Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
1 This study evaluates the potential for indirect competition between two phloem-feeding aphids as mediated by a shared host, pecan Carya illinoensis. In a greenhouse experiment, one of two aphid species, Monellia caryella and Melanocallis caryaefoliae, were introduced to pecan seedlings, removed for a period, and then introduced for a second time. Aphid performance and food quality, i.e. phloem amino acid concentration and composition, were measured in leaves after the first and second exposure to aphids. After the second exposure, leaves had been fed upon previously by either conspecifics or heterospecifics, and both direct and delayed effects were evaluated on adjacent leaves.
2 The performance of M. caryaefoliae was reduced by previous aphid feeding of both conspecifics and heterospecifics. The performance of M. caryella was unaffected by prior aphid feeding.
3 Feeding by M. caryaefoliae induced changes in amino acid content of the phloem. This alteration occurred within infested leaves and did not cause any changes in the phloem of adjacent leaves.
4 Feeding by M. caryella did not induce changes in phloem amino acid content, but seemed to inhibit M. caryaefoliae’s ability to alter the phloem. The inhibition caused by M. caryella was local and might be the cause of the indirect competition observed.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
We present a study of two related aphid species and their potential for interactions: the blackmargined pecan aphid Monellia caryella (Fitch) and the black pecan aphid Melanocallis caryaefoliae (Davis), feeding on pecan trees Carya illinoensis (Wang.) K. Kock. Both species are found primarily on C. illinoensis but are also found on other Carya spp. (Blackman & Eastop 1994). Both species overwinter as eggs on pecan trees. In the spring, fundatrices hatch from the eggs and initiate several alate parthenogenetic generations during the spring and summer. In the autumn, sexual individuals develop, then mate and lay overwintering eggs (Bissell 1978; Tedders 1978).
A characteristic feature of the distribution and abundance of the two aphid species on pecans is a spatial separation early in the growing season: M. caryella is more abundant in the outer parts of the canopy and M. caryaefoliae is more abundant in the shaded central parts. The aphids also differ slightly in their feeding sites: M. caryella tends to feed on primary leaf veins on the underside of the leaf, while M. caryaefoliae prefers tertiary and quaternary veins on both sides of the leaf (Tedders 1978; Kaakeh & Dutcher 1994). Field studies of their population biology show that the two species often co-occur later in the season but that they usually have population peaks at different times (Tedders 1978; Liao et al. 1984).
Although the aphids share the same resource (the phloem), they induce different types of damage to the leaves of their host. Feeding by M. caryaefoliae is followed by severe macroscopic changes in the leaves, i.e. chlorosis and necrosis, while feeding by M. caryella causes only insignificant macroscopic effects. However, histological studies have shown that M. caryella causes microscopic damage to the phloem cells (clogging); this is much more severe than the damage caused by M. caryaefoliae (Tedders & Thompson 1981; Wood, Tedders & Thompson 1985).
A study of the direct competition between M. caryella and M. caryaefoliae suggests that the outcome of competition depends on whether the foliage has been infested previously by one or both aphid species and whether the infestation reached a population outbreak level (Petersen & Hunter 2001). One possible mechanism for this observation could be indirect exploitative competition between the aphids, caused by changes in plant physiology during aphid feeding. To test this hypothesis, we measured the delayed effects on each aphid species’ performance on pecan seedlings following feeding by conspecific or heterospecific aphids. We also measured nutritional changes in the phloem, i.e. amino acid composition and concentration. Amino acids are the only nitrogen source for aphids; nitrogen is one of the basic nutrients required for growth and is frequently strictly limiting for phytophagous insects (Mattson 1980).
Materials and methods
Laboratory cultures of M. caryella and M. caryaefoliae were established from specimens collected in a pecan orchard at Sahuarita, Arizona, USA (31°57′ N, 111°00′ W) in August 1998. Aphids were reared on pecan seedlings germinated from nuts collected in an orchard with ‘Wichita’ variety intercropped with the pollinator variety ‘Western’. The cultures were kept in a greenhouse at 20–27 °C and provided with light for 15 h per day.
