SEARCH

SEARCH BY CITATION

Keywords:

  • digestibility;
  • feeding preference;
  • herbivory;
  • Lolium perenne;
  • herbivore–performance

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Silica, deposited as opaline phytoliths in the leaves of grasses, constitutes 2–5% of dry leaf mass, yet its function remains unclear. It has been proposed that silica may act as an antiherbivore defence by increasing the abrasiveness and reducing the digestibility of grass leaves, although there is little direct experimental evidence to support this.
  • 2
    We investigated the effects of manipulated silica levels on the abrasiveness of the leaves of five grass species. We also examined the effects of silica levels on the feeding preferences, growth performance and digestion efficiency of two folivorous insects and one phloem-feeding insect.
  • 3
    Silica addition resulted in increases to leaf abrasiveness in four of the five grass species studied. Silica addition also deterred feeding by both folivores and reduced their growth rates and digestion efficiency.
  • 4
    These effects resulted in lower pupal mass of the lepidopteron larvae Spodoptera exempta and compensatory feeding by the orthopteran, Schistocerca gregaria. In contrast, silica had no effects on the feeding preference or the population growth of the phloem feeder, Sitobion avenae.
  • 5
    Our results demonstrate that silica is an effective defence against folivorous insects, both as a feeding deterrent, possibly mediated by increased abrasiveness, and as a digestibility reducer. The effects of silica on pupal mass and development time may impact on herbivore fitness and exposure to natural enemies.
  • 6
    These results are the first demonstration of a direct effect of silica on the abrasiveness of grasses and the adverse impact of silica on herbivore preference and performance.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Silica in the leaves of grasses can constitute 2–5% dry matter, 10–20 times higher than levels found typically in dicotyledonous plants (Russel 1961). Silica is stored primarily as opaline phytoliths in the epidermis (Parry & Smithson 1964; Kaufman et al. 1985). Accumulation of such high levels of a single mineral element within this plant family suggests that it has a functional significance. Suggested functions include silica as a waste product resulting from uncontrolled uptake at the roots (Ellis 1979), or that silica acts a mechanism of structural rigidity forming a metabolically inexpensive alternative to carbon-based support (Iler 1979; McNaughton et al. 1985). Cell wall silification and increases in silica deposition around damage sites have also been proposed as a mechanism to reduce fungal attack (Jones & Handreck 1967).

Another view is that silica in the tissues of grasses acts as an antiherbivore defence mechanism to reduce levels of grazing by both vertebrate (McNaughton & Tarrants 1983) and invertebrate herbivores (O’Reagain & Mentis 1989; Vicari & Bazely 1993). Silica is thought to increase the abrasiveness of plant tissues, causing increased tooth or mouthpart wear (Baker, Jones & Wardrop 1959), and hence acting as a feeding deterrent, but thus far the support for this is mainly correlative as it is drawn from the fossil record. The evolution of continuously growing teeth (among members of the Rodentia and Lagomorpha) and folded enamel layers of hypsodont (high-crowned) teeth (among the Ungulata) have both been linked to a grass diet (Janis & Fortelius 1988; Jernvall & Fortelius 2002). A similar link has been suggested between an abrasive grass diet, due to silica, and the development of enlarged mandibles in both the Orthopterans and Lepidopterans (Chapman 1964; Drave & Lauge 1978; Patterson 1983, 1984). Less attention has been given to a second potential mechanism by which silica could have adverse effects on herbivores. Silica within cell walls may reduce the digestibility of leaves by preventing access to carbohydrates and nitrogen during digestion (Van Soest & Jones 1968; Smith, Nelson & Boggino 1971), although this has not been demonstrated experimentally. There is some evidence that increases in silica concentration can deter feeding by some mammals (Gali-Muhtasib, Smith & Higgins 1992) and stem-boring insects (Djamin & Pathak 1967; Moore 1984), but again investigations into the effects of silica on herbivore growth performance remain limited. No previous studies have made a direct test of any mechanism by which silica may defend against herbivores, such as by increases in leaf abrasiveness (Hochuli 1993).

