Oviposition preference and larval performance in Polygonia c-album (Lepidoptera: Nymphalidae): the choice between bad and worse
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Ecological Entomology ( 1993) 18, 394-398
S H O R T COMMUNICATION
Oviposition preference and larval performance
in Polygonia c-album (Lepidoptera: Nymphalidae):
the choice between bad and worse
S O R E N N Y L I N and N I K L A S J A N Z Department of Zoology, University of Stockholm,
Stockholm, Sweden
Key words. Butterfly, host plant, B e t h , evolution, history, oviposition,
preference, performance, growth rate.
Introduction (2) the preferred plant may be rare; (3) the poor host plant
may be a recent addition to the habitat, and sclection m a y
As Thompson ( 1988a) remarks, the relationship between not have had time to eliminate oviposition on the poor (or
adult oviposition preference and offspring performance is even fatal; Chew, 1977) plant species; (4) a favourable
the crux of the problem of understanding the evolution of host may grow in unfavourable habitats. Some additional
insect/plant associations. Females often display a hierarchy factors may be of importance in more specific cases.
of preferences for different hosts, so that they use a host Most of the studies on preference-performance relation-
lower in the hierarchy when the preferred species is not ships have been performed on oligophagous species.
available (e.g. Wiklund, 1981; Thompson, 1988b). The However, Nylin (1988) showed that the polyphagous
preference hierarchy may be more or less well correlated comma butterfly, Polygonia c-album L., displays a markcd
with larval performance; here used as a composite term preference, well correlated with larval performance (in
for offspring growth rate, survival and reproduction terms of survival and development time), for hosts in
(Thompson, 1988a). A good correlation suggests that Urticales (including the herbs Urtica and Humulus as well
plant characteristics, including plant chemistry and nutri- as the tree Ulmus), over hosts in Salicales (Salir), Rosales
tional value, is the most important factor that influences (Ribes) and, especially, over those in Fagales (Betulrr,
larval performance, limits host plant range and promotes Corylus). Thus, there is a hierarchy of preference for
specialization on only a few host plants - the general different host orders, similar to the preference hierarchy for
pattern in insects. This is the core of both coevolution species found in Papilio (Wiklund, 1975, 1981; Thompson.
theory and sequential evolution theory for the evolution of 1988b) and other oligophagous butterflies. This work
insect-host plant interactions (Ehrlich & Raven, 1964; provided the basis for the present study. Here, we demon-
Jermy, 1984; Ronquist & Nylin, 1990). Conversely, poor strate that the preference hierarchies and preference -
correlations can be used as evidence by those who consider performance correlations are found also in more rigorously
other ecological factors, such as selection for enemy-free designed experiments, with simultaneous presentation o f
space, to be of great importance in the evolution of host four potential host plants. We then focus on the lower end
plant choice and specialization (e.g. Atsatt, 1981; Bernays of the preference and performance hierarchies: Is there a
& Graham, 1988; see also Strong, 1988). difference in larval performance on two species of the
Thompson (1988a) and Thompson & Pellmyr (1991) lowest-rated genus, Betula? Do females discriminate
review some of the evidence on preference-performance between them, despite the fact that there should be only
relationships, and note that the correlations range all very slight selection in favour of discrimination among
the way from good to poor. Surprisingly often, poor cor- insignificant host plants? In short, has the hierarchy evolved
relations are found (e.g. Chew, 1977; Courtney, 1981; to perfection even at the lower levels and, it' so, how
Williams, 1983; Penz & Araujo, 1991; Valladares & and why?
Lawton, 1991). There is evidence that a number of factors
may cause poor correlations between preference and
performance (Thompson, 1988a; Thompson & Pellmyr, Methods
1991). These include: (1) selection for enemy-free space;
Host plant choice in P.c-album. ( I ) In the earlier study
Corrcspondcnce: Dr Soren Nylin. Department of Zoology, on host plant choice in P.c-album (Nylin, 1988) potential
University of Stockholm, S-106 91 Stockholm, Sweden. hosts were presented together with Urtica rlioicvr, which
394Oviposition preference and larval performunce 395
was used as a reference. A population-level preference for ten healthy larvae (reared o n U.dioica) was placed on
a specific host was calculated as the ratio of total number B.pendula at the beginning of the final (fifth) instar.