A greenhouse experiment was carried out during the winter of 1998. The experimental unit was an 8–10-week-old pecan seedling with 5–6 leaves (none of which were compound leaves). Each plant was grown in a 3·8 L pot with Premier® Pro Mix soil. Plants were watered regularly with tap water.
The experiment consisted of three periods of 9 days each: (i) day 1–9, an initial aphid infestation, hereafter called period I; (ii) day 10–18, an intervening period without aphids, and (iii) day 19–27, a second aphid infestation, hereafter called period II (Table 1).
Table 1. Experimental design for testing direct (same leaf) and systemic (adjacent leaf) effects of infestion with M. caryella and M. caryaefoliae on leaves of C. illinoensis and the effects of a second aphid population. The treatments during periods I and II were assigned randomly to the three upper leaves on each plant
Period I (days 1–9)
No aphids (days 10–18)
Period II (days 19–27)
Leaves cut and used for sampling of phloem; n = 8.
Before period I was initiated, two of the three upper and fully developed leaves were selected at random for aphid infestation, the third leaf remaining as the ‘no-aphid’ treatment. Five adult aphids, < 24-h-old, were placed on each of these leaves by tapping them out from the glass vial in which they had been collected. Each infested leaf was then immediately enclosed in a bag made of perforated plastic that was tied around the petiole using a piece of foam rubber and a twist tie. The whole plant was then covered with a bag made of nylon net, secured with rubber bands around the pot. After 9 days’ incubation, the degree of leaf damage (chlorotic and necrotic spots) was evaluated on the infested leaves: 0, no damage; 1, < 10%; 2, 10–25%; 3, 25–50%; 4, > 50%; 5, leaf dead. The number of nymphs produced on each infested leaf was counted and they were then removed. One of the two infested leaves was randomly cut off for sampling of leaf exudates (see below). The other infested leaf was washed carefully using moist cotton to ensure that all aphids were removed. Two leaves on each control plant (‘no aphids’) were treated similarly. Each plant was then showered with tap water, covered again and left in the greenhouse for 9 days. The length of the aphid infestation in each period was designed so that it terminated before the produced offspring became adults.
A second aphid infestation (see Table 1 for treatments) was carried out by placing aphids, as described above, onto the remaining (previously infested) leaf and onto the third (previously uninfested) leaf (treatments 1–4). The previously uninfested leaf was used to test for delayed systemic responses to aphid feeding. On the plants that did not receive an aphid treatment (treatments 5–6) during period I, only one leaf was infested during period II. The other leaf on the same plant was used to test for contemporary systemic responses to aphid feeding on the phloem of the adjacent leaf. The test of systemic effects always refers to between-leaf effects. On day 27, leaf damage was assessed again and the number of nymphs produced on each of the infested leaves was counted. Thereafter, the aphids were removed and the leaves cut for sampling of leaf exudates.
All combinations of aphid treatments (Table 1) were set up in two complete blocks. The full experimental design was seven treatments × two blocks × four replicates. The two blocks were separated in time due to the intensive amount of work; the second block was started 1 day after the first. The experiment was carried out in a greenhouse with 15 h of light and an average temperature of 19·8 °C (range 8·9–34·5 °C). Temperature was measured with a Tiny Talk temperature logger (Gemini Data Loggers Ltd, Chichester, UK) that was mounted on a separate pecan seedling covered with a fabric bag.
Estimation of the cumulative physiological time accumulated in the three periods was made using an estimated temperature threshold of aphid growth of T0 = 5·0 °C [from a tree-living aphid species belonging to the genus Myzocallis Passerini (Sternorrhyncha: Aphididae: Drepanosiphinae: Phyllaphidini) (P. Kindlmann, personal communication)]. The sum of day degrees was 133 °D during period I, 108 °D in the intervening period without aphids and 139 °D during period II.
Phloem amino acid content was assessed by taking leaf exudate samples. This was a more suitable method than taking exudates from aphid stylets, since a larger number of samples could be obtained and samples could also be taken from uninfested plants. The leaf exudate can be used for quantification of amino acid composition in phloem since it has a good correlation with phloem collected from cut aphid stylets (Weibull, Ronquist & Brishammar 1990; Girousse et al. 1991). Cutting leaves for exudate samples may affect the chemistry of the remaining leaves; however, all plants included in the experiment were treated similarly at the end of period I. We therefore assume the effect from the leaf cutting to be negligible.
