SEARCH

SEARCH BY CITATION

Keywords:

  • Artemisia tridentata;
  • communication;
  • fitness;
  • flowering;
  • neighbours;
  • plant–herbivore interactions;
  • reproduction;
  • seedlings;
  • survival;
  • volatiles

Summary

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

1. There are now approximately 10 examples of plants that use volatile cues emitted by damaged neighbours to adjust their defences against herbivores. For two of these examples, preliminary evidence suggested that plants may experience net benefits from such eavesdropping, although eavesdropping was uncommon in one case and estimates of plant fitness were ambiguous in the other case.

2. In the current study, we examined the long-term consequences of exposure to cues emitted by experimentally clipped sagebrush neighbours. In this sagebrush system we have repeatedly found that sagebrush plants that have experimentally clipped neighbours experience less herbivore damage over the season than plants with unclipped control neighbours under field conditions. We followed a cohort of young sagebrush plants from emergence in 1999 for 12 years. Neighbours of half of these plants were artificially clipped every spring from 2004–08 and survival and flowering was measured in each autumn from 1999–2011.

3. Survival of marked branches of young plants was not consistently affected by whether its neighbour was clipped. Plants near clipped neighbours produced more branches during this period than those near unclipped neighbours. There were no measurable treatment effects on plant survival over the 12 years. Branches near clipped neighbours produced more inflorescences than branches near unclipped neighbours.

4. Seedlings were more likely to survive to the end of their first dry season in two different years near clipped neighbours compared to unclipped neighbours.

5.Synthesis. The results suggest different effects of clipped neighbours that depend on plant age. Responding to the cues of experimental clipping may provide a slight net benefit, considering these results and other published studies, even though these cues provided little predictive value about actual risk of herbivory. Responding to reliable cues may be even more beneficial and may favour plants that eavesdrop on neighbours.


Introduction

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

Although volatile communication between plants has been controversial over its 30-year history (Baldwin & Schultz 1983; Rhoades 1983; Fowler & Lawton 1985; Dicke & Bruin 2001), there is now convincing evidence from approximately 10 plant species that individuals respond to airborne cues to adjust their defences against herbivores and pathogens (reviewed by Heil & Karban 2010). Unlike animals, plants lack central nervous systems, but they have faced similar selective pressures to modify their phenotypes to match current and future environmental conditions (Karban 2008). This has presumably led to the evolution of hormonal systems that allow plants to respond to reliable cues which indicate a high risk of attack and become more resistant or more tolerant to attack compared to unresponsive individuals. According to this adaptive hypothesis, plants are not maintained at maximal resistance because resistance is costly either in terms of resources that cannot be invested in other tissues or processes that increase reproductive fitness or else in terms of missed ecological opportunities (Karban & Baldwin 1997). For example, enhanced defence against some pathogens may make plants more vulnerable to attack by caterpillars (Thaler et al. 2002; Thaler, Agrawal & Halitschke 2010).

The adaptive hypothesis predicts that plants with elevated induced defences will experience a net fitness benefit when herbivore attack is likely but will suffer a net cost when attacks are uncommon. In other words, plants should experience a net fitness benefit when they respond appropriately to cues that are reliable and honest. Conversely, plants that respond inappropriately to cues that are unreliable and do not accurately predict future conditions should experience a net fitness cost.