It has been proposed that due to the histological distribution of silica within leaves it might be a more effective defence against localized attack by insect herbivores than generalist grazing by ungulates (O’Reagain & Mentis 1989; Vicari & Bazely 1993). It is also likely that phytoliths deposited at intervals throughout the leaf epidermis would be encountered more frequently by leaf-chewing insects such as Orthopterans and Lepidopteran larvae, rather than phloem-feeding insects of the Hemiptera, which probe leaf veins and hence may be able to avoid the location of the phytoliths. However, to date there is a lack of experimental evidence on the impacts of silica on both the preference and performance of herbivores that are most likely to be affected (folivores), compared with those likely to be unaffected (phloem feeders). In addition, previous work has not measured the parameter by which silica is proposed to act, namely abrasion. Here we examine the effects of manipulating silica levels in four commonly occurring European perennial grass species (Poaceae): Agrostis capillaris L., Brachypodium pinnatum L., Festuca ovina L. and Lolium perenne L.; and one annual species: Poa annua L. on leaf abrasiveness and the feeding preference and growth performance of two generalist insect grass folivores: African army worm Spodoptera exempta Walker (Lepidoptera, Noctuidae) and desert locust Schistocerca gregaria Forskal (Orthoptera, Acrididae) and one phloem-feeding insect: grain aphid Sitobion avenae Fabricius (Hemiptera, Aphididae). We tested five hypotheses: (1) increasing the silica content of grass leaves will increase abrasiveness; (2) increasing the silica content of grasses will reduce the palatability to folivorous insects; (3) interspecific differences in grasses ability to uptake silica could lead to changes in relative feeding preference ranks depending upon silica availability; (4) higher silica content in grass leaves will reduce the growth and feeding efficiency of folivorous insects; and (5) silica will have no negative effects on the feeding preference or population growth of a phloem feeding insect herbivore.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

study species

S. exempta, S. gregaria and S. avenae are all considered to be grass-feeding generalist herbivores (Fraser & Grime 1999; Lee et al. 2003), although in S. gregaria feeding is not restricted to grasses. Both S. exempta and S. gregaria are found extensively throughout Africa as pest species on cereal crops and grasslands (Parker & Gatehouse 1985; Lee et al. 2003). S. exempta larvae used in this study came from a culture at the University of Lancaster; collected originally from Tanzania, and S. gregaria nymphs were supplied by a local pet shop. S. avenae are a native British herbivore found extensively on grasses, especially cereal crops (Watt 1979). S. avenae used in this study were from a single clone collected from a field site on University of Sussex campus, to reduce between-insect variability. All herbivores were reared on leaves and seedlings of Triticum aestivum before the experiments, ensuring that none of the herbivores had prior exposure to grass study species to avoid prejudicing results (Fraser & Grime 1999). Grass species were selected to be of varying palatability (Grime et al. 1996), growth rates and natural levels of silica (Table 1).

Table 1.  Growth rate, leaf characteristics, abrasiveness and chemical composition of grass study species under high and low silica treatments
SpeciesSilica treatmentSilica content (% DM)Abrasiveness (Rz in µm)Growth rate (mg DM d−1)Specific leaf area (cm2 mg−1 DM)Water content (%)Leaf nitrogen content (% DM)Carbon : nitrogen ratio
  • NS = not significant,

  • *

    P < 0·05,

  • **

    P < 0·01,

  • ***

    P < 0·001.

Agrostis capillaris Low0·46 ± 0·032·56 ± 0·1514·8 ± 1·70·29 ± 0·0269·2 ± 1·62·31 ± 0·2320·3 ± 1·7
High2·51 ± 0·143·38 ± 0·1614·1 ± 1·40·29 ± 0·0170·4 ± 1·22·24 ± 0·1219·3 ± 0·9
Brachypodium pinnatum Low0·47 ± 0·032·59 ± 0·10 2·5 ± 0·60·25 ± 0·0165·9 ± 1·91·27 ± 0·1238·0 ± 3·7
High2·87 ± 0·143·94 ± 0·16 2·1 ± 0·40·22 ± 0·0165·7 ± 1·11·49 ± 0·2033·6 ± 4·2
Festuca ovina Low0·52 ± 0·043·14 ± 0·12 4·8 ± 1·20·27 ± 0·0174·3 ± 4·52·18 ± 0·1821·3 ± 1·4
High2·44 ± 0·164·03 ± 0·13 4·3 ± 0·70·31 ± 0·0277·8 ± 3·22·18 ± 0·1822·3 ± 3·6
Lolium perenne Low0·54 ± 0·102·78 ± 0·2738·8 ± 4·00·25 ± 0·0174·0 ± 1·03·10 ± 0·1013·8 ± 0·5
High4·68 ± 0·343·95 ± 0·3033·8 ± 5·90·28 ± 0·0177·1 ± 1·63·12 ± 0·1013·2 ± 1·0
Poa annua Low0·81 ± 0·152·59 ± 0·1437·3 ± 5·40·35 ± 0·0176·9 ± 0·53·50 ± 0·2212·3 ± 1·1
High1·95 ± 0·222·95 ± 0·1428·7 ± 5·20·33 ± 0·0276·9 ± 0·83·28 ± 0·1612·6 ± 0·6
anova Species (S) *** *** *** *** *** *** ***
Treatment (T) *** *** NSNS * NSNS
S × T *** NSNS * NSNSNS