of eggs laid by all females on this host over the number of
eggs laid on U.dioica. As a complement and test of this
relative measure of preference, in the present study four Results
hosts were simultaneously presented to five females of
Host plant choice in P.c-album
P.c-album wild-caught in the Stockholm area at the very
beginning of the flight season (after adult hibernation), (1) The results of simultaneous presentation of four
between 4 and 12 May. The plants used were U.dioica, hosts to five wild-caught females of P.c-album (% of all
Ulmus glabra, Salix caprea and Betula pubescens. eggs laid o n each plant) can be seen in Fig. 1. As in a
The females were kept in separate cages of about previous study (Nylin, 1988), where host plant choice was
0.5 x 0.5 x 0.5 m. In each cage, the host plants (freshly measured only relative to U.dioica, the hosts in Urticales
cut twigs of roughly equal size, placed in bottles of water) (U.dioica and U.glabra) were preferred hosts. The host in
were presented at equal distance from the central light Salicales, Xcaprea, and especially the host in Fagales,
source and food source, forming a square in the cage. B.pubescens, were less preferred hosts. There was some
Eggs were counted every 2 days (females of P.c-album variability in the preference hierarchies between females,
almost always lay their eggs singly, so each egg can be seen but four out of five females had U.dioica or U.glabra as the
as representing a choice of host) and a new presentation of most preferred host, whereas the fifth preferred S.caprea.
fresh hosts was then made, in a random manner. This All females had B.pubescens at one of the two lowest
procedure was repeated for the total life-span of the females, places in the hierarchy. This paper is concerned only with
so that almost all eggs laid by these five females were the lower levels of the hierarchy, and we will not here
sampled. discuss correlations between preference and performance
(2) To investigate the extent of discrimination between at the level of individual females. This result was very
the species of Betula, the two birch species Betula pubescens similar to earlier results (Nylin, 1988) and thus it appears
and B.pendula were simultaneously presented in four that painvise choices can b e acceptable alternatives to
different cages to groups of three to five female P.c-album. simultaneous presentations in experimental design. The
The females represented three different stocks originating fofmer technique is useful because it is often impractical
from females caught in the Stockholm area. Eggs were or impossible to present all potential host plant species at
counted after 3 days. A presentation was then made in the the same time.
same cages of only B.pendula for 2 days, whereafter eggs (2) When the two species of Betula were presented
were counted. together all eggs but a single exception were laid on B.
Larval performance in P. c-album. (1) Larvae (offspring pubescens in the four cages (Table 1). When only B.pendula
of the five wild-caught females used in preference exper-
iment I ) were divided among the host plants used in this
experiment. Thirty larvae from each female were reared
singly on each of these plant species. Larvae were reared
until pupation in environmental chambers under long-day
conditions (22h), fifteen larvae from each female at
20-22°C and the remaining fifteen at 26-28°C. For com-
"r
parison with a host plant known to be poor for larval al
growth (Nylin, 1988), a small number of larvae also were
reared on Corylus avellana at both temperatures. Rearings
p
al
Y-
took place in plastic jars in which the host plant could
receive water from a lower jar. Plants were changed at
frequent intervals and whenever they showed signs of c
$20
.-
.)
senescence. Development time and weight at pupation -
m
al
[r
were noted, and from these data and data on hatchling
weights larval relative growth rates were calculated (% 10
mean daily weight gain), i.e.: ((exp ((log (pupal weight)
- log (hatchling weight))/larval time)) - 1) x 100
I I
(2a) Eggs laid on B.pendula and B.pubesceris in pre- 0' I 1
ference experiment 2 were allowed to hatch in situ and U. glabra B. pubescens
larval performance followed during the first instar. (b) U. dioica S. caprea
Forty eggs laid on other hosts were removed and allowed
to hatch in plastic jars, after which the larvae were im- Fig. 1. The relative preference (i.e. the percentage of eggs laid
mediately transferred to B.prndu/a (twenty larvae) and on a particular host) for four host plant species, when presented
simultaneously to females of folygoniu c-album kept singly in
B.puhesc.ens (twenty larvae) and an attempt was made to cages. Data given are means +SE based on N = 5 females. N
rear the larvae individually on twigs of Betula spp. (c) (number of eggs) for females 1-5 were 207, 234, 166. 191 and
To investigate if older larvae could survive on B.pendulu, 323. respectively.396 .Wren Nylin and Niklas Janz
Table 1. Numbcr of eggs laid on the two species of Betula by were high on the preferred hosts U.dioica and U.gluhru,
groups of fcmalc Polygonia c-album of three stocks. P = statistical intermediate on the less preferred host S.iuprcw iund
significance as dctermined by chi2-tests. low o n the species lowest in the preference hierarchy,
B.pubescens (and also on C.avellana; a host for which
Stock Eggs on Eggs on P females also have low preference: Nylin, 198%). As i n
B.pubescens B.pendula
the earlier study, this was seen mostly as variation in
1 48 0Oviposition preferetice and larval performance 397
Table 2. Host plants used by species of Polygonia and related genera. Sources: Seppanen
(1970), Larsen (1974), Scott (1986), Ackery (1988). ‘Dubious’ records (Scott, 1986)
omitted. Butterflies of uncertain species status or with unknown hosts omitted.