The fitness consequences of responding to volatile cues has been examined in two systems – wild tobacco (Nicotiana attenuata) and wild lima beans (Phaseolus lunata) (Karban & Maron 2002; Kost & Heil 2006). Annual wild tobacco individuals growing near experimentally clipped sagebrush neighbours experienced less herbivory and produced as many or more flowers and seed capsules over their lifetimes as individuals near unclipped sagebrush neighbours in each of 5 years during which experiments were conducted (Karban & Maron 2002). Sagebrush and wild tobacco share many generalist herbivores and sagebrush is often attacked earlier in the season at a time when annual tobacco seeds are just germinating. Nonetheless, experimental damage in this case may represent misinformation, incorrectly signalling a higher risk of herbivore attack than otherwise perceived or actually experienced by the tobacco plants. Lima bean shoots that were treated in the field with artificial volatile cues lost less leaf area to herbivores and produced more leaves and inflorescences than untreated control tendrils (Kost & Heil 2006; Heil & Silva Bueno 2007). As with the previous tobacco example, the artificial cues provided misinformation but were still associated with increased correlates of fitness. When jasmonic acid, a signal that coordinates responses to herbivores, was artificially applied to lima bean shoots, net benefits were minimal perhaps because JA causes other costs such as a reduction in photosynthetic activity (Kost & Heil 2008).

These two examples both have significant limitations. Volatile signals between sagebrush and tobacco are active over distances of only 10–15 cm and most tobacco plants do not grow within this distance of a sagebrush neighbour (Karban et al. 2003). Wild lima beans are perennial vines and distinguishing individuals was not possible; in addition, the correlation between shoot growth and flower production over one ‘snapshot’ of time and lifetime fitness is unclear (Kost & Heil 2006; Heil & Silva Bueno 2007).

Communication between sagebrush individuals occurs over longer distances (60 cm) than that between sagebrush and tobacco (Karban et al. 2006). As such, most sagebrush individuals have the potential to communicate with a neighbour. Communication in this system reliably reduces the endemic levels of herbivore damage experienced by individuals that grow in close proximity to experimentally clipped neighbours (Karban et al. 2006; Shiojiri, Karban & Ishizaki 2009). The experimental clipping applied to emitter plants had the positive short-term effect of reducing levels of herbivore damage for the clipped emitter and the neighbouring plants that responded to volatile cues of clipping. The experimental clipping lacks the predictive value about the likelihood of future attacks that natural herbivore damage should provide. Therefore, the question remains: does this reduction in herbivory translate into a net fitness advantage or disadvantage for plants that are responding to artificially generated cues? We conducted two experiments to address this question: (i) We measured survival of marked branches and entire plants of an even-aged cohort of young sagebrush over 12 years in the field when a neighbouring plant was experimentally clipped for five consecutive years. We compared branches on plants with experimentally clipped neighbours to control plants with unclipped neighbours. We also measured the production of new branches and inflorescences for plants with clipped or unclipped neighbours. (ii) We measured the survival of individuals in two even-aged cohorts of seedlings that grew in proximity to experimentally clipped or unclipped adult neighbours from germination through the end of their first dry season.

Materials and methods

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

Long-term survival and reproduction

We measured the survival and reproduction of a cohort of young sagebrush plants that emerged in 1999 along the south floodplain of Convict Creek at the University of California’s Sierra Nevada Aquatic Research Lab, south of Mammoth Lakes, CA. Sagebrush is a long-lived perennial shrub with rare and episodic natural recruitment. In 20 years of studying this system, we have observed three natural recruitment events, all associated with rainy seasons that were well above average in terms of total precipitation and duration. In this study, we also generated two artificial recruitment events in 1999. One was the unintended consequence of growing a garden of wild tobacco during the 1997, 1998, and 1999 seasons (37° 36.8N 118°49.9W, 7089 ft. elev). To irrigate this garden we pumped water from Convict Creek which flooded a clearing in the shrub habitat that had been created a decade earlier. Since the pump had to operate continuously for several hours, it flooded the clearing three times a week for approximately 3 weeks each summer causing germination of sagebrush seeds from the seed bank. The second recruitment event resulted from a prescribed burn on 7 January 1999 in 5 acres of sagebrush scrub along the western edge of the reserve. Although sagebrush is generally intolerant of hot natural fires (Young & Evans 1978), this cooler winter burn caused germination and/or sprouting of new shoots during the spring of 1999 particularly along the edge of the burn closest to the stream bank (37° 36.8N 118°50.2W, 7102 ft. elev). By spring of 2003, these two disturbed patches had many young sagebrush plants that had emerged in 1999 and were vigorously growing interspersed among existing adult shrubs.