plant growth conditions

Grasses were grown under glasshouse conditions (15–25 °C, 16 : 8 light : dark). Grasses for insect preference trials were grown individually in 5 × 5 × 5 cm plugs of washed Perlite (an inert growth medium) for 12 weeks. Half the plants were watered every 3 days with 100 mL of Hoagland's solution; the other half received Hoagland's solution containing 150 mg/L of soluble silica in the form of NaSiO39H2O (Cid et al. 1990). For the insect performance trials and efficiency of food utilization tests, grasses were grown in 9-cm plant pots of washed Perlite (approx. 30 seeds per pot) for 12–15 weeks, with 100 mL Hoagland's solution, with or without 150 mg/L of silica, added every 3 days. All plants received tap water ad libitum.

chemical and physical analyses of foliage quality

Foliar silica content (n = 10 per silica treatment) was determined by fusing oven-dried leaf samples (approximately 0·2 g) in sodium hydroxide followed by analysis using the colorimetric silicomolybdate technique (Allen 1989). Foliar nitrogen and carbon content (n = 10 per grass species per silica treatment) was analysed using flash combustion of dried leaf samples (approximately 2·5 mg) followed by gas chromatographic separation (Elemental Combustion System; Costech Instruments, Milan, Italy) calibrated against a standard of composition C26H26N2O2S. Specific leaf area (ratio of leaf area to dry leaf mass) and leaf water content were measured for each species and silica treatment (n = 10).

Abrasiveness of grass samples (n = 10 per silica treatment) was determined using a development of the method described by Hammond & Ennos (2000). A layer of fresh grass leaves was mounted flat beneath the upper sample holder of a modified Martindale abrasion and pilling tester (Model 404, James H. Heal, Halifax, UK), giving an exposed leaf area of 5 cm2. The sample holder was attached to the top plate of the machine and a weight placed above the top plate applied a pressure on the grass sample of 12 kPa. The grass sample was rested on a 70 mm square block of Perspex, which was mounted flat on the modified abrading table of the machine. Therefore, the grass sample was sandwiched between the upper (mobile) sample holder and the Perspex. The machine was then set in motion, the top plate moving in a Lissajous figure of 24 mm stroke, causing the grass to rub against, and hence abrade, the Perspex. This process was continued until 200 rubs had been performed. The Perspex sheet was then removed and wear was measured by determining the roughness of the central part of each sheet using a laser perthometer. This scanned 15 mm across the plate four times and measured the average peak to valley heights for five successive sample lengths (Rz) on each scan, finally calculating a mean Rz. The more abrasive the grass, the rougher the Perspex plate would be after abrasion and, hence, the larger the value of Rz.

insect feeding preference

We conducted intraspecific paired feeding choice tests for all grass species between high and low silica treatments using three plants, which were matched for size, per treatment, per trial. The leaf area of each plant was scanned at the start of the trials (AM-200 leaf area meter, ADC) and the plants arranged in a grid design (3 × 2 individuals) within an insect cage. Sawdust was added to the bottom of the cage up to the base of the grasses, so insects could move freely between grasses. Two fourth-instar S. gregaria nymphs or S. exempta larvae were then placed randomly in the cage and left for 6–12 h at 25 °C, until approximately 50% of total leaf area was removed, after which time the remaining leaf area of each plant was measured. Ten replicate trials were completed for each grass species with each folivore.

Interspecific multiple preference trials on either high or low silica grasses of all five species were carried out in the same manner with two individuals of each grass species per trial, and 10 replicate trials with each folivore. For S. avenae feeding preference trials, only one plant per treatment or species was used in each trial. Twenty apterous, adult aphids were then added and the pot, grass and aphids covered with a muslin bag. After 24 h the number of individuals on each plant was recorded. Ten replicates of the intraspecific preference trials were carried out for each grass species and eight replicate trials for interspecific preference; however, due to its low palatability to aphids, intraspecific preference trials were not carried out using B. pinnatum.

insect growth performance

Ten second-instar nymphs of either S. gregaria or larvae of S. exempta were weighed then caged individually, under muslin netting, on a high or low silica treatment plant of each study species and kept at 25 °C throughout development, replacing grass when necessary. Larval mass of S. exempta was recorded every 3 days until pupation, at which point pupae were sexed and weighed. S. gregaria nymphal mass was recorded every 7 days over a 4-week period. Growth performance of herbivores was calculated as the relative growth rate [RGR = (change in biomass/time in days)/initial body mass] and for S. exempta only, as pupal mass and time taken to reach pupation. L. perenne was omitted from S. exempta growth performance trials due to high levels of mortality of larvae on this species, regardless of treatment. S. avenae growth performance was measured by caging 20 individuals on each of the grass species and silica treatments for 2 weeks (replicated 10 times), after which the number of individuals were counted. Performance was calculated as the relative population growth rate [RPGR = (change in aphid population/time in days)/initial number of aphids].

