8/18/2019 1-s2.0-S1872203214000547-main

1/12

Interactions between plants and herbivores: A review of plant defense

Bin Gong  a, Guangfu Zhang  a,b,*a Jiangsu Key Laboratory of Biodiversity and Biotechnology, School of Life Sciences, Nanjing Normal University, Nanjing 210023, Chinab State Key Laboratory of Palaeobiology and Stratigraphy (Nanjing Institute of Geology and Palaeontology, CAS), Nanjing 210008, China

A R T I C L E I N F O

 Article history:

Received 7 January 2013

Revised 5 June 2013

Accepted 23 July 2013Available online

Keywords:

Herbivore

Plant anti-herbivore defense

Plant–animal interaction

Resistant

Tolerance

Escape strategies

A B S T R A C T

Ecologists have long ignored or underestimated the importance of plant–herbivore interactions owing

to the diversities of herbivores, plant defensive strategies and ecological systems. In this review, we briefly

discussed the categories of herbivores. Then we reviewed the major types of plant defenses against her-

bivores. Selective forces of herbivore pressures have led to the evolution of various defensive mechanismsin plants, which can be classified into (i) resistance traits that reduce the amount of damage received,

including physical, chemical, and biotic traits; (ii) tolerance mechanisms that decrease the impact of her-

bivore damage, and (iii) escape strategies that reduce the probability of plants to be found by herbivores.

These strategies have been studied at different levels from molecular genetics and genomics, to chem-

istry and physiology, to community and ecosystem ecology. We summarized the development of the

methodology for studying plant defenses against herbivores. Particularly, 24 of those hypotheses and models,

which are influential in the international community concerning the relationship between plants and

herbivores, including the defensive mimicry hypothesis, the compensatory continuum hypothesis, the

slow-growth-high-mortality hypothesis, etc, were introduced and grouped into four categories accord-

ing to plant defense strategies in the present review. Finally, we also reviewed the research progress of 

plant–herbivore interactions in China, and discussed the perspectives of studies on plant–herbivore

interactions.

© 2013 Ecological Society of China. Published by Elsevier B.V. All rights reserved.

1. Introduction

The interactions between plants and herbivores are among the

most important ecological interactions in nature [1]. As primary pro-

ducers, almost all plants inevitably avoid being eaten by herbivores

[2]. Thus, these relationships will affect nutrient cycles and energy

flows of food chains [3]. It is reported that these herbivores consume

over 15% of the whole plant biomass produced annually in tem-

perate and tropical ecosystems. Accordingly, this makes herbivory

the major conduit by which energy enters food chains  [1,4].

Morethanthree-quarters of animals areherbivores in nature, which

play a significant role in shaping ecosystem structure and function

[5,6]. Herbivores have a strongeffect on their distributions andabun-

dances by consuming plants [7,8]. They also exert a strong selective

pressure on plant population by increasing its mortality and deplet-

ing biomass which can be used for plant growth and reproduction

[9]. On thecontrary,such habitatconditions as community type, plant

density and light intensity, will result in spatial variation of plant-

eating insects. Therefore, the interactions between plants and

herbivores notonly affect the structure and dynamics of plant popu-

lations, but alsoaffect community composition and diversity, as well

as ecosystem through food web and nutrient cycles [10].

When attacked by herbivores, plants can take various defensive

measures, which are essential in the research field of interactions

between plants and herbivores. Firstly, plant defense has played a

critical role in the long-term co-evolution of plants and herbivores.

For this reason, understanding the evolution and ecology of plant

defenses is nearly equivalent to understanding the originand func-

tion of extant ecosystems [1]. Secondly, the plant defense research

deals withmultiple subdisciplines and different scales, for example,

from genetics and genomic to chemistry and physiology, to com-

munity ecology, ecosystem sciences and global patterns of herbivory

and defense [1]. Another reason for studying plant defense against

herbivores is that every year herbivory causes world economies to

lose billions of dollars of revenue related to agriculture, horticul-

ture and forestry [11]. Therefore, the study of plant defense is

particularly necessary. It is quite common to carry out studies about

plant defense characteristics, defense mechanisms and other re-

spects abroad; however, there arevery fewrelated researchesat home.

In this review, we briefly discussed the categories of herbi-

vores. Then we reviewed the major types of plant defenses against

herbivores from an ecological point of view, classified them into

three categories including resistance traits, tolerance mechanisms

and escape strategies. We also summarized the development of the

* Corresponding author. Jiangsu Key Laboratory of Biodiversity and Biotechnology,

School of Life Sciences, Nanjing Normal University, Nanjing 210023, China. Tel:

+86-25-13915978931; fax:  +86-25-85891839.

E-mail address: zhangguangfu@njnu.edu.cn (G. Zhang).

http://dx.doi.org/10.1016/j.chnaes.2013.07.010

1872-2032/© 2013 Ecological Society of China. Published by Elsevier B.V. All rights reserved.

Acta Ecologica Sinica 34 (2014) 325–336

Contents lists available at ScienceDirect

Acta Ecologica Sinica

j o u r n a l h o m e p a g e :   w w w . e l s e v i e r. c o m / l o c a t e / c h n a e s

mailto:zhangguangfu@njnu.edu.cnhttp://dx.doi.org/10.1016/j.chnaes.2013.07.010http:///reader/full/http://www.sciencedirect.com/science/journal/01678809http://www.elsevier.com/locate/CHNAEShttp://crossmark.dyndns.org/dialog/?doi=10.1016/j.chnaes.2013.07.010&domain=pdfhttp://www.elsevier.com/locate/CHNAEShttp:///reader/full/http://www.sciencedirect.com/science/journal/01678809http://dx.doi.org/10.1016/j.chnaes.2013.07.010mailto:zhangguangfu@njnu.edu.cn

8/18/2019 1-s2.0-S1872203214000547-main

2/12

methodology for studying plant defenses against herbivores. Nu-

merous theoretical models and hypotheses, which are influential

in the international community concerning the relationship between

plants and herbivores, were introduced in the present review. They

can be grouped into four categories according to plant defense strat-

egies; meanwhile most of them were reviewed within each category.

Finally, we also reviewed the research progress of plant–herbivore

interactions in China, and then discussed the perspectives of studies

on plant–herbivore interactions to provide a theoretical basis for

our future research.

2. Categories of herbivores

 2.1. According to zoological classification criteria

According to zoological classification criteria, herbivores can be

divided into herbivorous vertebrates and invertebrates. Most part

of the former is generally herbivorous mammals (mainly ungu-

lates), which is widely recognized as an important factor in

maintaining the biodiversity of grasslands [12–14]. Meanwhile the

latter mainly consists of Arthropoda (including herbivorous insects

and crustaceans) and Mollusca (usually Gastropoda, such as snails,

slugs, etc.). Initially many authors reported important relation-

ships between mammalians and plants. By contrast, little attentionwas paid to the role of invertebrate herbivores in shaping plant com-

munity and population dynamics. However, such studies becoming

a great part of ecology have been well documented in the litera-

ture in the past decades. Many studies have demonstrated that

invertebrate herbivores have an important effect on secondary suc-

cession of plant communities [15–20].

Most mammals and mollusks feed on plant seedlings while

insects do great damage to adult plants in the field. Interaction

between plants and insects from different forest ecosystems has been

widely carried out. Because of their different mouthparts, leaf 

damages by insects include chewing, skeletonizing, insect galling,

mining, rolling, and sucking [21]. Plants and insects comprise most

part of the organisms on Earth, and their interactions have pro-

found implications not only for both ecological and evolutionaryprocesses [22–24], but also for ecosystem nutrient cycling and energy

flow [22,25]. Currently, researches on interactions between mol-

lusks and plants are not as many as those relationships between

insects and plants, but most studies on mollusk herbivory have sug-

gested that mollusks, consuming little biomass, do enhance seedling

mortality of subdominant herbs [14,26]. Generally mollusks are likely

to feed on seedlings instead of adult plants, causing a great influ-

ence on plant individuals which is disproportionate to the biomass

removed [27,28]. For example, a mollusk can kill a whole seedling

with the removal of one bite of the hypocotyl while a similar bite

to a mature leaf would have a negligible effect on the survival of 

the plant [29]. Therefore mollusks have a great impact on commu-

nity composition of herb layer [30,31], especially for seedlings since

their establishment is the crucial point in a species’ life cycle [31–33].