We selected 100 of these 4-year old (juvenile) focal plants which had canopies that were within 10–15 cm east of another adult sagebrush individual. This site is walled in by steep glacial moraines so that wind flow is consistently down slope, from the west. We marked the branch of the downwind focal plant that was closest to the upwind neighbour. In 2003, all the selected focal plants were immature and had not produced any inflorescences. Focal plants had approximately three branches at this time (mean branches per plant ± 1 SE = 3.04 ± 0.11). We randomly assigned half of the upwind adult neighbours to be experimentally clipped (n = 50) and half of the upwind neighbours to serve as unclipped controls (n = 50). During May or June of each spring from 2003–08 following snow melt, we clipped with scissors the distal half of one third of the leaves of the closest branch of each upwind plant that was assigned to the clipped treatment. This damage regime has been found consistently to cause release of volatile cues that induce resistance in the downwind branch at this study site (Karban et al. 2006; Shiojiri & Karban 2006) and to be similar to responses to actual herbivory (Shiojiri & Karban 2008). We assessed survival of the marked branch of each focal plant in autumn from 2003–11. We recorded the number of inflorescences produced during the season on each of the focal branches during each autumn from 2003-2011. We assessed the survival of each plant with a marked focal branch. We also recorded the total number of branches that were produced by each marked plant from 2003–11.

We compared the survival of marked branches of the 100 focal plants with clipped or unclipped neighbours with failure-time analysis (Muenchow 1986; Fox 2001) using the ‘Fit Parametric Survival’ command in JMP 9.0. We treated time to death as our response variable and the two clipping treatments of neighbours (clipped or control) as two levels of a fixed effect variable. All 100 focal branches were included in the analysis and branches that had not died at the end of the 2011 season were right-censored. We used a likelihood ratio test to compare models (fitting a Weibull distribution) that included neighbour clipping treatments and models without this variable. We compared the survival of entire plants with marked branches in a similar manner. We counted the number of new branches added to each marked plant between 2003 and 2011. Production of new branches was compared using a GLM with a Poisson distribution of sampling errors and a natural log link function in JMP 9.0 for plants with clipped or unclipped neighbours as a fixed factor.

The effect of clipping neighbours on time to first reproduction was evaluated using failure-time analysis in JMP, any plants that failed to flower were censored. A likelihood ratio test compared a model with a Weibull distribution that included clipping treatments with a similar model that lacked these treatments. Lifetime numbers of inflorescences for focal branches with clipped or unclipped neighbours were best modelled by a Poisson distribution. We analysed lifetime production of inflorescences using GLM with a log link function and neighbour clipping as a fixed effect. We correlated the number of inflorescences produced on our focal branches in 2011 with the number produced on the entire plant in 2011 using a GLM. We did not quantify levels of herbivory for each year although cecidomyiid galls (Rhopalomyia spp.) were particularly common in 2006 and gelechiid caterpillars (Aroga websteri) were particularly common from 2007–09.

Survival of seedlings

Natural recruitment events occurred at one of our field sites (Taylor meadow, UC Sagehen Creek Field Station, 39° 25.99N 120°14.20W, 6345 ft) during the springs of 2010 and 2011. The ground between existing sagebrush shrubs became a carpet of newly germinated seedlings following snowmelt in May. We marked 1890 seedlings that were within 20 cm of the foliage of one or more adult sagebrush plants on 26 May 2010. We clipped the distal edge of the foliage of neighbouring adult sagebrush plants for 905 seedlings and had 985 seedlings near unclipped control adult sagebrush plants. We recorded the number of seedlings that had survived on 26 July and again on 26 September 2010. The winter rainy season began at this site in mid-September. On 4 June 2011, we marked 1680 and 2040 seedlings near clipped and unclipped neighbouring adults respectively. We recorded the number of seedlings that survived on 19 July and again at the start of the rainy season on 3 October 2011.