efficiency of food utilization

Individual third-instar S. gregaria nymphs or S. exempta larvae were starved for 12 h and weighed, before being placed in a container with a known weight of fresh grass leaf material (n = 10 per grass species per silica treatment, detailed above). Insects were kept at 21–25 °C and allowed to feed for 24 h, after which time they were starved for a further 12 h, to allow all the frass to be passed, before being reweighed. The remaining grass and frass was separated, dried at 60 °C for 48 h and weighed. Values of water content, derived from leaf samples of the same plants, were then used to convert the initial fresh mass of grass to dry mass. Food utilization efficiency measures were calculated according to Slansky (1985) and Raps & Vidal (1998):

  • • 
    Efficiency of conversion of ingested food (ECI), which estimates the percentage of food ingested converted to body mass, was calculated as: mass gained (mg change in fresh body mass)/food ingested (mg change in dry mass) × 100
  • • 
    Relative consumption (RC), which estimates the mass of food ingested over 24 h relative to initial body mass, was calculated as: food ingested (mg change in dry mass)/initial body mass (mg fresh mass).

statistical analysis

The leaf area removed by S. gregaria and S. exempta for each silica treatment in the intraspecific preference trials was calculated as a proportion of total leaf area removed per trial. Data were arcsine-square root-transformed before carrying out paired t-tests on trial means. For the interspecific preference trials, the leaf area removed per species was calculated as a proportion of total leaf area removed per trial and arcsine-square root-transformed before carrying out one-way anovas of trial means with Tukey's post-hoc analysis. For the preference trials with S. avenae, we compared the numbers of aphids per plant between treatments using paired t-tests in the intraspecific preference trials and the number of aphids between species using one-way anovas in the interspecific preference trials. The effects of grass species and silica addition on the foliar nitrogen and silica concentrations, leaf C : N ratio, leaf abrasiveness, specific leaf area, leaf water content and plant growth rates, and on the RGR, RPGR, ECI and RC for each herbivore species separately, were tested using general linear models. Development time to pupation and pupal mass of S. exempta from the performance trials were compared between silica treatments within each grass species and for each sex using two-sample t-tests.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

leaf silica content and abrasiveness

Silica content of grass leaves was up to seven times higher for those plants in the high silica treatment for all species (Table 1, anova, treatment, F1,88 = 499·3, P < 0·001). However, the amount of silica in grass leaves differed greatly between species, with P. annua the lowest and L. perenne the highest in the silica addition treatment (anova, species, F4,88 = 19·09, P < 0·001, species × treatment, F4,88 = 22·64, P < 0·001). The silica treatment also increased the abrasiveness of leaves for all species by between 28% and 52%, with the exception of P. annua (Table 1, anova, species, F4,88 = 6·13, P < 0·001, treatment, F1,88 = 63·34, P < 0·001, species × treatment, F4,88= 2·24, P = 0·071), which had the lowest levels of silica in leaves. Leaves had abrasiveness to some extent without the addition of silica, and this also varied between species (Table 1).

insect feeding preference

Both folivorous insects studied, S. exempta and S. gregaria, avoided plants to which silica had been added in the intraspecific preference tests (Fig. 1a,b). These plants contained higher levels of silica, but did not differ in terms of nitrogen content, C : N ratio, specific leaf area or water content (Table 1). The difference in relative proportions of leaf material eaten was greatest in those grass species that were able to take up the greatest levels of silica and had the greatest difference in abrasiveness between treatments, i.e. L. perenne and B. pinnatum. Silica addition to plants also altered the preference ranks between grass species for both S. exempta and S. gregaria (Fig. 2a–d).

image

Figure 1. Proportion of total leaf area removed by (a) S. exempta larvae and (b) S. gregaria nymphs, and (c) number of S. avenae on low (closed bars) and high (open bars) silica treatment plants of A. capillaris (Ac), B. pinnatum (Bp), F. ovina (Fo), L. perenne (Lp) and P. annua (Pa). Values are means of trial means from pair wise intraspecific preference tests ± SE. Degrees of significance from paired t-tests using trial means for each species are indicated as follows: ns = not significant, *P < 0·05, **P < 0·01, ***P < 0·001.

Download figure to PowerPoint

image

Figure 2. Proportion of total leaf area removed by (a, b) S. exempta larvae and (c, d) S. gregaria nymphs, and (e, f) number of S. avenae on each species of grass for low silica (a, c, e) and high silica treatment plants (b, d, f) for A. capillaris (Ac), B. pinnatum (Bp), F. ovina (Fo), L. perenne (Lp) and P. annua (Pa). Values are means of trial means from interspecific preference tests ± SE. Bars within each graph not sharing a common letter differ significantly (Tukey's test P < 0·05).