Lodge [34] once pointed out that aquatic herbivores had little

effect on the aquatic plants. However, studies hereafter have shown

that aquatic herbivores have a strong impact on aquatic plant

biomass [35] and species composition [36]. As common aquatic her-

bivores, some snails (from Gastropoda) and crustaceans (from

Crustacea), like crayfishes, are distributed widely in the field. Now-

adays, most of them have been applied as generalist herbivores

during bioassay experiment to elucidate the relationship between

aquatic herbivores and plants [8,37–39].

 2.2. According to herbivores’ preference for plant species

According to herbivores’ preference for plant species, herbi-

vores can be divided into generalists and specialists (includingoligophagous and monophagous). Most herbivorous mammals and

mollusks usually belong to generalist consumers while most plant-

eating insects belong to specialist consumers [39]. For example,

crayfishes (from crustacean) are often used as generalist herbi-

vores in experiments (Fig. 1). Generalist herbivores refer to animals

that can feed on most plants, and will not give up feeding on some

certain plant species because of their special feeding preferences.

Plant defenses and natural enemies are widely believed to be the

main reasons why specialist herbivores only rely on one food source

[3].

In view of the fact that generalists and specialists have differ-

ent dependence and effect on plants [40], there are two viewpoints:

some scholars believe that generalist consumers have greater effects

on plant fitness and community composition [8,39,41,42]. On thecontrary, other scholars hold that specialist consumers cause more

damage to plants because they have superiority to generalists in food-

searching and food utilization, food location and detoxifying, with

fewer chance of being exposed to natural enemies  [43]. Such dis-

parity may be attributed to the different research content which is

Herbivorous vertebrates

Herbivorous

Mammals (mainly)

Insecta

Arthropod 

Molluscs

Generalist herbivores

Specialist herbivores

(Oligophagous and monophagous)

Crustacea

Herbivore’s preference for plants

Zoological classification criterion

H e r  b i   v or  e  s 

Fig. 1.  Categories of herbivores.

326   G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

3/12

focused on generalists or specialists, as well as to the evolutionary

stage of plants.

In temperate forest ecosystems, generalists take up a large pro-

portion of herbivores while in tropical forest ecosystems, specialists

take up a large proportion of herbivores. Basset  [44] pointed out

that this may be related with higher leaf palatability in temperate

forest. Take for example the toughness of mature or immature leaves

in tropical forest, which was twice as much as that in temperate

forest; however, both nitrogen and water contents of mature leaves

in tropical forest were significantly lower than those in temperate

forest [45].

3. Plant defense strategies

In general, herbivores have negative impacts on plant fitness

[46,47]. The damage caused by them is a kind of natural selection

pressure, which enables plants constantly to develop effective defense

strategies against herbivores in the long-term interactive and co-

evolutionary process. As a rule, each plant species has more than

one defense characteristics. Based on the resource allocation tradeoffs

between different body parts, plant species may well have various

defense characteristics coexisting in different individuals instead of 

investing all defensive features in one individual. Agrawal and

Fishbein (2006) predicted that a continuum of anti-herbivory defensecontained three types of syndromes: (1) poorly defended plants with

phenological escape mechanisms; (2) plants with nutritious, edible

leaves having physical and chemical defenses; and (3) plants with

tough and inedible leaves [48]. In addition, plants can change their

defense features across life-history stages to meet the require-

ments of resource allocation.

In this paper, in light of the latest classification posed by Boege

et al. [49], plant defense strategies will be divided into the follow-

ing three categories: resistance traits, tolerance mechanisms and

escape strategies (Fig. 2).

 3.1. Resistance traits

Resistance traits include physical features (e.g., trichomes, spines,thorns or leaf toughness), chemical features (main secondary me-

tabolites), or biotic features (e.g., maintaining or enhancing the

activity of natural enemies of herbivores), and all these traits can

be used to reduce the amount of damage from herbivores.

 3.1.1. Physical resistance traits

Physical resistance traits refer to morphological or structural

modifications that plant species make when attacked by herbi-

vores [50]. One of the most possible parts is leaf blade owing mainly

to its structure, such as trichomes, LMA (leaf mass per area), thick-

ness, texture and cell structure [51]. Among them, LMA, which is

often used as indicators of leaf physical defense, is one of the most

widely measured functional traits [52,53]. In addition, leaf thick-

ness is considered to be the best effective defense measure [54]

because many studies have shown that leaf toughness is negative-

ly correlated with herbivory [24,49,55,56]. It is worthy to note that

some plant species own physical mimicry in structure, like some

orchid flowers, which are able to mimic bees or wasps to deter large

herbivorous mammals and insects  [57]. However, such kinds of 

flower traits may positively or negatively influence foraging pref-

erences of pollinators to a great extent. For instance, root herbivory

might positively influence pollinator behavior; nevertheless, her-

bivore damage to leaves and flowers might negatively affect foraging

preferences of pollinators [58].

 3.1.2. Chemical resistance traits

Chemical resistance traits refer to physiological modifications that

plant species make when attacked by herbivores [50], which chiefly

involve a great variety of plant secondary metabolites (PSMs). The

total number of PSMs whose structures have been elucidated is about

50 000, and this is only a small fraction of all PSMs existing in nature

[59,60].   According to Kang  [61],  the chemical defense compo-

nents, which are produced by plants against herbivores, can be

divided into seven categories: terpenoids; phenolic compounds; ni-

trogen compounds; tannins, lignin and cellulose; plant hormones

and lectin; protease inhibitors; and volatile compounds. Of all, phenol

and terpene are secondary metabolites with carbon but without ni-

trogen, and both of them are made to defend herbivores when there

is redundant carbon in plants. The total phenolic within a plant in-

dividual can be used as indicators showing chemical defense capacity,

and the content of tannin and protein is correlated with carbon or

nitrogen-based plant defense [51]. Studies concerning other sec-

ondary metabolites such as saponin, alkaloids, amino acids, cyanide

and other studies are still very few. In recent years, many scholars

have focused on the effect of plant enzymes (such as amino acid

degrading enzymes, proteases, etc.) on herbivores after feeding. Someamino acids cannot be synthesized by herbivores, and therefore they

must be obtained from the diet. If these essential amino acids are

destroyed by plant enzymes in the gut of herbivores, their devel-

opment will be impaired [60]. The role of anthocyanins in plant

defense against herbivores has been a disputed topic for a long time,

in which some scholars believe that the development of anthocy-

anin is a response against pathogens [54]. Furthermore, some flowers

are able to emit carrion and dung odors, an olfactory mimicry of a

Resistance traits

Tolerance mechanisms

Escape strategies

Physical resistance (Leaf thickness, leaf texture, physical mimicry, etc.)

Chemical resistance (Tannins, alkaloids, chemical mimicry, etc.)

Biological resistance (Mututalistic and non-mututalistic indirect defenses)

Spatial escape

Increases in photosynthetic area, stored resources, bud bank, etc.

Temporal escapeLag time in the herbivore colonization of young

trees and phenology, etc.

Associational resistance and Janzen-Connell

hypothesis, etc.

 S  t  r  a  t   e  gi   e  s  of   pl   a n t  

Fig. 2.  Categories of plant defense against herbivores (Adapted from Reference 49).

327G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

4/12

danger of predators, through chemical mimicry defense to deter the

attack of mammalian herbivores [62].

 3.1.3. Biological resistance traits

There are two types of chemical resistance traits. (1) Indirect de-

fenses involving defensive mutualisms. It means that plant species

are able to attract herbivores’ natural enemies (predators and para-

sitoids) by providing food rewards, nesting space or chemical cues

so that they can defend themselves against herbivores. The most

famous case is myrmecophytism which involves plant–ant

mutualisms. In this case, the myrmecophytic plants provide nesting

space for ants, and sometimes offer extrafloral nectar or nutri-

tious food [63]. In return, ants protect host plants from herbivores

[64–67]. Such defensive mutualisms also occur between wasps and

plants, and in these cases the plant provides nectar for wasps as a

reward to obtain protection  [68]. (2) Non-mututalistic indirect

defense. In some cases, plant species are defended by animals that

are not engaged in mutualistic interactions with the plant. Com-

pared with the first type, this type has small probability of 

occurrence. van Bael et al. (2003) investigated how predators af-

fected herbivore abundance and levels of herbivory in saplings and

adult trees in three tropical tree species. By using cages to prevent

access to bird predators, they found that the practice of caging sig-

nificantly decreased herbivore abundance and levels of herbivoryon trees but had no effect on saplings [69]. Boege and Marquis [70]

compared the effect of caging and non-caging on saplings and mature

trees, and found that the foraging intensity of bird predators in

mature trees was significantly higher than in samplings. Further-

more, Boege [71] noted that foraging of parasitoid wasps was almost

restricted to the canopies of mature trees in rain forest.