Treatment effects of experimentally clipping adult neighbours on the survival of seedlings were evaluated using a failure-time analysis in JMP. The seedlings were located in 20 clearings in 2010 and 30 different clearings in 2011 (25 clearings of each treatment) and we conducted analyses on survival of seedlings in these spatial blocks (each clearing was approximately 1 m2). This assumes that each clearing, rather than each seedling, is an independent replicate and is conservative. We used a likelihood ratio test to compare models that included neighbour clipping with those without this effect.

Results

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

Long-term survival and reproduction

We observed no mortality of focal branches during the 2003, 2004 and 2005 seasons (Fig. 1). However, following the 2005 season, some of the focal branches began to die. We detected no differences in the longevities of branches with clipped neighbours and those with unclipped neighbours (mean ± 1 SE survival time for branches with clipped neighbours = 8.07 ± 0.24 years; for plant with unclipped neighbours = 8.66 ± 0.24; Likelihood ratio test of effect of neighbour clipping χ2 = 0.20, d.f. = 1, = 0.66). By the end of the census (2011), most plants with marked focal branches were still alive (77/100), and likelihood of dying and time to death did not differ for plants with clipped and unclipped neighbours (Table 1; Likelihood ratio test for effect of clipping on time to death χ2 = 0.03, d.f. = 1, = 0.87). Plants with clipped neighbours produced approximately twice as many new branches between 2003 and 2011 as those with unclipped neighbours (Fig. 2; GLM effect of clipping treatments χ= 21.0, d.f. = 1, < 0.0001).

image

Figure 1.  Survivorship for a cohort of sagebrush with neighbouring plants that were either experimentally clipped or unclipped controls. Survival was measured for one branch from each plant. The cohort of plants included in this survey emerged in 1999 and neighbours were clipped every year from 2004–08 as indicated by the arrows above the figure.

Download figure to PowerPoint

Table 1.   Survival of plants from 1999–2011 with clipped and unclipped neighbours
 Clipped neighbourUnclipped neighbourTotal
Dead101323
Alive403777
Total5050100
image

Figure 2.  Number of new branches produced between 2003 and 2011 by plants with clipped or unclipped neighbours (mean ± 1 SE).

Download figure to PowerPoint

Some young plants first began flowering in 2004 when they were approximately 5.5 years of age. By 2011, when the cohort of plants was almost 13 years of age, 53/100 focal branches had produced some flowers. Most (48/53) of the branches that flowered during the 13 years of the experiment produced their first flowers when they were 5–7 years of age. There was no indication that the two treatments differed in the timing of first flowering (Likelihood ratio test for clipping treatments χ= 2.52, d.f. = 1, = 0.11). Proximity to a clipped neighbour did not influence the likelihood that a branch would flower as 27/50 branches with clipped neighbours produced some flowers and 26/50 branches with unclipped neighbours produced some flowers (χ= 0.04, d.f. = 1, = 0.84).

Lifetime flower production (up to 13 years of age) was quite variable among individuals as many branches failed to produce any inflorescences and the most prolific branch produced 50 inflorescences (Fig. 3a). Branches with clipped neighbours that survived to flower did not produce significantly more inflorescences than branches with unclipped neighbours (clipping effect χ= 3.04, d.f. = 1, = 0.08). However, when we included all 100 plants, those that flowered and those that failed to flower during their first 13 years, branches with experimentally clipped neighbours produced more inflorescences than branches with unclipped neighbours (Fig. 3B; clipping effect χ= 5.85, d.f. = 1, = 0.02). These estimates of reproductive output involve the number of inflorescences on a marked focal branch. We do not have estimates of lifetime production for whole plants although the number of inflorescences on the focal branch was highly correlated with the number of inflorescences on the entire plant in 2011 (R= 0.62, χ= 25.4, d.f. = 6, < 0.001).

image

Figure 3.  (a) Number of inflorescences produced by surviving sagebrush branches with clipped or unclipped control neighbours (mean ± 1 SE). All plants in the survey emerged in 1999 and neighbours were clipped every year from 2004–08 as indicated by the arrows above the figure. (b) Number of inflorescences produced during the lifetime for plants with clipped and unclipped neighbours (mean ± 1 SE). This includes plants that died before flowering.