Leaf exudates were collected from pecan seedlings treated as described in Table 1. The leaf was cut from the plant at the petiole and washed carefully in 8 mm ethylenediaminetetraacetic acid (EDTA), pH 7·0. Thereafter, it was cut again with a razor blade about 2–3 cm from the base of the petiole while still immersed in the EDTA solution. The EDTA enhances exudation by inhibiting the phloem-sealing mechanism. After a few minutes’ washing, the cut surface of the leaf was immersed in a vial with 300 µL EDTA solution and incubated in darkness at 24 ± 1 °C and 90–95% relative humidity (RH) for 6 h. The leaf was then removed, the exudates were centrifuged at 10 000 × g for 5 min and the supernatant was mixed 1 : 1 with Beckman Li-S buffer. The samples were then stored at −20 °C until analysis.
Free amino acids in exudate samples were analysed with an ion-exchange amino acid analyser by the ninhydrin method (Beckman, model 7300; Beckman, Fullerton, CA, USA). Concentrations were calculated by comparing peak heights to an external standard. Total amino acid concentration in the leaf exudates was divided by leaf area and the exudation time to yield an estimate of pmol h−1 cm−2. This was justified by a positive correlation between leaf area and amount of amino acids in exudates in control samples (R2 = 0·40, P < 0·001). The amount of leaf exudate is correlated linearly with extraction time within 24 h (Van Helden, Tjallingi & Van Beek 1994).
An analysis of variance using aphid treatment and block as the main factors was used to analyse the number of nymphs produced per cm2 on plants infested for the first time. A generalized linear mixed model (SAS macro GLIMMIX; SAS Institute Inc., Cary, NC, USA) (Littell et al. 1996) was used to analyse the number of nymphs produced during period II. It was assumed that the number of nymphs followed a Poisson distribution. The model also included a logarithmic link function (to adjust for overdispersion) and the leaf area was used as an offset function. Results were then backtransformed to number of nymphs per cm2 and approximate standard errors were given in percentages. No significant effect was found from the blocks and therefore this factor is omitted from the presentation of results.
Total concentrations of amino acids were analysed by a non-parametric comparison of samples Kruskal–Wallis test, followed by multiple comparison tests (Zar 1984).
PERFORMANCE OF APHIDS
When plants were infested for the first time (treatments 1–4 during period I and treatments 5–6 during period II), M. caryaefoliae produced more nymphs per cm2 than M. caryella (Fig. 1).
The number of M. caryaefoliae produced per cm2 at the end of period II was greatest on leaves that did not experience a direct aphid infestation during period I. No such difference was found for M. caryella (Table 2). The number of nymphs produced per cm2 leaf at the end of period II (Table 2) was unaffected by the treatment (period I × period II) (F4,19 = 2·24, P = 0·103). However, both the type (same leaf, adjacent leaf, no aphids) of the aphid infestation in the first period (F1,19 = 16·46, P < 0·001) and the interaction term between the treatment and the type of infestation (F3,19 = 3·29, P = 0·043) was of significant importance for the number of aphids produced per cm2 at the end of period II. The individual pecan seedling was of significant importance as a covariate in the analysis of the number of aphids produced per cm2 leaf (P = 0·026).
Table 2. Performance of the aphids M. caryella and M. caryaefoliae on leaves of C. illinoensis earlier infested with the same (or the alternative) aphid species. Performance was evaluated after a period of 9 days’ infestation. Approximate standard errors are given as percentages. ‘Direct’ refers to leaves that were earlier infested directly with aphids and ‘systemic’ refers to leaves that were earlier uninfested but adjacent to an infested leaf on the same plant. Statistical test for comparison between ‘direct’ and ‘systemic’ shown to the right. Statistical test for comparison with control shown below (the same control value was used in comparison with values from direct or systemic samples). The terms ‘direct’ and ‘systemic’ only refer to treatments with aphid infestations in both periods
Aphid species during period I
Effect of aphid species during period II (nymphs per cm2 leaf)
Only the leaves that were infested with M. caryaefoliae during period I had developed chlorotic and necrotic spots at the end of this period (the median index value was 2·0 on a scale from 0 to 5). The leaf damage effects continued to increase during the intervening period without aphids, and several of the severely damaged leaves fell off during period II, thereby reducing the number of replicates in these treatments.