 3.2. Tolerance mechanisms

Tolerance mechanisms are defined as the capacity of plants to

reduce the negative effects of damage on fitness  [72]. During the

long-term evolution, plants are likely to establish tolerance mecha-

nisms to reduce the impact of herbivore damage once it has occurred.

Early research on this respect regarded plant tolerance as a part of defense mechanism [73], but later studies have shown that there

is a tradeoff between tolerance and defense [74–76]. Although plants

are able to resist herbivores through defense mechanisms, they can

hardly reduce the damage. By contrast, tolerance mechanisms can

enable plants to compensate or replace damaged tissues, e.g., en-

hancing photosynthetic efficiency, activating dormant meristems,

making use of reserved resource, changing resource allocation mode,

etc. Generally, seedlings that experience significant reductions in

growth following herbivory are unable to compensate for lost tissues.

Later on as juveniles develop from seedlings to saplings, they can

show significant increases in compensation for herbivory, proba-

bly due to increases in photosynthetic area, stored resources, and

abundant bud bank. Finally, mature plants may experience lower

compensation for herbivory than saplings because of having rela-tively more senescent leaves [49]. Besides, on the basis of tradeoff 

between defense and tolerance, plants at the stage of reproduc-

tion will have much more tolerance than vegetative growth  [76],

because they will allocate more resources to their reproduction with

low investment in other respects. Though many a scholar assume

that compensation mechanisms may enable plants to increase their

fitness, there are few related studies and it remains unclear which

mechanism is the most important [77]. Therefore, plant tolerance

mechanisms still need further study. For instance, injured plants can

increase their photosynthetic efficiency, but are unlikely to achieve

the same level as the uninjured plants. In fact, the lack of knowl-

edge about the mechanisms of tolerance has constrained the study

of tolerance in real ecological condition, limited the experimental

and genetic manipulation of tolerance to herbivory indoors, and

impeded understanding the role of environmental factors and genetic

backgrounds in tolerance research [72]. In view of the fact that there

are much more theoretical reasons than experimental explana-

tions, experimental researches should be strengthened in order to

supplement and improve the tolerance mechanisms of plants.

 3.3. Escape strategies

The reason why plants develop escape strategies during evolu-

tion is to reduce the probability of plants to be found by their

consumers. There are a variety of escape strategies, including as-

sociational resistance, distance from conspecific trees, lag time in

the herbivore colonization of young trees, phenology, and limited

access to trees as they grow. All these strategies often show bio-

geographical variation of patterns.

It seems that associational resistance is especially important for

seedlings which lack their own resistance or tolerance to her-

bivory [78]. Associational resistance occurs when highly susceptible

plants grow close to well-defended plants so as to escape from her-

bivory. Such plants are called nurse plants, providing physical defense

[78–80] and chemical defense [81] for small and/or young plants

that obtain associational resistance from neighbors  [82].

Spatial distance between individuals within species is also im-

portant for escape from herbivory. As proposed by the Janzen–Connell hypothesis [83,84], seedlings and juveniles growing close

to conspecific adults may suffer more damage from specialist her-

bivores than those adult trees, resulting in relatively high mortality

near adults and a decrease of herbivory and mortality with dis-

tance from conspecific adults. This pattern occurs commonly in

tropical forests. In contrast, tree species tend to have higher sur-

vival nearby conspecific adults in temperate forests because parent

trees can produce genetically variable offspring. Meanwhile com-

pared to the adult trees, the surviving seedlings have distinct

secondary chemical components, and herbivores usually con-

sumed only those seedlings with components similar to their parent

trees, leaving chemically differentiated seedlings to survive.

According to associational resistance and Janzen–Connell hy-

pothesis, plant species can escape from herbivory in terms of space.Likewise, they can also do so in terms of time, such as changes in

leaf phenology. Specifically, most plants can reduce their damage

from herbivores through the following three main methods: early

leaf expansion, synchronous leaf expansion, and rapid leaf expan-

sion [85,86]. Another pattern is that in tropical forests many species

can resort to delayed greening to escape herbivores [3].  Further-

more, recent studies have demonstrated that   Arabidopsis

synchronizes jasmonate-mediated defense with circadian behav-

ior of cabbage loopers (Trichoplusia ni) to protect themselves from

herbivory [87].

4. Methodology of plant defenses against herbivores

There are numerous research approaches and techniques con-cerning plant–herbivore relationships which deal with different

interdisciplines and is conducted at different scales. Overall, current

methodology covers the following three key issues for plant de-

fenses against herbivores.

4.1. Molecular genetics

Plants have evolved a great number of defensive characteris-

tics for resistance or tolerance to herbivory, due to the marked

selection pressure produced by herbivores to plants. Accordingly,

to reveal the molecular mechanisms of plant defense, traits can con-

tribute to further understanding of the evolution of plant defense.

Ecogenomics, a new interdisciplinary approach proposed in recent

years, plays an increasingly important role in revealing molecular

328   G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

5/12

mechanisms of plant defense traits. It integrates across disciplines

including evolutionary biology, ecology, and genetics so it can be

adopted to explain genetic and phenotypic variation in intraspe-

cies and interspecies, and to analyze the genetic structure of plant

defense traits. Its common methods mainly include quantitative trait

loci (QTL) mapping, transcription profiling, population genomics and

transgenic approaches [88].

In addition, epigentic variation, other than phenotypic varia-

tion, is an alternative important resource of plant variations because

plants can produce noticeable defensive phenotypes in different gen-

erations under the attack of herbivores or pathogens. Some chemical

modifications such as DNA methylation, chemical modification of 

histones and siRNA may cause transgenerational defense initia-

tion of plants. And it has been proved that this kind of 

transgenerational defense can last in many generations [89]. Inview

of the fact that DNA sequence variation and epigenetic variation

covary in most natural systems, it becomes much difficult to test

the phenotypic effects of epigenetic variation of plant species. At

present, there are four testing methods: to study natural epialleles,

to manipulate DNA methylation, to study systems that naturally lack

DNA sequence variation, and to study epigenetic recombinant inbred

lines [90]. Recent epigentic research with respect to plant defense

suggests that there are three approaches adopted widely in exper-

iment: bisulfite conversion method, affinity chromatography andimmunoprecipitation [89].

4.2. Chemistry and physiology

Many secondary metabolites are essential for the survival of the

plant individual [91]. Among them, tannin and jasmonic acid are

widely used in plant defense. In most cases, tannin is considered

to be one of the most important secondary metabolites in plant de-

fenses against herbivores. Tannin is changeable in molecular

structure, and hence there are a variety of tannins in plants. Poly-

condensation tannin is easy to be tested, and its commom measuring

approach is spectrophotometric method. With the ability of pro-

ducing a variety of hydrolyzate by hydrolysis, hydrolyzed tannin is

usually measured by the rhodanine assay, the sodium nitrite methodand the modified potassium iodate technique. And oxidized tannin

is generally measured by Forint-phenol colorimetric test  [91].

 Jasmonic acid also plays an important role in plant chemical de-

fenses. It can be produced by plants attacked by herbivores, leading

to decreased photosynthetic electron transport and gas exchange,

and inducing plants to produce other secondary metabolites. In this

way, plants thereby can resist herbivores. Such methods as gas ex-

change, chlorophyll fluorescence and thermal spatial patterns are

generally applied to test the defensive effects of jasmonic acid [92].

Recently, the reserach of  Arabidopsis thaliana defenses against two

herbivores has indicated that jasmonic acid and salicylic acid path-

ways may well have considerable interaction effect, and its research

method is mass spectrometry [93].

4.3. Community and ecosystems

Currently, at the community and ecosystem scales, research ap-

proaches concerning the relationships of plants and herbivores are

field observations and indoor control experiments. Field observa-

tion is one of the most efficacious methods of plant defense

researches in forest ecosystems, which has two categories, namely,

discrete random sampling and continuous fixed-site observation

[22,94,95]. With the recent advance in data collection, field obser-

vation method has improved. For example, each target tree species

can be classified into two categories: observation trees and sam-

pling trees. For each observation tree, several randomly selected

branches are located and then herbivore damages are recorded at

fixed periods through continuous observation. At the same time for

each sampling tree, certain leaves are collected and leaf traits are

measured respectively to explain the plant defense characteristics

[51,56,96]. In this way, the interaction of plants and herbivores can

be analyzed. This method can truly reflect the plant survival and

plant defense against different herbivores in the field, but seems dif-

ficult to explain the effect of a specific herbivore on plants.