Download figure to PowerPoint

Survival of seedlings

The vast majority of seeds that germinated in spring failed to survive through their first summer. By late July 2010, only 1% (18/1890) of seedlings remained and only 0.3% (5/1890) survived to late September. By July 2011, only 0.6% (24/3720) of seedlings survived and only 0.2% (8/3720) remained at the end of the summer. At both sampling times in both years, seedlings near experimentally clipped adults were more likely to survive than those near unclipped adults (Fig. 4). Including the effect of clipping treatments in a failure-time model for seedlings improved the fit of the model (Likelihood ratio test χ= 4.85, n = 50, d.f. = 1, = 0.03).

image

Figure 4.  The number of living seedlings (log scale) as the dry season progressed that were near adult sagebrush plants that were either clipped in spring or remained as unclipped controls. Seeds germinated in May and the first rains came in September 2010 or October 2011.

Download figure to PowerPoint

Discussion

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

An assessment of the fitness consequences of eavesdropping is essential if we are to understand the evolutionary and ecological ramifications of the phenomenon. This information will help us determine whether eavesdropping on cues produced by damaged neighbours can be considered as an adaptation that enables plants to fine tune their defences to match or even anticipate their risk of attack by herbivores. An assessment of fitness should take into account both survival and reproduction over the entire lifetime of individuals and should be conducted for a large sample of individuals that vary only in their eavesdropping experiences (Endler 1986). This is a very tall order and the two systems (wild tobacco and wild lima beans) for which we had estimates of fitness effects did not come close to fulfilling these requirements (see Introduction).

In the current experiments we found that sagebrush seedlings tended to survive at a higher rate near clipped adults compared to seedlings near unclipped adults (Fig. 4). We failed to detect differences in the survival of focal branches (Fig. 1) or whole plants (Table 1) that depended on whether they grew near a neighbour that had been experimentally clipped. However, plants with clipped neighbours produced more new branches (Fig. 2) and focal branches produced more inflorescences (Fig. 3).

These results present a picture of inconsistent plant responses to an unnaturally amplified signal – exposure to artificially generated cues was beneficial for seedlings, inconsequential for young established plants in terms of survival, but beneficial in terms of production of new branches and inflorescences. There are many reasons why we would not expect sagebrush plants to show consistently strong costs or benefits to experimental clipping.

First, in these experiments we artificially generated volatile cues that are generally associated with high levels of foliar damage caused by herbivory. However, these were not completely honest signals in this case because there were not actually high levels of herbivores present (Krebs & Dawkins 1984; Bradbury & Vehrencamp 1998). Despite the ‘misinformation’, these artificial cues consistently reduced levels of herbivory (Karban et al. 2006; Shiojiri, Karban & Ishizaki 2009). If natural selection has already resulted in matching of plant defences with levels of herbivore damage then our artificial cues should result in inappropriate plant responses, more analogous to ‘cry wolf’ cues than to honest signals (Munn 1986; Shiojiri et al. 2010). It is interesting to note that our artificial cues were relatively more beneficial during 2007–09 (Fig. 1 and 3a), years of relatively high rates of herbivory by caterpillars, when the cues may have been closer to honest signals. It is also interesting to note that responding to inappropriate artificial cues resulted in increased lifetime flowering and capsule production for eavesdropping wild tobacco (Karban & Maron 2002), and increased growth and production of inflorescences for lima bean tendrils (Kost & Heil 2006; Heil & Silva Bueno 2007). Some theoretical models predict that honest signals are not evolutionarily stable and that we might expect alternating cycles of honest and dishonest signals in ‘cry wolf’ games (van Baalen & Jansen 2003; Traulsen & Nowak 2007; Shiojiri et al. 2010). Although these models consider multiple generations of players, they highlight inconsistent selection over time. The conflicting effects on survival at different developmental stages observed herein may also reflect inconsistent selection.