After period II, leaves infested with M. caryaefoliae had developed visible damage while none of the leaves infested with M. caryella had developed any visible macroscopic symptoms throughout the experiment (Table 3). Leaves that were first infested with M. caryella and then infested with M. caryaefoliae were less damaged than all other leaves infested with M. caryaefoliae. Leaves adjacent to those infested with M. caryella during period I and then infested with M. caryaefoliae showed a degree of damage at the end of period II (Table 3) similar to that of plants only infested with M. caryaefoliae during period II, suggesting no systemic effect from the first infestation.
Table 3. Median of leaf damage (range in brackets) caused by feeding by M. caryella and M. caryaefoliae on leaves of C. illinoensis seedlings after period II, days 19–27. The degree of leaf damage (chlorosis and necrosis) was evaluated on the infested leaves on a scale of 0 to 5. ‘Direct’ refers to leaves that were earlier directly infested with aphids and ‘systemic’ refers to leaves that were earlier uninfested but adjacent to an infested leaf on the same plant. n = 8
Type of infestation during period I
During period I
During period II
AMINO ACIDS IN THE PHLOEM
Leaf exudates taken at the end of period I from leaves infested with M. caryaefoliae exuded seven times the total amount of amino acids compared with uninfested leaves. The total amount of amino acids exuding from leaves infested with M. caryella was similar to that from uninfested leaves (Fig. 2). In leaf exudates from plants infested for the first time during period II, similar changes after infestation with M. caryaefoliae were observed – three times the amount of total amino acid compared with samples uninfested in both periods – while samples from plants with M. caryella were similar to uninfested leaves (Fig. 2).
The increase in the amount exuding from the leaves after infestation with M. caryaefoliae was dependent on the number of aphids the individual leaf supported, since a significant positive correlation was found between aphid density and the amount of amino acids exuding (log y = 1·14 + 1·24 log x, R2 = 0·31, P = 0·003; where x is the number of aphids per cm2 and y is the concentration of amino acids per h and cm2). This relationship explains the greater alteration by M. caryaefoliae during period I compared with period II (Fig. 2), since the number of M. caryaefoliae produced per leaf during period II was only two thirds of the number produced during period I. The correlation for M. caryella was not significantly different from zero (log y = 1·18 + 0·41 log x, R2 = 0·07, P = 0·122), and the intercept for both lines was near the value for control leaves (log y = 1·29).
Leaf exudate samples taken at the end of period II from leaves that were infested with M. caryaefoliae during both periods I and II showed a more than six-fold increase in total amino acids (Q8 = 3·01, P < 0·05) (Table 4) compared with leaves from control plants uninfested in both periods I and II. The corresponding value for leaves infested twice with M. caryella was not significantly different from the control (Q8 = 0·83 NS) (Table 4). When the leaf was infested first with M. caryaefoliae and then with M. caryella (treatment 2), only one replicate remained for exudate sampling due to aphid damage. The concentration in this sample was elevated, and thus showed similarities with those infested only with M. caryaefoliae (treatment 1). In the treatment where the leaf was infested first with M. caryella and then with M. caryaefoliae (treatment 3), the concentration recorded was not significantly different from control plants (treatment 7) (Q8 = 1·38 NS) (Table 4). Thus, a previous infestation with M. caryella seems to prevent the effects of M. caryaefoliae on total amino acid concentration in leaf exudates.