Consequently, this would limit the application of relationships

between plants and herbivores to solve ecological problems. In con-

trast, experimental manipulation in laboratory can make up the

shortfall in field observation to a great extent. This method is used

to assess the anti-herbivore traits in laboratory by calculating pal-

atability index (PI) of each plant species. Specifically, this approach

is devised as follows: first, to use insects, slugs, or other generalist

animals to make bioassay experiments; second, to measure leaf traits

of the plants; then to analyze the correlation of PI with leaf traits

[8,38,39,97]. This method is appicable to those generalist herbi-

vores that are small in size, move slowly, with a limited range of 

activity. Currently, there has been little research in which such gen-

eralists are used as tested animals in China, suggesting that the need

for making such studies is becoming much urgent.

5. Theoretical models and hypotheses concerning interactions

between plants and herbivores

The interactions between plants and herbivores are related with

plant invasion, ontogeny, dynamics and evolution of plant popu-

lation and community, and therefore numerous hypotheses and

theoretical models have been proposed by a large number of re-

searchers at different times over the past few decades (see  Table 1).

Some of them seem contradictory, resulting from the various

ecosystem-types, plant communities at different successional stages

from which plant species are sampled, herbivores varying from in-

vertebrates to vertebrates in different research papers. For example,

Xiong et al.  found that a native generalist snail (Radix swinhoei)

showed preference for feeding on native aquatic plants from local

lakes [8]. On the contrary, experiments from Morrison and Hay [39]

demonstrated that local snails, as generalist herbivores, preferred

consuming exotic aquatic plants. The contradictory results can bemainly ascribed to the fact that the tested plant species came from

totally different stages of succession, and that plants from early stage

were more susceptible to herbivory since during that time they had

not yet evolved defensive characteristics effective enough to resist

herbivores. The experiment conducted by Parker and Hay[38] is also

contrary to Xiong et al., owing probably to a different animal taxon

which was used as a generalist. In fact, they used crayfish

(Procambarus spiculifer ) which was a generalist herbivore, instead

of snail. Last but not least, lacking a unified theory concerning plant

defenses at present is an underlying reason, leading to miscella-

neous seemingly conflicting hypotheses. Indeed, every hypothesis

or theoretical model plays an important role within its own field.

5.1. Aspects of resistance traits

The hypotheses which can be applied to aspects of resistance

traits are the defensive mimicry hypothesis, the carbon–nutrient

balance hypothesis, the oxidative stress hypothesis, the defensive

mutualisms hypothesis, the natural enemies hypothesis, the biotic

resistance hypothesis, the adaptive convergence hypothesis, the

optimal defense theory, and the induced defense theory.

The defensive mimicry hypothesis is applicable for physical re-

sistance. It refers to the fact that some plants can efficiently resist

against herbivores by mimicking the shape of animals or leaf trace

after chewing by herbivores, and by emitting carrion and dung odors

[98]. It has been almost neglected for a long time except that plant

defensive Batesian mimicry was believed to have an effect on plant

pollination based on field observation or theoretical reason. But

329G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

6/12

recently, more and more evidence has indicated that plant mimicry

plays a significant role in plant defense   [57]. Studies show that

flowers of plants in several orchid genera have the ability to mimic

female bees, by producing similar pheromone or pretending to be

them in appearance. Interestingly, what attracts the male bees to

pollinate is the specific chemical mimicry, rather than the flower

color or appearance polymorphism [114] which can cause the male

bees to misrecognize the deceptive flowers, leading to lack of pol-

lination [115,116]. In fact, recent studies indicate that mimicry flower

in color or shape is crucial for orchids to prevent herbivorous

mammals and insects from attacking [57].

The carbon–nutrient balance hypothesis and the oxidative stresshypothesis are applicable for chemical defense. The former means

that under the condition of high light and lower nitrogen, plant

species can invest excess carbon in plant defense if carbohydrates

are more than requirements for development. If not, plants will

reduce their defense with decreasing carbohydrates [99]. The latter

means that oxidative activation of phenolics in ecological interac-

tions can be used to explain plant defense at the levels of individuals

and ecosystems, and that measurements of oxidative conditions can

improve predicting the activity of phenolic derivatives [100].

The defensive mutualisms hypothesis, the natural enemies hy-

pothesis, and the biotic resistance hypothesis are applicable for

biological resistance. The defensive mutualisms hypothesis refers

to traits facilitating the visitation or colonization of mutualistic

animals that defend the plants against herbivores, such as plant foodrewards, nesting space or chemical cues that can attract herbi-

vores’ natural enemies (predators and parasitoids) [64]. This is one

of the most famous indirect defenses. Seedlings and juveniles are

susceptible to herbivory because they can hardly provide extra nectar

or nesting space. Therefore, this hypothesis is probably not appli-

cable to young plants. Instead, plants during juvenile stage will rely

mainly on direct defense [67,117]. Researchers propose different hy-

potheses concerning how to explain the successful invasion of exotic

plants, of which the enemy release hypothesis and the biotic re-

sistance hypothesis are influential and contradictory. The enemy

release hypothesis postulates that non-native plants entering novel

environments will escape their co-evolved, native enemies and that

this escape may free resources and facilitate the spread of exotic

plants [47]. The biotic resistance hypothesis contends that native

species can function as natural enemies (consumers, pathogens, com-

petitors) of non-native invaders and suppress their establishment

and spread in the new habitat [101]. Accordingly the effects of her-

bivores on the invading plants may be determined by the net effect

of escaping old herbivores and obtaining new ones [39]. Upon in-

vading a new habitat, a non-native plant will escape many specialist

herbivores from its previous habitat (enemy release), but it may also

encounter many new generalist herbivores to deter (biotic resis-

tance). Therefore, this net effect may depend to a great extent on

the relative impact of generalist versus specialist herbivores on plant

fitness [118]. Apparently, however, all these disputes seem ulti-

mately to boil down to the question of whether generalist orspecialist herbivores have more effect on plants. If generalists play

a more significant role, the leading mechanism is the biotic hy-

pothesis; otherwise, it is the enemy release hypothesis.

The adaptive convergence hypothesis, the optimal defense theory,

and the induced defense theory may be involved in more than one

type of resistance traits. The adaptive convergence hypothesis means

that association with specific ecological interactions can result in

convergence on suites of covarying defensive traits. It predicts that

plant defense traits can consistently covary across species, due to

shared evolutionary ancestry or adaptive convergence [48]. Accord-

ing to its characteristics, this hypothesis can be applicable to the

species-rich plant communities, as well as closely related species

in different evolutionary lineages. Agrawal and Fishbein found that

three different shrub communities shared similar plant defenses withconvergence characteristic [48]. The optimal defense theory and the

induced defense theory are very important within the field of 

induced plant defense. The former means that the susceptible plant

tissues contain high constitutive defense and low induced defense

while insusceptible ones contain low constitutive defense and high

induced defense [102]. The latter means that compared to unaf-

fected plants, the initially attacked plants by herbivores are induced

to develop new plant defense, enabling other plants to effectively

resist subsequent attacks [103].

5.2. Aspects of tolerance mechanisms

There are two theoretical model and related hypotheses con-

cerning tolerance mechanisms due to few current researches. Firstly,

 Table 1

The major theories or hypotheses of interactions between plants and herbivores.

Application Name Literature source

Resistance defense The defensive mimicry hypothesis Benson et al. (1975) [98]

The carbon–nutrient balance hypothesis Bryant et al. (1983) [99]

The oxidative stress hypothesis Appel (1993) [100]

The defensive mutualisms hypothesis Janzen (1966) [64]

The natural enemies hypothesis Crawley (1997) [47]

The biotic resistance hypothesis Maron and Vila (2001) [101]

The adaptive convergence hypothesis Agrawal and Fishbein (2006) [48]

The optimal defense theory Feeny (1975) [102]

The induced defense theory Agrawal (1998) [103]

Tolerance mechanisms The compensatory continuum hypothesis Maschinski and Whitham (1989) [104]

The limiting resource model Wise and Abrahamson (2005) [105]

Escape strategies The Janzen–Connell hypothesis Janzen (1970) [83]; Connell (1971) [84]

The trees and grazer satiation hypothesis Silvertown (1980) [106]

The herbivore-adaptation hypothesis Rathcke (1985) [107]

The plant-predictability hypothesis Rathcke (1985) [107]

The slow-growth-high-mortality hypothesis Clancy and Price (1987) [108]

Others The switching of defensive mechanisms during ontogeny Boege et al. (2011) [49]

The escape/defense continuum hypothesis Kursar and Coley (2003) [24]

The co-evolution hypothesis Ehrlich and Raven (1964) [109]

The growth rate model Hilbert et al. (1981) [110]

The resource availability hypothesis Coley et al. (1985) [9]

The grazing optimization hypothesis Williamson et al. (1989) [111]

The g row th–differentiation ba lance hypothesis Herms and Mattson (1992) [112]

The nutrition hypothesis Koyama et al. (2004) [113]

330   G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

7/12

the compensatory continuum hypothesis refers to a continuum of 

compensatory responses to vertebrate herbivory, depending on plant

competition, nutrient availability, and timing of grazing [104]. Sec-

ondly, the limiting resource model means that confronted with

consuming of herbivores, plant species will have seven pathways

to three potential outcomes: greater tolerance, equal tolerance, or

lower tolerance in low- vs high-resource environments [105].