Download figure to PowerPoint

For three of the four species studied, silica addition to grasses had no effect on the feeding preference of S. avenae, but for L. perenne there was a 78% increase in feeding on high silica plants (Fig. 1c). Relative feeding preference ranks for S. avenae were unaffected by silica addition, with a clear preference for P. annua over L. perenne, both of which were preferred to A. capillaris, B. pinnatum and F. ovina regardless of the treatment (Fig. 2e,f).

insect performance

S. exempta displayed between 40% and 66% reduction in relative larval growth rate on high silica plants compared with low silica plants on all four species tested (Table 2, Fig. 3a). This resulted in longer development times for female larvae on A. capillaris and male larvae on F. ovina (Table 3), but no changes to development times were detected for larvae reared on the other grass species. The reduced growth rates on high silica plants also resulted in lower pupal mass for both sexes of S. exempta larvae, except for male larvae on F. ovina plants, which had a longer development time but no difference in pupal mass. Female larvae performed badly on B. pinnatum, which was a particularity unpalatable species regardless of silica treatment (Table 3, Fig. 3a). We found increases of 33% and 99% in the consumption rates of larvae on high silica plants of P. annua and L. perenne, respectively, but no difference for the other species (Table 2, Fig. 4a), and 84%, 68% and 51% reductions in the efficiency with which ingested material was converted to body mass (ECI) on B. pinnatum, F. ovina and L. perenne, respectively (Table 2, Fig. 4c).

Table 2.  General linear model (GLM) analyses on the effects of grass species (A. capillaris, B. pinnatum, F. ovina, L. perenne and P. annua) and treatment (high and low silica) on herbivore performance measurements (RGR: relative growth rate, RPGR: relative population growth rate, ECI: efficiency of conversion of ingested material, RC: relative consumption) of S. exempta larvae, S. gregaria nymphs and S. avenae.
HerbivoreFactorRGR or RPGRECIRC
d.f. F P d.f. F P d.f. F P
  • RGR analysis excludes L. perenne.

S. exempta Species (S) 3 5·07     0·003  417·36 < 0·001  4 9·78 < 0·001
Treatment (T) 128·56 < 0·001  115·65 < 0·001  1 1·90    0·171
S × T 3 1·99    0·125 4 1·72    0·152 4 2·39    0·056
Error59  90  90  
S. gregaria Species (S) 466·00 < 0·001  411·92 < 0·001  4 8·37 < 0·001
Treatment (T) 127·32 < 0·001  117·97 < 0·001  185·84 < 0·001
S × T 4 4·36    0·002 4 2·94    0·025 4 3·15    0·018
Error90  88  88  
S. avenae Species (S) 441·22 < 0·001       
Treatment (T) 1 0·39    0·536      
S × T 4 0·58    0·678      
Error90        
image

Figure 3. Relative growth rate of (a) S. exempta larvae and (b) S. gregaria nymphs, and population relative growth rate of (c) S. avenae on low (closed bars) and high (open bars) silica treatment plants of A. capillaris (Ac), B. pinnatum (Bp), F. ovina (Fo), L. perenne (Lp) and P. annua (Pa). Values are means ± SE.

Download figure to PowerPoint

Table 3.  Performance of male and female S. exempta larvae in terms of development time and pupal mass reared on plants of A. capillaris (Ac), B. pinnatum (Bp), F. ovina (Fo) and P. annua (Pa) with high and low silica levels. Means ± SE and results of two-sample t-tests are given, with numbers in brackets indicating replicates
Grass speciesHerbivore sexDevelopment time (days to pupation)Pupal mass (mg)
Low silicaHigh silica P Low silicaHigh silica P
AcFemale22·7 ± 0·3 (4)27·3 ± 0·3 (3) < 0·001 184·1 ± 1·2 (4)124·0 ± 0·7 (3)    0·018
AcMale23·0 ± 0·9 (6)25·0 ± 1·2 (4)    0·220151·4 ± 1·0 (6)111·1 ± 1·0 (4)    0·026
BpFemale21·9 ± 0·4 (6)24·0 ± 1·0 (4)    0·118190·6 ± 7·8 (6)165·3 ± 7·6 (4)    0·049
BpMale21·7 ± 0·7 (3)23·8 ± 0·8 (5)    0·096160·9 ± 4·8 (3)145·3 ± 3·7 (5)   0·062
FoFemale23·7 ± 1·8 (3)24·0 ± 1·0 (4)    0·880208·4 ± 11·3 (3)203·3 ± 21·0 (4)   0·840
FoMale20·6 ± 1·0 (5)25·0 ± 1·2 (4)     0·027 223·0 ± 4·3 (5)180·2 ± 12·2 (4)    0·042
PaFemale17·8 ± 0·8 (5)19·1 ± 2·3 (7)    0·289244·4 ± 19·0 (5)159·7 ± 18·1 (7)    0·010
PaMale19·4 ± 1·0 (5)18·3 ± 1·3 (3)    0·554202·4 ± 11·0 (5)112·7 ± 6·7 (3) < 0·001
image