Table 4. Amount of amino acids (pmol h−1 cm−2 leaf) (means ± SE) exuding from a single C. illinoensis leaf after period II, days 19–27. ‘Direct’ refers to exudates from leaves that were infested with aphids during both periods, ‘systemic’ refers to leaves that were not exposed directly to aphids during period I, days 1–9, but that were situated on a plant where adjacent leaves were infested with aphids
To investigate if the plants experience a contemporary systemic effect of aphid infestation between leaves, exudates were taken from uninfested leaves adjacent to infested leaves during the second aphid infestation period. These plants were uninfested during the first aphid infestation period (treatments 5 and 6). Total amino acid concentration in leaves adjacent to those infested with M. caryaefoliae during the second aphid infestation (treatment 5) (21·7 ± 3·7, n = 8) were similar to control leaves from plants uninfested during both periods (treatment 7) (19·7 ± 2·0, n = 7 (Q3 = 0·15 NS). Likewise, no differences were observed between uninfested plants and samples from leaves adjacent to those infested with M. caryella during the second aphid infestation (treatment 6) (19·5 ± 1·5, n = 8) (Q3 = 0·08 NS). Thus, neither aphid species’ feeding had a contemporary systemic effect on total amino acid concentration.
To investigate if the plants experienced a delayed systemic effect between the leaves, exudates were taken from leaves adjacent to infested leaves during period I and then infested during period II. In leaves uninfested but adjacent to infested leaves in period I, and infested with M. caryaefoliae during period II, increased concentrations of amino acids were recorded compared with uninfested plants, irrespective of which aphid species had been present on the adjacent leaves during period I (treatment 1: Q8 = 3·08, P < 0·05; treatment 3: Q8 = 3·65, P < 0·01) (Table 4). In corresponding leaves that were infested with M. caryella during period II only, a slight increase was observed compared with leaves from control plants; however, it was not statistically significant (treatment 2: Q8 = 1·75 NS; treatment 4: Q8 = 1·35 NS). Thus, there were no obvious delayed systemic effects of aphid feeding from period I. It should be observed that the inhibitory effect of a previous M. caryella infestation on a later M. caryaefoliae phloem alteration is not systemic. During period II, M. caryaefoliae was able to alter phloem in leaves positioned adjacent to those previously infested with M. caryella (Table 4).
Amino acid composition in leaf exudates taken from control plants (‘no aphids’) and from leaves infested with M. caryella once (either period I or period II) were indistinguishable, indicating that M. caryella does not change the composition (Fig. 3). Leaf exudates from plants infested once with M. caryaefoliae showed dramatic changes in amino acid composition compared with control plants, especially for the amino acids aspartic acid (ASP), glutamine (GLN), glycine (GLY), methionine (MET), serine (SER) and valine (VAL) (Fig. 3).
Leaves infested with M. caryella in both periods I and II showed a composition very similar to that found after a single aphid infestation. Leaves infested with M. caryaefoliae in both periods I and II showed a similar change in composition to that seen after a single period with M. caryaefoliae; however, it was more drastic. Those leaves infested with M. caryella during period I and with M. caryaefoliae during period II showed a weaker change than after a single M. caryaefoliae infestation, suggesting an inhibitory effect of the preceding infestation with M. caryella. From the treatment with M. caryaefoliae during period I and with M. caryella during period II, only one replicate remained, and it showed similarities with exudates from leaves infested only with M. caryaefoliae.
We found: (i) that M. caryaefoliae’s performance was reduced on pecan foliage by prior feeding of both conspecifics and heterospecifics. Melanocallis caryaefoliae thus experiences indirect intra- and interspecific competition. (ii) Melanocallis caryaefoliae’s feeding altered the pecan seedlings’ physiology and caused a change in amino acid content and profile of the phloem. This alteration was only local (within a leaf) in its effects. (iii) A previous infestation with M. caryella appeared to inhibit M. caryaefoliae’s ability to alter phloem amino acid content. This inhibition was also local in its effects.