5.3. Aspects of escape strategies

The hypotheses which can be applied to aspects of escape strat-

egies are the Janzen–Connell hypothesis, the trees and grazer

satiation hypothesis, the herbivore-adaptation hypothesis, the plant-

predictability hypothesis, and the slow-growth-high-mortality

hypothesis. The Janzen–Connell hypothesis means that seedlings

and juveniles growing near conspecific adults receive high loads of 

specialist herbivores from nearby adult trees, leading to relatively

high mortality near adults [83,84]. This hypothesis indicates that

plant species can escape from herbivory in terms of space. The trees

and grazer satiation hypothesis means that all trees of one species

within a region can produce large crops of seeds at odd intervals

– mast years – and that seed predators cannot respond fast enough

reproductively, so many seeds survive and sprout [106].

Thereafter Rathcke proposed the herbivore-adaptation hypoth-esis and the plant-predictability hypothesis [107]. The former means

that from the herbivore’s point of view, herbivores should evolve

to consume the most available (i.e., most predictable) plants in their

environments. In other words, predictable plants should be the most

acceptable to herbivores. The latter means that from the plant’s point

of view, the most predictable plants should have the greatest risk

of herbivory and have evolved the most effective defenses. That is,

predictable plants should be the least acceptable to herbivores. Ad-

ditionally, according to the slow-growth-high-mortality hypothesis,

plant species can escape herbivores by prolonging development in

herbivorous insects, which results in greater exposure to natural

enemies such as predators or parasites and a subsequent increase

in mortality [108].

5.4. Other respects

The switching of defensive mechanisms during ontogeny is as-

sociated with resistance defense, tolerance mechanisms, and escape

strategies. Plant species can switch from one defensive strategy to

another as they develop under a variety of environmental condi-

tions, or individuals of the same species in different environments

can take distinct plant defense strategies   [49].  Take arbores for

example; from seeds to adult trees, plants have undergone tremen-

dous change in morphological and physiological features. As a rule,

during seed or seedling stage plant resources mainly rely on en-

dosperm or cotyledons, and at that time plants can hardly allocate

energy to develop defensive characteristics. Therefore, as ex-

plained above, it seems likely that trees can be defended throughassociational resistance, a certain distance away from the same in-

dividuals, production of secondary metabolites or early leaf 

expansion as seedlings, and then switch from escape to secondary

chemistry, physical defenses or tolerance later in development [119].

With the development of seedlings, plants will achieve dominant

position by giving priority to investment in the rapid growth of the

individuals. At this moment, plants may well resort to tolerance

mechanisms. For example, plants then will increase the rate of pho-

tosynthesis or activate dormant meristems if their leaves are

damaged by herbivores. In fact, when seedlings then become ju-

veniles and mature trees, switches between tolerance and chemical

defense are also likely to occur, driven by the risk of attack and re-

source allocation tradeoffs between growth and defense [120,121].

It is found that the level of secondary metabolites probably de-

crease with individual age [53].   When plants reach adulthood,

accumulated plant resource becomes abundant enough to enable

plants to invest more energy in physical defensive characteristics,

such as enhancing leaf toughness, offering extrafloral nectar or

nesting space to attract herbivores’ natural enemies. Thus plants

defend themselves against herbivores through various direct and

indirect defense mechanisms [67,68], depending on different on-

togeny. Plants in different habitats can also switch from one defensive

strategy to another, or change their resource allocation. For in-

stance, the indirect defense of adult plants with canopies was

different from counterparts without canopies. Under the cover of 

canopies the predators of herbivores were much fewer than those

without canopies [67], leading plant individuals with canopies to

invest more resources in defensive measures. The studies by Cates

[122] demonstrated that Asarum caudatum would invest more energy

in defense, otherwise it would invest in growth and reproduction.

When entering a new habitat, plant species would allocate more

resources to develop morphological structure and synthesize chem-

ical composition relative to the consumption of generalist herbivores

[123].

The escape/defense continuum hypothesis is associated with plant

escape and defense. Plant species may take two extreme kinds of 

defensive mechanisms: at one extreme are species with a ‘defense’

strategy, and at the other extreme are ‘escape’ species. Actually, mostspecies fall along an escape/defense continuum in the field  [24].

The co-evolution hypothesis is associated with plant evolu-

tion. It refers to such an evolution of two or more species in which

the evolutionary changes of each species influence the evolution

of the other species [109].

Besides those stated above, there are some other hypotheses pro-

posed to explain the relationship between plant growth and defense,

including the growth rate model, the resource availability hypoth-

esis, the grazing optimization hypothesis, and the growth–

differentiation balance hypothesis. The growth rate model means

that under certain conditions plant growth rate, especially for

aboveground biomass, increases with an increase in grazing inten-

sity; under other conditions, very large increases in relative growth

rate after grazing can occur but the biomass may not increase, oreven less than that of ungrazed plants  [110]. The resource avail-

ability hypothesis refers to cause and effect between intrinsic growth

rate and plant defense characteristics against herbivores. When the

resource is limited, plants with low growth rate will allocate more

energy to resist herbivores than those with high growth rate [9].

The grazing optimization hypothesis suggests that the grazing in-

tensity within a reasonable range will incite plants to enhance their

net productivity [111]. The growth–differentiation balance hypoth-

esis holds that plants have to make a tradeoff between growing fast

and defending herbivores and pathogens efficiently because they

are confronted with the dilemma: to grow or defend during their

development [112].

The nutrition hypothesis can be used to explain the formation

of aphid galls. Koyama et al .   reported that some aphid speciesinduced leaf galls, which had to accumulate high concentrations of 

amino acids and provide the aphids with sufficient nutrients. So

those aphids would have survival advantage over the counter-

parts consuming leaves [113].

6. Current issues and problems of plant defense in China

Researches of plant–herbivore relationships abroad are full-

fledged relatively to counterparts in China. The studies abroad cover

diversified plant defensive characteristics, involving numerous her-

bivores applied in bioassay experiment and a wide variety of 

ecosystem types in the field. Qin  [124]  discussed insect–plant

interactions and their co-evolution in his famous book  The Rela-

tionship between Insects and Plants, which can be recognized as a

331G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

8/12

milestone, causing a significant impact in China. Hereafter one after

another herbivory research cases popped out, including the rela-tion of interaction between insects and plants to evolution [125,126];

anti-herbivore defenses of young leaves in tropical forests  [54];

leaves’ positive effect and their defensive mechanism under the stress

of phytophagous insects [127]; effects of plant on insect diversity

[128]; feeding level of folivorous insects in forest canopy [129]; in-

teractions between herbivores and plant diversity in grassland

ecosystem [130]; new discovery about plant defense: plant–plant

communication [131]. Based on comparing the reviews at home and

abroad concerning plant defense published in mainstream ecolog-

ical journals during the past ten years, it is noted that there are much

fewer papers in China than those abroad, in several years lacking

reviews at home (Fig. 3). Consequently, this suggests that related

researches should be strengthened in China.

6.1. Categories of herbivores

In most cases, recently there have been few herbivorous species

used in plant defense researches at home except phytophagous

mammals like cows [132], grazing sheep [133,134] and goats [135]

from grassland ecosystem or  Radix swinhoei  from aquatic ecosys-

tem [8]. The vast majority of studies in this area have concentrated

on insect-eating patterns, herbivory intensity and dynamics

[2,10,94,96,136–138] , and most of them have focused on forest eco-

system, rather than other ecosystems. In addition, Zhu et al. reported

the effects of large herbivore grazing on meadow steppe plant and

insect diversity [139].