A second hypothesis is that the divergent survival responses represent different stage-specific consequences of the same cues. Although we do not know the active cues that are involved in eavesdropping, methyl jasmonate is a possible candidate. Sagebrush emits relatively large quantities of methyl jasmonate when damaged and this plant hormone is a potent germination inhibitor as well as a regulator of plant defences (Farmer & Ryan 1990; Creelman & Muller 1997; Preston, Betts & Baldwin 2002). Previous results showed that volatiles from experimentally clipped sagebrush reduced germination of seeds of other plant species but had no detectable effects on germination of sagebrush (Karban 2007). Sagebrush seeds that germinate near a clipped adult may experience less competition from other plant species and therefore may accrue a net benefit that is reflected in terms of greater survival (Fig. 4). A preliminary comparison of the composition and density of other plant species in the 50 arenas used in the seedling survival experiments failed to reveal any obvious differences between those near clipped and unclipped sagebrush (data not shown). We cannot distinguish whether these volatile cues evolved as allelopathic agents or if their primary function was to stimulate plant defences against herbivores either among branches or neighbouring plants (Karban, Shiojiri & Ishizaki 2011).

The cues may have induced changes in established plants that would have been appropriate in situations where herbivore pressure was much more intense but were inappropriate, and hence costly, at the herbivore levels that these plants actually experienced. These costs may have been associated with the reductions in survival for focal branches in the clipped treatment during the 2005 and 2006 seasons (Fig. 1). If the same cues have multiple consequences (reduced competition for seedlings and also increased defences and flowering for older plants), then potential selection against increasing defences at inappropriate times may be reduced. It is not unreasonable to imagine that cues may have different effects on plants at different developmental stages and that the dangers and selection pressures may differ with stage (Boege & Marquis 2005; Barton & Koricheva 2010). This line of reasoning assumes that herbivores can depress plant fitness and that defences are costly. Recent experiments that removed herbivores from sagebrush plants using insecticides showed large treatment effects on plant growth and reproduction (Takahashi & Huntley 2010). Experimental evidence for costs of defences in this system is much more limited.

Third, the inconsistent results may be a consequence of the sampling design that we used. Despite starting this experiment with over 5000 seedlings and 100 young plants, the effect sizes of costs and benefits may have been too small to detect at some stages. Similarly, although the 100 focal plants were followed for 12 years, sagebrush is slow-growing and long-lived so a longer experiment would be more informative. We chose to monitor survival and flowering on one branch per plant; a design that includes more independent replicates that are sampled imprecisely maximizes statistical power but sacrifices a more complete assessment of each plant (Zschokke & Ludin 2001). We assume that the fate of a statistical sample of single branches reflects the fate of individual plants, particularly since these plants had relatively few branches. This assumption was supported by data from 2011, the only year in which we tested it. We chose to measure the number of inflorescences produced by each branch although this measure may be a less informative than the number of flowers or seeds produced (Takahashi & Huntley 2010). In addition, experimental clipping of neighbouring sagebrush may have caused clipped plants to change in a variety of characteristics (Karban & Baldwin 1997). Any of these undocumented changes may have affected the seedlings and focal branches that we assayed. As mentioned earlier, there is no reason to conclude that the effects of clipped neighbours that we observed necessarily involved changes in resistance to herbivory.