Melanocallis caryaefoliae produced fewer nymphs on leaves that had had a previous aphid infestation, independent of the aphid species. The indirect competition from conspecifics is not so surprising, since many leaves were damaged by the first infestation and the damage was exaggerated during the second infestation. It is interesting that the previous infestation with M. caryella also reduces M. caryaefoliae’s performance, because this species neither altered the amino acid content of the phloem nor caused any macroscopic damage to the leaf tissue. However, previous research suggests that M. caryella causes extensive injuries to phloem cells, i.e. clogging by callose (Tedders & Thompson 1981; Wood et al. 1985). These clogged phloem cells or some linked alteration may be responsible for the inhibition of M. caryaefoliae’s ability to alter amino acid composition and concentration following M. caryella feeding. The performance of M. caryella was not affected by the previous infestation by M. caryaefolia despite leaf damage and changes in phloem amino acids. Monellia caryella did not exaggerate the damage to the leaf and therefore perhaps it did not experience reduced performance. Changes in phloem sap caused by M. caryaefoliae might not be accessible to M. caryella or the changes could have diminished after the 9 days’ intervening period. In a field experiment where pecan leaves were exposed to extensive and long-term aphid infestation, prior feeding by M. caryella was found to ‘condition’ the leaves in a way that reduced the performance of another pecan aphid, Monelliopsis pecanis (Bissel). The opposite order of infestation with the two aphid species did not result in a reduced performance of M. caryella (Bumroongsook & Harris 1992). Similarly, Sluss (1967) reports that the walnut aphid, Chromaphid juglandicola (Kalt.) causes ‘conditioning’ of the leaves that affects later infestations by conspecifics.
HOST PLANT PHLOEM
The ability of aphids to alter their host plants’ phloem has been suggested in several studies (Way & Banks 1967; Way & Cammell 1970; Forrest 1971; Dorschner et al. 1987; Riedell 1989), and has actually been demonstrated by analysis of leaf exudates (Poehling 1985) and by direct analysis of phloem collected from aphid stylets (Telang et al. 1999; Sandström et al. 2000). The leaf exudate samples taken from pecan leaves infested with M. caryaefoliae suggest that this aphid species changes the composition and increases the total concentration of amino acids in phloem. However, the observed seven-fold increase in total amino acid concentration in exudates from cut leaves may be somewhat exaggerated compared with the actual concentration ingested by the aphid. A six-fold increase in leaf exudates from wheat infested with the aphid Schizaphis graminum (Rondani) corresponded to a two-fold increase in samples from cut aphid stylets (Sandström et al. 2000).
The type of alteration that M. caryaefoliae caused in the pecan phloem is likely to be nutritionally advantageous to the individual feeding for some days. However, when the alteration results in severe leaf damage the advantage may be lost, resulting in negative effects on the population, seen as reduced aphid growth on infested leaves and finally leaf abscission due to severe macroscopic damage. The mobility of the aphids (i.e. all adults are alate) might circumvent this disadvantage. The amino acid alterations caused by M. caryaefoliae may be compared with the senecence-like alterations that S. graminum was found to cause in wheat (Dorschner et al. 1987). Dorchner et al. hypothesized that the aphids take advantage of an increased translocation, resulting from breakdown of leaf proteins. The drastic increase in the amino acid glutamine in our study (Fig. 3), also found in the case of S. graminum (Sandström et al. 2000), supports the hypothesis that M. caryaefolia’s feeding induces senescence-like changes. Glutamine is considered the major nitrogen form translocated from senescent leaves to sink organs in rice and possibly other plants (Kamachi et al. 1991; Watanabe et al. 1997).
Melanocallis caryaefoliae is probably ingesting a nutritionally richer diet than M. caryella. This should make it possible for M. caryaefoliae to reduce its intake of phloem since its growth rate is not generally higher than M. caryella’s (Kaakeh & Dutcher 1994; Petersen & Hunter 2001). This scenario is supported by the much larger amounts of carbohydrates excreted by M. caryella compared with M. caryaefoliae (Tedders & Wood 1987); the surplus of carbohydrates could result from large volumes of phloem ingested to satisfy the demand for amino acids. Similarly, S. graminum also has a reduced ingestion of phloem compared with other grass-feeding aphids, which is probably a consequence of its enhancement of the amino acid composition and concentration in its grass host (Sandström & Moran 2001).
The alteration by M. caryaefoliae seems to be restricted to a single leaf: no direct or delayed systemic effect was observed on adjacent leaves of the same plant. Systemic effects are supposed to be most pronounced in leaves on the same orthostichy, i.e. leaves vertically aligned on the stem and thereby directly connected by the vascular system (Jones et al. 1993). However, the few developed leaves on our pecan seedlings did not allow such a comparison. Older pecan seedlings would develop compound leaves, and a study carried out on such plants would require another year for raising the plants. Within a leaf, the increased concentrations of amino acids in the exudates indicate that a large area of the leaf is affected. It is not only the phloem vessels the aphids are actually feeding on that are affected – change in a single phloem vessel would hardly be detected in an exudate sample from multiple vessels. However, higher density of M. caryaefoliae will increase the effect on a leaf, suggesting that a few aphids cannot change a whole leaf maximally.