6.2. Plant defense strategies

Studies concerning plant defense strategies focus on forest eco-

systems in China. In tropical areas, Cai and Cao  [54] reviewed the

advances in anti-herbivore defenses of tropical forest plants. In frigid

zone, Deng [140]  reported the effects of volatile chemical sub-

stances on needles of   Pinus massoniana   seedlings. However, the

majority of plant defense studies are conducted in the subtropical

forests. Liu et al. found that there existed effects of early-season her-

bivory on leaf traits of  Schima superba and subsequent insect attack

in Mt. Meihua, southern China. Sun et al. found out a direct rela-

tionship between herbivory and leaf expansion of  Castanopsis fargesii

from evergreen broad-leaved forest in Tiantong National Forest Park

of Zhejiang, China [86]. Liu et al. compared leaf mass per area, pho-

tosynthetic capacity and chemical defense traits of four evergreen

broad-leaved tree species under different light conditions, and dem-

onstrated that there was a tradeoff between physical and chemicaldefense strategies  [51]. Liu et al. found that there was a correla-

tion between leafing phenology and leaf traits of woody species of 

evergreen broad-leaved forests in subtropical China, suggesting that

plants with different leaf size probably took different defense mea-

sures against herbivores [141]. Xia et al. found that leaf herbivory

damage differed between the first and second sets of shoots in five

evergreen woody species from Tiantong National Forest Park of Zhe-

 jiang, China [142].

 Just as stated above, it is obvious that most studies are related

to resistance defense and escape strategies. As far as the contents

are concerned, physical and chemical defenses are the main issues

in these studies of resistance defense; leaf phenological escape is

the main issue in these studies of escape strategies. Actually, there

is little research concerned with tolerance mechanisms, biotic re-sistance pertaining to the category of resistance defense, and other

escape strategies except leaf phenology.

6.3. Research approaches

It seems that no study of plant defense at the genetic level has

been reported in China so far. But at the chemical and physiolog-

ical levels, such approaches of measuring leaf tannin or total phenolic

content are widely used to analyze plant defense strategies [51,143].

Main approaches at the community and ecosystem levels are field

observation of leaf damage by herbivory, coupled with measuring

leaf traits at laboratory   [51,86,96,136,141]. Nevertheless, her-

bivory experiment indoors is particularly few besides the study made

by Xiong et al. in which  Radix swinhoei was cultured as generalistherbivore [8]. Therefore, herbivory experiment indoors and study

of plant defense at the genetic level are extremely insufficient in

China, suggesting that we should make efforts to adopt laboratory

techniques in molecular biology in future research of plant defense.

6.4. Theoretical models and hypotheses

Scholars abroad have proposed a number of theoretical models

and hypotheses concerning the relationships of plants and herbi-

vores over the past few decades, facilitating the development in this

field. Until recently, China has begun research on this area. Most

of the scanty studies at home are carried out to test one or two of 

the hypotheses above. For example, the result of experiment per-

formed by Xiong et al. supported the natural enemies hypothesis

0

2

4

6

8

10

12

14

16

18

2 00 0 2 00 1 2 00 2 2 00 3 2 00 4 2 00 5 2 00 6 2 00 7 2 00 8 2 00 9 2 01 0 2 01 1 2 01 2 2 01 3

   N  u  m   b  e

  r  o   f  r  e  v   i  e  w  s

Year

SCI

CNKI

Fig. 3.  The comparison of review number from domestic and international mainstream ecology journals concerning plant–herbivore interactions.

332   G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

8/18/2019 1-s2.0-S1872203214000547-main

9/12

[8]. It remains unclear whether those models and hypotheses are

applicable to Chinese situations where there are various ecosys-

tem types with rich species diversity. Accordingly, the lack of 

theoretical progress in relationships of plants and herbivores may

restrict in-depth understanding of plant defense mechanisms. There-

fore, it is necessary to carry out theoretical research in this area.

In summary, it seems likely that current herbivory studies in

China are characterized by much few herbivorous animals used in

experiment, only little and poor means and technique, and the lack

of original theoretical models and hypotheses. All of them will con-

strain the development of herbivory research in China. In addition,

other characteristics are as follows: current researches focus on her-

bivory from forest or grassland ecosystem, few from other

ecosystems; herbivory researches focus on species in tree and shrub

layers, few in herb layer; these researches focus on the plant–

insect and plant–mammal relationships, few on the other

relationships. Therefore, the relationships between plants and her-

bivores in China should be strengthened in the future.

7. Research prospects

Considerable progress about plant–herbivore relationships have

been made in China in the past decades. However, there still exists

a big gap in this respect between China and the developed coun-tries in the world. For example, compared with other countries, it

is much later for China to conduct researches concerning interac-

tions of plants and herbivores. Based on the current research status,

we recommend that future researches should be focused on the fol-

lowing aspects:

(1) Selections for experimental herbivores should be diversi-

fied. Different herbivores have different effect on plant species.

In general, vertebrates have an important role in maintain-

ing the biodiversity of grassland, insects can increase plant

community richness by feeding on dominant species to reduce

their productivity, and mollusks may significantly affect the

species richness of herb layer [14]. In view of the fact that

current researches are mainly focused on plant–mammal orplant–insect relationships in China with a high plant and

animal species diversity, therefore we highly recommend that

coming researches of plant–mollusk and plant–crustacean re-

lationships should be stimulated in the future.

(2) Experimental manipulation and field observation should be

combined in practice. Many environmental factors may in-

fluence plant fitness in the wild. Actually, most current

researches are conducted by sampling for one time rather than

continuous observation, so we suggest that it is of impor-

tance to make continuous observation in a fixed site to reflect

the consuming dynamics of herbivores in realty. Besides,

efforts should be made to find out the mechanisms of certain

herbivores to plants. The big advantage of experimental ma-

nipulation in laboratory is that it can make up the shortfallin field observation of interaction between herbivores and

plants. In the present use, field observation is one of the few

research approaches for plant–herbivore relationship in China;

in contrast the lack of experimental manipulation may seri-

ously affect the related researches of plant defense

mechanisms against herbivores. Therefore, a trend in plant

defensive mechanisms seems likely to combine experimen-

tal manipulation with field observation in China.

(3) Researches of plant–herbivore relationships across different

ecosystems should be encouraged in China where there are

diversified ecosystem types with rich flora and fauna, espe-

cially for herbivore types and plant species. However, current

researches of plant–herbivore relationships only focus on those

species from forest or grassland ecosystem, so we should

strengthen interactions of species from water ecosystem and

other systems. In addition, due to human-induced ecosys-

tem perturbations it is worthy to carry out the studies on

how the relationships between plants and herbivores alter

under the man-made treatments and in unaffected nature

within the same ecosystem.

(4) Future researches should also be encouraged to assess the her-

bivory in the scenario of global climate change. Human

activities are severely changing the composition and func-

tion of ecosystems at the global level  [144]. For example,

climate warming and increased atmospheric nitrogen depo-

sition may exert strong bottom-up effects on primary

producers and ecosystems; moreover, it probably depends on

herbivores that respond differently to these changes  [145].

Thus, there is an urgent requirement to improve our under-

standing of the herbivory under climate change.

(5) More attentionshould be paid to plant roots during the studies

of interactions between plants and herbivores. Currently, our

knowledge about the interaction mostly comes from

aboveground herbivory (AGH), rather than the below ground

herbivory (BGH) which has long been neglected in the past

decades. Actually, recent study shows that belowground her-

bivores have substantial damage to the roots by significantly

impacting overall plant fitness. Furthermore, roots play a sig-nificant role in defending against aboveground herbivory.

Roots can be used as useful organ not only to store the toxic

substances, but also photoassimilates, enabling plants to tol-

erate the herbivory. In addition, the interaction between roots

and rhizosphere microorganisms can also affect plant–

herbivore relationships   [146].   Therefore, a better

understanding of the contribution of roots to aboveground

herbivory will shed light on the mechanisms by which plants

and herbivores interact intertwiningly.

 Acknowledgements

This study was supported by the Project for National Basic SciencePersonnel Training Fund (J1103507 and J1210025), Priority Aca-

demic Program Development of Jiangsu Higher Education Institutions

(PAPD), and State Key Laboratory of Palaeobiology and Stratigra-

phy (Nanjing Institute of Geology and Palaeontology, CAS) (No:

083111). We are also grateful to Professor Rodolfo Dirzo for kindly

providing us with some useful articles, and in particular to Profes-

sor Sun Shucun for critically reviewing the abstract.