Although imperfect, this study adds to our understanding of the fitness consequences of receiving volatile cues produced by experimentally damaged neighbours. Our results suggest that the ‘misinformation’ provided by these volatile cues can aid survival of recently germinated seedlings, may or may not reduce the survival of established young plants, and can increase production of new branches and inflorescenses. We expect that plant responses to reliable cues of high herbivore risk, in contrast to the ‘misinformation’ provided by our experimental clipping, will be even more beneficial for plants that respond. Along with previous studies of wild tobacco and lima bean plants that eavesdrop on neighbours, these results suggest a slight net benefit of exposure to volatile cues, even those cues that provide misinformation.

Acknowledgements

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

These experiments were conducted at the UC Sierra Nevada Aquatic Research Lab and Sagehen Creek Natural Reserve and we thank Dan Dawson and Jeff Brown for facilitating our work at those field stations. We thank John Maron and Richard Perloff for assistance with the prescribed burn that produced the young sagebrush plants that we used. This manuscript was improved by Truman Young, Will Wetzel and Masashi Ohara. We were supported by NSF DEB-0121050 and we have no conflicts of interest.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • van Baalen, M. & Jansen, V.A.A. (2003) Common language or tower of Babel? On the evolutionary dynamics of signals and their meanings Proceedings of the Royal Society B, 270, 6976.
  • Baldwin, I.T. & Schultz, J.C. (1983) Rapid changes in tree leaf chemistry induced by damage: evidence for communication between plants. Science, 221, 277279.
  • Barton, K.E. & Koricheva, J. (2010) The ontogeny of plant defense and herbivory: characterizing general patterns using meta-analysis. American Naturalist, 175, 481493.
  • Boege, K. & Marquis, R.J. (2005) Facing herbivory as you grow up: the ontogeny of resistance in plants. Trends in Ecology and Evolution, 20, 441448.
  • Bradbury, J.W. & Vehrencamp, S.L. (1998) Principles of Animal Communication. Sinauer, Sunderland, Massachusetts.
  • Creelman, R.A. & Muller, J.E. (1997) Biosynthesis and action of jasmonates in plants. Annual Review of Plant Physiology and Plant Molecular Biology, 48, 355381.
  • Dicke, M. & Bruin, J. (2001) Chemical information transfer between plants: back to the future. Biochemical Systematics and Ecology, 29, 981994.
  • Endler, J.A. (1986) Natural Selection in the Wild. Princeton Univiversity Press, Princeton, New Jersey.
  • Farmer, E.F. & Ryan, C.A. (1990) Interplant communication: airborne methyl jasmonate induced synthesis of proteinase inhibitors in plant leaves. Proceedings of the National Academy of Sciences, 87, 77137716.
  • Fowler, S.V. & Lawton, J.H. (1985) Rapidly induced defenses and talking trees: the devil’s advocate position. The American Naturalist, 126, 181195.
  • Fox, G.A. (2001) Failure-time analysis: studying events and rates at which events occur. Design and Analysis of Ecological Experiments, 2nd edn (eds S.M. Scheiner & J. Gurevitch), pp. 235266. Oxford University Press, Oxford.
  • Heil, M. & Karban, R. (2010) Explaining the evolution of plant communication by airborne signals. Trends in Ecology and Evolution, 25, 137144.
  • Heil, M. & Silva Bueno, C. (2007) Within-plant signaling by volatiles lead to induction and priming of an indirect plant defense in nature. Proceedings of the National Academy of Sciences, 104, 54675472.
  • Karban, R. (2007) Experimental clipping of sagebrush inhibits seed germination of neighbors. Ecology Letters, 10, 791797.
