elevated atmospheric CO2 concentrations increased production of rhizobial root nodules in white clover (T. repens) with corresponding increases in clover root weevil (S. lepidus) populations and development rates (Johnson and McNicol, 2010). Given the attraction of S. lepidus to rhizobial root nodules (Gerard, 2001), it seems likely that greater numbers of nodules would increase overall concentrations of the chemical cues underpinning this attraction.

This review has shown the breadth of chemicals that elicit behavioral responses in root-feeding insects. To date, little attempt has been made to exploit this research to manage pest populations despite some tentative evidence of success (e.g., Bernklau et al., 2004). ‘Enemy recruitment chemicals’ that some insects elicit in their host plants (Rasmann et al., 2012, this issue), show potential application in pest management (Hiltpold and Turlings, 2012, this issue). We suggest that host plant location cues might also play a role in the management of subterranean insect pests, and represent an, as yet, untapped area of chemical ecology.

Acknowledgments

The authors are grateful for the invitation to write this article and to the two anonymous reviewers for their constructive comments and suggestions.

Copyright information

© Springer Science+Business Media, LLC 2012