References

[1]   M.T.J. Johnson, Evolutionary ecology of plantdefencesagainst herbivores, Funct.Ecol. 25 (2) (2011) 305–311.

[2]   Z.G. Liu, Y.L. Cai, K. Li, L. Yang, C. Sun, Insect herbivory patterns on leaves of Castanopsis fargesii during leaf expansion in evergreen broad-leaved forest in

eastern China, Chin. J. Plant Ecol. 33 (5) (2009) 919–925.[3]   P.D. Coley, J.A. Barone, Herbivory and plant defenses in tropical forests, Annu.

Rev. Ecol. Syst. 27 (1) (1996) 305–335.[4]   A.A. Agrawal, Current trends in the evolutionary ecology of plant defence,

Funct. Ecol. 25 (2) (2011) 421–433.[5]   N.Huntly, Herbivores and the dynamics of communities and ecosystems, Annu.

Rev. Ecol. Syst. 22 (1) (1991) 477–503.[6]   J. Cebrian, Role of first-order consumers in ecosystem carbon flow, Ecol. Lett.

7 (3) (2004) 232–240.[7]   J.E. Houlahan, C.S. Findlay, Effect of invasiveplant species on temperate wetland

plant diversity, Conserv. Biol. 18 (4) (2004) 1132–1138.[8]   W. Xiong, D. Yu, Q. Wang, C.H. Liu, L.G. Wang, A snail prefers native over exotic

freshwater plants: implicationsfor the enemy release hypotheses, Freshw. Biol.53 (11) (2008) 2256–2263.

[9]   P.D. Coley, J.P. Bryant III, F.S. Chapin, Resource availability and plantantiherbivore defense, Science 230 (4728) (1985) 895–899.

[10]   H. Jiang, Y.L. Cai, K. Li, H. Wang, L. Wang, Intensity and patterns of leaf areaeaten of  Lithocarpus glaber  by insects, at Tiantong Forest Park, Zhejiang, Chin.

 J. Ecol. 24 (9) (2005) 989–993.

333G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0010http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0010http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0020http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0020http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0025http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0025http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0030http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0030http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0035http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0035http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0040http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0040http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0050http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0050http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0055http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0050http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0050http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0045http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0040http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0040http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0035http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0035http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0030http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0030http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0025http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0025http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0020http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0020http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0015http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0010http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0010

8/18/2019 1-s2.0-S1872203214000547-main

10/12

[11]  E.C. Oerke, H.W. Dehne, Safeguarding production-losses in major crops andthe role of crop protection, Crop Prot. 23 (4) (2004) 275–285.

[12]   S.J. McNaughton, M. Oesterheld, D.A. Frank, K.J. Williams, Ecosystem-levelpatterns of primary productivity and herbivory in terrestrial habitats, Nature341 (6238) (1989) 142–144.

[13]  S.L. Collins, A.K. Knapp, J.M. Briggs, J.M. Blair, E.M. Steinauer, Modulation of diversity by grazingand mowing in native tallgrass prairie, Science 280 (5364)(1998) 745–747.

[14]   E. Allan, M.J. Crawley, Contrasting effects of insect and molluscan herbivoreson plant diversity in a long-term field experiment, Ecol. Lett. 14 (12) (2011)1246–1253.

[15]   H. McBrien, R. Harmsen, A. Crowder, A case of insect grazing affecting plantsuccession, Ecology 64 (5) (1983) 1035–1039.[16]   P.J. Edwards, M.P. Gillman, Herbivores and plant succession, in: A.J. Gray, M.J.

Crawley, P.J. Edwards (Eds.), Colonization, Succession and Stability, The26th Symposium of the British Ecological Society Held with the LinneanSociety of London, Blackwell Scientific Publications, Oxford, 1987, pp. 295–314.

[17]  C.W.D. Gibson, V.K. Brown, M. Jepson, Relationships between the effects of insect herbivory and sheep grazing on seasonal changes in an earlysuccessional plant community, Oecologia 71 (2) (1987) 245–253.

[18]   V.K. Brown, A.C. Gange, Differential effects of above- and below-ground insectherbivory during early plant succession, Oikos 54 (1) (1989) 67–76.

[19]   V.K. Brown, A.C. Gange, Secondary plant succession: how is it modified byinsect herbivory? Vegetatio 101 (1) (1992) 3–13.

[20]   M.E. Hanley, M. Fenner, P.J. Edwards, An experimental field study of the effectsof mollusc grazing on seedling recruitment and survival in grassland, J. Ecol.83 (4) (1995) 621–627.

[21]   D.C. Allen, J.E. Coufal, Introduction to Forest Entomology: A Training Manualfor Forest Technicians, Syracuse University Press, Syracuse, 1984.

[22]   P.D. Coley, Herbivory and defensive characteristics of tree species in a lowlandtropical forest, Ecol. Monogr. 53 (2) (1983) 209–233.

[23]   S.E. Hartley, T.H. Jones, Plant diversity and insect herbivores: effects of environmental change in contrasting model systems, Oikos 101 (1) (2003)6–17.

[24]   T.A. Kursar, P.D. Coley, Convergence in defense syndromes of young leaves intropical rainforests, Biochem. Syst. Ecol. 31 (8) (2003) 929–949.

[25]   C.M. Herrera, O. Pellmyr, Plant Animal Interactions: An Evolutionary Approach,Wiley, Oxford, 2002.

[26]  H. Buschmann, M. Keller, N. Porret, H. Dietz, P.J. Edwards, The effect of sluggrazing on vegetation development and plant species diversity in anexperimental grassland, Funct. Ecol. 19 (2) (2005) 291–298.

[27]   C.M.G. Duthoit, Slugs and food preferences, Plant Pathol. 13 (2) (1964)73–78.

[28]   M. Fenner, M.E. Hanley, R. Lawrence, Comparison of seedling and adultpalatabilityin annual and perennial plants, Funct. Ecol. 13 (4) (1999) 546–551.

[29]  R. Dirzo, J.L. Harper, Experimental studies on slug-plant interactions. II. Theeffect of grazing by slugs on high density monocultures of   Capsella bursa-

 pastoris and Poa annua, J. Ecol. 68 (3) (1980) 999–1011.[30]   M.J. Crawley, Insect herbivores and plant population dynamics, Annu. Rev.

Entomol. 34 (1) (1989) 531–564.[31]  P.E. Hulme, Seedling herbivory in grassland: relative impact of vertebrate and

invertebrate herbivores, J. Ecol. 82 (4) (1994) 873–880.[32]   P.E. Hulme, Herbivores and theperformance of grassland plants: a comparison

of arthropod, mollusc and rodent herbivory, J. Ecol. 84 (1) (1996) 43–51.[33]   U. Scheidel, H. Bruelheide, Effects of slug herbivory on the seedling

establishment of two montane Asteraceae species, Flora 200 (4) (2005)309–320.

[34]   D.M.Lodge,Herbivory on freshwater macrophytes, Aquat.Bot. 41 (1–3) (1991)195–224.

[35]   M. Søndergaard, L. Bruun, T. Lauridsen, E. Jeppensen, T.V. Madsen, The impactof grazing waterfowl on submerged macrophytes: in situ experiments in ashallow eutrophic lake, Aquat. Bot. 53 (1–2) (1996) 73–84.

[36]   E. van Donk, Switches between clear and turbid water states in abiomanipulated lake (1986 – 1996): the role of herbivory on macrophytes,in: E. Jeppensen, M. Sondergaard, M. Sondergaard, K. Christoffersen (Eds.), TheStructuring Role of Macrophytes in Lakes, Springer-Verlag, New York, 1998,

pp. 290–297.[37]   G. Cronin, D.M. Lodge, M.E. Hay, M. Miller, M.A. Hill, T. Horvath, et al., Crayfish

feeding preferences for freshwater macrophytes: the influence of plantstructure and chemistry, J. Crustacean Biol. 22 (4) (2002) 708–718.

[38]  D.J. Parker, E.M. Hay, Biotic resistance to plant invasions? Native herbivoresprefer non-native plants, Ecol. Lett. 8 (9) (2005) 959–967.

[39]   W.E. Morrison, M.E. Hay, Herbivore preference for native vs. exotic plants:generalist herbivores from multiple continents prefer exotic plants that areevolutionarily naïve, PLoS ONE 6 (3) (2011) e17227.

[40]   W.G. Abrahamson, Plant-animal Interactions,McGraw-Hill Inc, New York, 1989,pp. 123–162.