  • Karban, R. (2008) Plant behaviour and communication. Ecology Letters, 11, 727739.
  • Karban, R. & Baldwin, I.T. (1997) Induced Responses to Herbivory. University of Chicago Press, Chicago.
  • Karban, R. & Maron, J. (2002) The fitness consequences of interspecific eavesdropping between plants. Ecology, 83, 12091213.
  • Karban, R., Shiojiri, K. & Ishizaki, S. (2011) Plant communication – why should plants emit volatile cues? Journal of Plant Interactions, 6, 8184.
  • Karban, R., Maron, J., Felton, G.W., Ervin, G. & Eichenseer, H. (2003) Herbivore damage to sagebrush induces resistance in wild tobacco: evidence for eavesdropping between plants. Oikos, 100, 325332.
  • Karban, R., Shiojiri, K., Huntzinger, M. & McCall, A.C. (2006) Damage-induced resistance in sagebrush: volatiles are key to intra- and interplant communication. Ecology, 87, 922930.
  • Kost, C. & Heil, M. (2006) Herbivore-induced plant volatiles induce an indirect defence in neighbouring plants. Journal of Ecology, 94, 619628.
  • Kost, C. & Heil, M. (2008) The defensive role of volatile emission and extrafloral nectar secretion for lima bean in nature. Journal of Chemical Ecology, 3, 213.
  • Krebs, J.R. & Dawkins, R. (1984) Animal signals: mind-reading and manipulation. Behavioural Ecology: An Evolutionary Approach (eds J.R. Krebs & N.B. Davies), pp . 380402. Blackwell, Oxford.
  • Muenchow, G. (1986) Ecological use of time-failure analysis. Ecology, 67, 246250.
  • Munn, C.A. (1986) Birds that cry wolf. Nature, 319, 143145.
  • Preston, C.A., Betts, H. & Baldwin, I.T. (2002) Methyl jasmonate as an allelopathic agent: sagebrush inhibits germination of a neighboring tobacco, Nicotiana attenuata. Journal of Chemical Ecology, 28, 23432369.
  • Rhoades, D.F. (1983) Responses of alder and willow to attack by tent caterpillars and webworms: evidence for pheromonal sensitivity of willows. Plant Resistance to Insects(ed. P.A. Hedin), pp. 5568. American Chemical Society,Washington, D.C.
  • Shiojiri, K. & Karban, R. (2006) Plant age, communication and resistance to herbivores: young sagebrush plants are better emitters and receivers. Oecologia, 149, 214220.
  • Shiojiri, K. & Karban, R. (2008) Seasonality of herbivory and communication between individuals of sagebrush. Arthropod-Plant Interactions, 2, 8792.
  • Shiojiri, K., Karban, R. & Ishizaki, S. (2009) Volatile communication among sagebrush branches affects herbivory: timing of active cues. Arthropod-Plant Interactions, 3, 99104.
  • Shiojiri, K., Ozawa, R., Kugimiya, S., Uefune, M., van Wijk, M., Sabelis, M.W. & Takabayashi, J. (2010) Herbivore-specific, density-dependent induction of plant volatiles: honest or “cry wolf” signals? PLoS ONE, 5, e12161.
  • Takahashi, M. & Huntley, N. (2010) Herbivorous insects reduce growth and reproduction of big sagebrush (Artemisia tridentata). Arthropod-Plant Interactions, 4, 257266.
  • Thaler, J.S., Agrawal, A.A. & Halitschke, R. (2010) Salicylate-mediated interactions between pathogens and herbivores. Ecology, 91, 10751082.
  • Thaler, J.S., Karban, R., Ullman, D.E., Boege, K. & Bostock, R.M. (2002) Cross-talk between jasmonate and salicylate plant defense pathways: effects on several plant parasites. Oecologia, 131, 227235.
  • Traulsen, A. & Nowak, M.A. (2007) Chromodynamics of cooperation in finite populations. PLoS ONE, 2, e270.
  • Young, J.A. & Evans, R.A. (1978) Population dynamics after wildfires in sagebrush grasslands. Journal of Range Management, 31, 283289.
  • Zschokke, S. & Ludin, E. (2001) Measurement accuracy: how much is necessary? Bulletin of the Ecological Society of America, 82, 237243.