Figure 4. (a, b) Relative consumption of grass [= food ingested (mg change in dry mass)/initial body mass (mg fresh mass)], and (c, d) efficiency of conversion of ingested food [ECI = mg change in fresh body mass/food ingested (mg change in dry mass) × 100] for S. exempta larvae (a, c) and S. gregaria nymphs (b, d) fed on low (closed bars), and high (open bars) silica leaves of A. capillaris (Ac), B. pinnatum (Bp), F. ovina (Fo), L. perenne (Lp) and P. annua (Pa). Values are means ± SE.

Download figure to PowerPoint

The relative growth rate of S. gregaria nymphs was between 17% and 33% lower on high silica grasses compared with low silica grasses (Table 2, Fig. 3b). S. gregaria nymphs increased consumption rates by 149%, 78% and 77% when fed on high silica plants for A. capillaris, B. pinnatum and L. perenne, respectively, and although the same trend can be seen in F. ovina and P. annua, these differences were found to be non-significant (Table 2, Fig. 4b). The efficiency by which S. gregaria nymphs were able to convert leaf mass to body mass was reduced by 62%, 52% and 69% when reared on high silica plants of A. capillaris, F. ovina and L. perenne, respectively (Table 2, Fig. 4d). No differences in digestability were found with S. gregaria on P. annua, which was the most palatable species for this herbivore and which does not uptake large quantities of silica, and for B. pinnatum, which had very low palatability on both high and low silica treatments.

Population growth of S. avenae was unaffected by silica treatments, but differed markedly between grass species, with P. annua sustaining much greater population growth rates than any other species tested (Table 2, Fig. 3c).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Our study demonstrates that silica in the leaves of grasses resulted in increased leaf abrasiveness. Silica addition also acted as a feeding deterrent and reduced the growth performance of both folivorous insects studied, but we found no detrimental effects of silica on the feeding preference or population growth of this phloem-feeding species. Increased levels of silica deterred feeding by both S. gregaria and S. exempta on all the five grass species studied, and also changed the palatability ranks between them. The change in palatability ranks and the degree to which silica affected palatability within species reflected the amount of silica taken up into the leaves, and the subsequent increase in abrasiveness.

In addition to reducing the palatability of grasses, we have demonstrated that increases in the levels of silica in leaves of host plants can reduce the growth of folivorous insects. Larval growth rates of S. exempta were reduced on each of the four grasses studied, resulting in increased development times, potentially increasing exposure time to predators and other adverse consequences (Clancy & Price 1987; Williams 1999). Possibly more significant in terms of herbivore populations, however, there was also a decrease in the pupal mass, especially in female pupae. Lower pupal mass has been shown previously to affect both the fecundity and dispersal distances of emerging adults, therefore reducing the fitness of individuals (Woodrow, Gatehouse & Davies 1987). S. exempta also showed a reduction in digestion efficiency on high silica plants of L. perenne, F. ovina and B. pinnatum, but did not alter the rate of consumption on plants with different silica content. It has been reported previously that more specialist herbivores, such as S. exempta, have limited ability to increase consumption to compensate for feeding on poorer quality host plants (Lee et al. 2003), and this could have resulted in the greater impact of silica on its performance compared with S. gregaria.

The levels of silica in host plants also affected the growth rate of S. gregaria nymphs, although not to the same extent as with S. exempta; the relative growth rates of nymphs on the most unpalatable species were not significantly different between silica treatments. This may be explained partly by S. gregaria's ability to compensate partially for poor quality food by increasing consumption rates (Lee et al. 2003; Raubenheimer & Simpson 2003). On all species studied S. gregaria fed at a higher rate on high silica plants. However, on only three species, F. ovina, L. perenne and P. annua, did we find lower efficiency of conversion of ingested material between treatments. The reduced growth rates and therefore longer development times, as well as increases in the time spent feeding to compensate for poor quality food, would be expected to increase exposure time of S. gregaria to predation (Clancy & Price 1987; Williams 1999).

In contrast to both the folivores, we found no detrimental effects of silica in the leaves of grasses on the feeding preference or population growth performance of the phloem-feeding insect S. avenae. As silica increases the abrasiveness of grass leaves and has been suggested to reduce the lignin content of leaves (McNaughton et al. 1985), it is possible that the method of feeding by S. avenae is unaffected by silica. Inserting stylets into the leaf veins would be unimpeded by small silica bodies that are isolated in the epidermis. Aphid growth performance and feeding preference were higher on grass species with higher growth rates and foliar nitrogen content.