In this study, we have looked at changes in amino acid composition because they are of major nutritional importance for aphids (Mattson 1980). The observed changes in amino acids may be associated with, or secondary to, other changes in the plant, such as the possible effects of allelochemicals on aphid behaviour and performance.
OUTCOME OF COMPETITION
In the present greenhouse study, we have found that M. caryella appears to be able to inhibit M. caryaefoliae’s ability to alter the phloem, and M. caryaefoliae seems to experience competition from M. caryella. These observations could also have ecological effects under field conditions, since aphid population densities can be approximately the same as in our study (i.e. up to 2·0 aphids of each species per cm2 during field outbreaks) (M.K. Petersen, personal communication). Direct competition between the two aphid species was also observed to be asymmetric when they were released simultaneously in both field and greenhouse cage experiments (Petersen & Hunter 2001). When the aphids were released in equal numbers on previously uninfested pecan seedlings in the greenhouse, M. caryella was a superior competitor. A similar result was found in a field experiment in which the leaves were also previously uninfested (Petersen & Hunter 2001). The reason for M. caryella’s competitive superiority could be that the microscopic damage caused by this aphid occurred rapidly relative to the alteration of the phloem that M. caryaefoliae initiated. An early inhibition of heterospecifics’ alteration of a shared host could determine the outcome of competition between the two species in the short term. However, an outbreak of M. caryella a couple of months before a field experiment did not result in the same inhibitory effect: in this situation, both aphid species performed equally well (Petersen & Hunter 2001). In these field and greenhouse experiments, aphids have no real opportunity to select between host plants with different previous infestation histories. Under field conditions, aphids might discriminate between leaves that have not been fed upon as opposed to leaves that have been previously exposed to one of the two aphid species. Such a reaction could strengthen the outcome of the interaction between the two aphid species. A study of aphid feeding behaviour performed by Prado & Tjallingii (1997) shows that some aphid species are able to react to changes caused by a previous aphid infestation.
The observed indirect competition between the two aphid species is mediated by responses in the host plant. However, the aphids’ reactions to changes in the plant might be secondary: the primary purpose of their manipulation of the plant may be something else. Plants probably mobilize defence reactions towards phloem-feeding herbivores. The different reactions observed in this study may be as a result of the aphids’ counteractions to avoid these defence systems; M. caryaefoliae induces senescence that turns off defence reactions (i.e. no active defence during the breakdown process of the leaf). Monellia caryella may use a less apparent action that inhibits the defence reaction as well as the senescence induction caused by M. caryaefoliae.
This study of indirect competition between two aphid species in the pecan system provides an example of herbivore exploitative competition mediated through the host (Wootton 1994). The outcome is a consequence not only of one herbivore removing a food resource required by the other, but of its alteration of the quality so that the other herbivore’s alteration with the host plant is inhibited, causing a reduction in its fitness (Inbar et al. 1995). In future studies, it would be of interest to investigate what mechanisms (see Schoener 1983) are involved in the interspecific competition, how quickly M. caryella inhibits M. caryaefoliae’s alteration of the phloem and for how long this lasts. Such information would be of value for the general understanding of the coexistence and population dynamics of the two aphid species.
We would like to thank Martha Hunter and Elizabeth Bernays for constructive criticism of earlier versions of this manuscript. Thanks to Wallace Clark for help with the analysis of amino acids and to Kristian Kristen for statistical advice. This work was partly funded by USDA South-western irrigated pecan research. J.P.S. was funded by a postdoctoral fellowship from the Swedish Council for Forestry and Agricultural Research and M.K.P. was funded by a postdoctoral fellowship from the Danish Agricultural and Veterinary Research Council.
Received 2 November 2000; revised 26 March 2001; accepted 28 March 2001