[41] A.A. Agrawal, P.M. Kotanen,C.E. Mitchell, A.G.Power, W.Godsoe, J. Klironomos,Enemy release? An experiment with congeneric plant pairs and diverse above-and belowground enemies, Ecology 86 (11) (2005) 2979–2989.

[42]   J.D. Parker, D.E. Burkepile, M.E. Hay, Opposing effects of native and exoticherbivores on plant invasions, Science 311 (5766) (2006) 1459–1461.

[43]   T.S. Zhang, Insect Diversity in Subtropical Evergreen Broad-leaved Forests[PDR],East China Normal University, Shanghai, 2008.

[44]   Y.Basset, Palatability of treefoliage to chewinginsects: a comparisonbetweena temperate and a tropical site, Acta Oecologica 15 (1994) 181–191.

[45]   P.D. Coley, T.M. Aide, Comparison of herbivory and plant defenses in temperateand tropical broad-leaved forests, in: P.W. Price, T.M. Lewinsohn, G.W.Fernandes, W.W. Benson (Eds.), Plant-animal Interactions: EvolutionaryEcology in Tropical and Temperate Regions, Wiley & Sons, New York, 1991,pp. 25–49.

[46]   J.L. Harper, Population Biology of Plants, The Blackburn Press, Caldwell, 2010.[47]   M.J. Crawley, Plant-herbivore dynamics, in: M.J. Crawley (Ed.), Plant Ecology,

2nd ed., Blackwell Science, Cambridge, 1997, pp. 401–475.[48]   A.A. Agrawal, M. Fishbein, Plant defense syndromes, Ecology 87 (2006)

S132–S149.[49]   K. Boege, K.E. Barton, R. Dirzo, Influence of tree ontogeny on plant-herbivoreinteractions, in: F.C. Meinzer, B. Lachenbruch, T.E. Dawson (Eds.), Size- andAge-related Changes in Tree Structureand Function, Tree Physiology 4, SpringerScience and Business Media, Berlin, 2011, pp. 193–214.

[50]   J.H. Gowda, Physical and chemical response of juvenile Acacia tortilis trees tobrowsing: experimental evidence, Funct. Ecol. 11 (1) (1997) 106–111.

[51]  F.J. Liu, S. Xiang, X.C. Yang, S.C. Sun, Comparison of leaf mass per area,photosynthetic capacity and chemical defense traits of four evergreenbroad-leaved tree species under contrasting light conditions, Chin. J. Appl.Environ. Biol. 16 (4) (2010) 462–467.

[52]   C.H. Lusk, P.B. Reich, R.A. Montgomery, D.D. Ackerly, J. Cavender-Bares, Whyare evergreen leaves so contrary about shade? Trends Ecol. Evol. 23 (6) (2008)299–303.

[53]   B.L. Webber, I.E. Woodrow, Chemical and physical plant defence acrossmultiple ontogenetic stages in a tropical rain forest understorey tree, J. Ecol.97 (4) (2009) 761–771.

[54]   Z.Q. Cai,K.F. Cao, Advances in thestudies on anti-herbivore defenses of tropicalforest plant young leaves, Guihaia 22 (5) (2002) 474–480.

[55]   P.V.A. Fine, I. Mesones, P.D. Coley, Herbivores promote habitat speciali-zation by trees in Amazonian forests, Science 305 (5684) (2004) 663–665.

[56]   E.G. Pringle, R.I. Adams, E. Broadbent, P.E. Busby, C.I. Donatti, E.L. Kurten, et al.,Distinct leaf-trait syndromes of evergreen and deciduous trees in a seasonallydry tropical forest, Biotropica 43 (3) (2011) 299–308.

[57]   S. Lev-Yadun, G. Ne’eman, Does bee or wasp mimicry by orchid flowers alsodeter herbivores? Arthropod Plant Interact. 6 (3) (2012) 327–332.

[58]   D. Lucas-Barbosa, J.J.A. van Loon, R. Gols, T.A. van Beek, M. Dicke, Reproductiveescape: annual plant responds to butterfly eggs by accelerating seedproduction, Funct. Ecol. 27 (1) (2013) 245–254.

[59]   V. De Luca, B. St Pierre, The cell and developmental biology of alkaloidbiosynthesis, Trends Plant Sci. 5 (4) (2000) 168–173.

[60]   M.S. Chen, Inducible direct plant defense against insect herbivores: a review,Insect Sci. 15 (2) (2008) 101–114.

[61]   L. Kang, The chemical defenses of plants on phytophagous insects, Chin. Bull.Bot. 12 (4) (1995) 22–27.

[62]  S. Lev-Yadun, G. Ne’eman, U. Shanas, A sheep in wolf’s clothing: do carrionand dung odours of flowers not only attract pollinators but also deterherbivores? Bioessays 31 (1) (2009) 84–88.

[63]   B. Fiala, U. Maschwitz, Food bodies and their significance for obligate ant-association in the tree genus Macaranga (Euphorbiaceae), Bot. J. Linn. Soc. 110(1) (1992) 61–75.

[64]   D.H. Janzen, Coevolution of mutualism between ants and acacias in CentralAmerica, Evolution 20 (3) (1966) 249–275.

[65]  E.W. Schupp,  Azteca protection of  Cecropia: ant occupation benefits juveniletrees, Oecologia 70 (3) (1986) 379–385.

[66]  B. Fiala, H. Grunsky, U. Mashwitz, K.E. Linsenmair, Diversity of ant-plantinteractions: protective efficacy in  Macaranga species with different degreesof ant association, Oecologia 97 (2) (1994) 186–192.

[67]  M. Heil, D. McKey, Protective ant-plant interactions as model systemsin ecological and evolutionary research, Annu. Rev. Ecol. Syst. 34 (2003)425–453.

[68]   E. Narbona, R. Dirzo, A reassessment of the function of floral nectar in Crotonsuberosus (Euphorbiaceae): a reward for plant defenders and pollinators, Am.

 J. Bot. 97 (4) (2010) 672–679.[69]  S.A. van Bael, J.D. Brawn, S.K. Robinson, Birds defend trees from herbivores

in a neotropical forest canopy, Proc. Natl Acad. Sci. U.S.A. 100 (14) (2003)8304–8307.

[70]   K. Boege, R.J. Marquis, Plant quality and predation risk mediated by plantontogeny: consequences for herbivores and plants, Oikos 115 (3) (2006)559–572.

[71]   K. Boege, Herbivore attack in Casearia nitida influenced by plant ontogeneticvariation in foliage quality and plant architecture, Oecologia 143 (1) (2005)117–125.

[72]  J. Fornoni, Ecological and evolutionary implications of plant tolerance toherbivory, Funct. Ecol. 25 (2) (2011) 399–407.

[73]   R.H. Painter, Resistance of plants to insects, Annu. Rev. Entomol. 3 (1958)367–390.

[74]   J. Fornoni, P.L. Valverde, J. Nuñez-Farfán, Population variation in the cost andbenefit of tolerance and resistance against herbivory in  Datura stramonium,Evolution 58 (8) (2004) 1696–1704.

[75]   R. Leimu, J. Koricheva, A meta-analysis of tradeoffs between plant toleranceand resistance to herbivores: combining the evidence from ecological andagricultural studies, Oikos 112 (1) (2006) 1–9.

334   G. Bin, Z. Guangfu/Acta Ecologica Sinica 34 (2014) 325–336

http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0060http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0060http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0065http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0065http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0065http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0070http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0070http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0070http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0075http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0075http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0075http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0080http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0080http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0085http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0085http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0085http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0085http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0085http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0090http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0090http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0090http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0095http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0095http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0100http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0100http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0105http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0105http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0105http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0110http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0110http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0115http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0115http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0120http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0120http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0120http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0125http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0125http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0130http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0130http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0135http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0135http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0135http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0140http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0140http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0145http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0145http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0150http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0155http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0155http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0160http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0160http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0165http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0165http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0170http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0170http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0170http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0175http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0175http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0180http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0180http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0180http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0185http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0185http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0185http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0185http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0185http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0190http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0190http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0190http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0195http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0195http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0200http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0200http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0200http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0205http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0205http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0210http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0210http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0210http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0215http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0215http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0220http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0220http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0225http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0225http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0230http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0230http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0230http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0230http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0230http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0235http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0240http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0240http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0245http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0245http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0250http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0250http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0250http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0250http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0255http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0255http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0255http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0255http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0260http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0260http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0260http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0260http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0265http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0265http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0265http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0270http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0270http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0270http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0275http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0275http://refhub.elsevier.com/S1872-2032(14)00054-7/sr0280http://refhu