This is the first study, of which we are aware, that has manipulated successfully the silica concentrations in the leaves of a range of grass species and demonstrated directly that increased silica leads to increased abrasiveness. Hence, our study contributes to our understanding of grass–grazer coevolution (McNaughton & Tarrants 1983; Herrera 1985). In addition, we have shown that silica deters feeding and also leads to changes in the relative feeding preference ranks between different grass species. Furthermore, silica significantly reduces the growth rates, digestion efficiency and pupal mass of insect folivores. Finally, we have found no evidence of silica affecting the growth performance or acting as a feeding deterrent on our chosen phloem feeder, although previous work has suggested that silica may have detrimental affects on xylem feeders (Kim & Heinrichs 1982). Silica content in grasses is spatially heterogeneous due to many biotic and environmental variables, including grazing history (McNaughton & Tarrants 1983; Brizuela, Delting & Cid 1986), soil pH and substrate type (O’Reagain & Mentis 1989). This variation in silica content in grass leaves creates in turn spatial variation in the interactions between grass and herbivores. For example, the relative palatabilities of L. perenne and P. annua may differ in low and high silica environments. The effects of silica on the growth rate and feeding efficiencies of folivores will impact on future fecundity and exposure to predators. The effects that silica in grass leaves has on herbivore growth, development and feeding efficiency have largely been neglected when studying the role of silica as a defence. Previous studies have focused on herbivore preference and have not examined either abrasiveness or performance indicators, and so may have underestimated the defensive properties of different grass species.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors would like to thank Andrew Dean for developing and conducting the silica digestion technique, Sam Valbonesi for C/N-analyses, University of Manchester School of Dentistry for use of their laser perthometer and Libby John, Alan Stewart, Kate Williamson and two anonymous reviewers for helpful comments on earlier drafts of the manuscript. The work was supported by the Natural Environment Research Council, UK (NER/A/S/2001/01144).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Allen, S.E. (1989) Chemical Analysis of Ecological Materials, 2nd edn. Blackwell Press, London.
  • Baker, G., Jones, L.H.P. & Wardrop, I.D. (1959) Cause of wear in sheep's teeth. Nature, 184, 15831584.
  • Brizuela, M.A., Detling, J.K. & Cid, M.S. (1986) Silicon concentration of grasses growing in sites with different grazing histories. Ecology, 67, 10981101.
  • Chapman, R.F. (1964) The structure and wear of the mandibles in some African grasshoppers. Proceedings of the Zoological Society of London, 142, 107121.
  • Cid, M.S., Detling, J.K., Whicker, A.D. & Brizuela, M.A. (1990) Silicon uptake and distribution in Agropyron smithii as related to grazing history and defoliation. Journal of Range Management, 43, 344346.
  • Clancy, K.M. & Price, P.W. (1987) Rapid herbivore growth enhances enemy attack − sublethal plant defenses remain a paradox. Ecology, 68, 733737.
  • Djamin, A. & Pathak, M.D. (1967) Role of silica in resistance to Asiatic rice borer Chilo suppressalis (Walker) in rice varieties. Journal of Economic Entomology, 60, 347–&.
  • Drave, E.H. & Lauge, G. (1978) The study on the action of silica on the wearing of mandibles of the Pyralid of rice: Chilo suppressalis (F. Walker).Bulletin of the Entomological Society of France, 83, 159162.
  • Ellis, R.P. (1979) A procedure for standardizing comparative leaf anatomy in the Poaceae. II. The epidermis as seen in surface view. Bothalia, 12, 641671.
  • Fraser, L.H. & Grime, J.P. (1999) Interacting effects of herbivory and fertility on a synthesized plant community. Journal of Ecology, 87, 514525.
  • Gali-Muhtasib, H.U., Smith, C.C. & Higgins, J.J. (1992) The effect of silica in grasses on the feeding-behavior of the prairie vole Microtus ochrogaster. Ecology, 73, 17241729.
  • Grime, J.P., Cornelissen, J.H.C., Thompson, K. & Hodgson, J.G. (1996) Evidence of a causal connection between anti-herbivore defence and the decomposition rate of leaves. Oikos, 77, 489494.
  • Hammond, T.A. & Ennos, A.R. (2000) Mechanical testing of the abrasiveness of grass leaves. Third Plant Biomechanics Conference Proceedings. Georg Thieme Verlag, Stuttgart.
  • Herrera, C.M. (1985) Grass grazer radiations − an interpretation of silica-body diversity. Oikos, 45, 446447.
  • Hochuli, D.F. (1993) Does silica defend grasses against invertebrate herbivores? Trends in Ecology and Evolution, 8, 418418.
  • Iler, R.K. (1979) The Chemistry of Silica. John Wiley & Sons, New York.
  • Janis, C.M. & Fortelius, M. (1988) On the means whereby mammals achieve increased functional durability of their dentitions, with special reference to limiting factors. Biological Reviews of the Cambridge Philosophical Society, 63, 197230.
  • Jernvall, J. & Fortelius, M. (2002) Common mammals drive the evolutionary increase of hypsodonty in the Neogene. Nature, 417, 538540.
  • Jones, L.H.P. & Handreck, K.A. (1967) Uptake of silica by Trifolium incarnatum in relation to the concentration in the external solution and to transpiration. Plant and Soil, 30, 7180.
  • Kaufman, P.B., Dayanandan, P., Franklin, C.I. & Takeoka, Y. (1985) Structure and function of silica bodies in the epidermal system of grass shoots. Annals of Botany, 55, 487507.
  • Kim, H.S. & Heinrichs, E.A. (1982) Effect of silica level on whitebacked planthopper. International Rice Research Newsletter, 7, 17.
  • Lee, K.P., Raubenheimer, D., Behmer, S.T. & Simpson, S.J. (2003) A correlation between macronutrient balancing and insect host-plant range: evidence from the specialist caterpillar Spodoptera exempta (Walker). Journal of Insect Physiology, 49, 11611171.
  • McNaughton, S.J. & Tarrants, J.L. (1983) Grass leaf silicification − natural-selection for an inducible defense against herbivores. Proceedings of the National Academy of Sciences USA, Biological Sciences, 80, 790791.
  • McNaughton, S.J., Tarrants, J.L., McNaughton, M.M. & Davis, R.H. (1985) Silica as a defense against herbivory and a growth promotor in African grasses. Ecology, 66, 528535.
  • Moore, D. (1984) The role of silica in protecting Italian ryegrass (Lolium multiflorum) from attack by dipterous stem-boring larvae (Oscinella frit and other related species). Annals of Applied Biology, 104, 161166.
  • O'Reagain, P.J. & Mentis, M.T. (1989) Leaf silification in grasses − a review. Journal of the Grassland Society of South Africa, 6, 3743.
  • Parker, W.E. & Gatehouse, A.G. (1985) The effect of larval rearing conditions on flight performance in females of the African armyworm, Spodoptera exempta (Walker) (Lepidoptera, Noctuidae). Bulletin of Entomological Research, 75, 3547.
  • Parry, D.W. & Smithsonian, F. (1964) Types of opaline silica depositions in leaves of British grasses. Annals of Botany, 28, 169185.
  • Patterson, B.D. (1983) Grasshopper mandibles and the niche variation hypothesis. Evolution, 37, 375388.
  • Patterson, B.D. (1984) Correlation between mandibular morphology and specific diet of some desert grassland Acrididae (Orthoptera). American Midland Naturalist, 111, 296303.
  • Raps, A. & Vidal, S. (1998) Indirect effects of an unspecialized endophytic fungus on specialized plant–herbivorous insect interactions. Oecologia, 114, 541547.
  • Raubenheimer, D. & Simpson, S.J. (2003) Nutrient balancing in grasshoppers: behavioural and physiological correlates of dietary breadth. Journal of Experimental Biology, 206, 16691681.
  • Russel, E.W. (1961) Soil Conditions and Plant Growth. Longmans Green, New York.
  • Slansky, F. (1985) Food utilization by insects − interpretation of observed differences between dry-weight and energy efficiencies. Entomologia Experimentalis et Applicata, 39, 4760.
  • Smith, G.S., Nelson, A.B. & Boggino, E.J.A. (1971) Digestibility of forages in-vitro as affected by content of silica. Journal of Animal Science, 33, 466–&.
  • Van Soest, P.J. & Jones, L.H.P. (1968) Effect of silica in forages upon digestibility. Journal of Dairy Science, 51, 1644–&.
  • Vicari, M. & Bazely, D.R. (1993) Do grasses fight back − the case for antiherbivore defenses. Trends in Ecology and Evolution, 8, 137141.
  • Watt, A.D. (1979) Effect of cereal growth-stages on the reproductive activity of Sitobion avenae and Metopolophium dirhodum. Annals of Applied Biology, 91, 147157.
  • Williams, I.S. (1999) Slow-growth, high-mortality − a general hypothesis, or is it? Ecological Entomology, 24, 490495.
  • Woodrow, K.P., Gatehouse, A.G. & Davies, D.A. (1987) The effect of larval phase on flight performance of African armyworm moths, Spodoptera exempta (Walker) (Lepidoptera, Noctuidae). Bulletin of Entomological Research, 77, 113122.