Species Host plants
Polygonia c-album Ribes, Ulmus, Urticu. Humulus, Salix, Populus, Betula, Corylus
P .faunus Ribes, Urticu, Salix, Betula, Alnus, Rhododendron, Vaccinium
P.gracilis Ribes, Rhododendron
P.egeu Ribes, Ulmus, Parietaria, Urtica. Corylus
P.progne Ribes, Betula, Rhododendron
P . c-aureum Humulus
P.satyrus Urtica, Humulus, Sulix
P . interrogutionis Ulmus, Celtis, Urtica. Boehmeria, Humulus
P.comma Ulmus. Urticu, Boehmeria, Laportea, Humulus
Kunisku canuce Smilax
Nymphalis vau-album Ribes, Ulmus, S u l k , Populus, Betula, Fagus
N.polychloros Pirus, Prunus, Ulmirs. Celtis, Salix, Populus
N. californica Ceanothus
N.antiopu Resales:* Pyrus, Sorbus; Urticales:* Ulmus, Celtis,
Salicales: Sulix, Populus; Fagales: Betula, Alnus
N.xanthomelas Salix
Agluis urticae Urrica
A . milberti Urtica
Inachis io Urtica
* From these two host orders there are only Nearctic reports. Several other hosts
reported in Scott (1986), also in plant orders other than those included.
choice in this species. Hosts in Fagales, i.e. species of the two species. B.pendula is fatal to larvae, whereas
Betula and Corylus, are the least preferred hosts in the growth on B.puhescens is poor to intermediate (worse
range of hosts used by P.c-album (Nylin, 1988; this study). than on S.caprea, but better than on C.avellaiza). Females
Presumably the reason for this is that these hosts lack the apparently use chemical cues to distinguish between the
special chemical properties (whatever they are) which species of Betula. This is the only reasonable explanation
makes hosts in Salix, Ribes and, especially, Urticales the for the high degree of discrimination against B.pendulu.
preferred hosts for Polygonia spp. (Larsen, 1974; Scott, It is interesting to note that such discrimination has been
1986; Nylin, 1988). Nevertheless, Betula spp. have been able to evolve despite the apparently very minor role of
recorded as a host of P.c-album in the field (Seppanen, 1970; Betula as a host. The penalty for a mistake is obviously
C. Wiklund, pers. comm.) and, as has been shown here, high for each particular egg, but should be of significance
some eggs are occasionally laid on them in the laboratory, for the fitness of ovipositing females only if oviposition on
especially when there is no alternative. To what extent has Betula takes place at a reasonably high rate in the field.
the preference hierarchy evolved to correlation with larval Field studies are needed to clarify this point, but will be
performance at these lowest levels of the hierarchy? very difficult in the case of P.c-album because females are
There are two common tree-forming birch species in the hard to follow for any extended periods of time. At present
western Palearctic, B.pubescens Ehrh. (including the arctic all evidence suggest that Betula is an unimportant host
subspecies tortulosa Ledeb.) and B.pendula Roth (syn. genus for P.c-album (e.g. Pratt, 1986, 1987; Nylin, 1988).
B . verrucosa Ehrh.). The similarities between them, and The probable reasons why total discrimination (and per-
their tendency to form hybrids, have often been stressed fect preference-performance correlation) has been able to
in the past (Linneaus considered them a single species, evolve in the case of P.c-album can be found by reversing
B.ulba L.). As a consequence the two species are often the two causes of poor preference-performance corre-
(although by no means always) lumped in insect host plant lations (see Introduction; Thompson, 1988a; Thompson &
records (e.g. Seppanen, 1970; connerning P.c-album). Pellmyr, 1991) which we find relevant in this case, i.e. the
However, the two species are in fact morphologically and abundance hypothesis (based on optimality criterions)
ecologically rather distinct, and in fact it is now thought and the time hypothesis (based on the assumption of
that about five speciation events divide the two species evolutionary constraints to optimal choice). The abund-
(Roskam, 1985). They belong to different species ‘series’ ance hypothesis states that when the best hosts are rare,
(Pubescetites and Verrucosae),and are more closely related poorer but more abundant hosts should be used instead.
to other species in the Nearctic and in the eastern Palearctic All of the plant species investigated in the present study
than to each other. Other insects are known to distin- are common in Sweden. Moreover, both species of Betula
guish between B.pendula and B.pubescen.s (e.g. Roskam, are very abundant in the typical open forest habitats of
1985; Claridge & Nixon, 1986). Larval performance and P.c-album, which (when coupled with the certainty of
oviposition preference in P.c-alhum is very different on death of offspring o n B.pendula) means that it is likely that308 Siiretr Nyliii wid Niklus Jutiz
females cannot afford to waste an egg on, or even alight study in coevolution. Evolution, 18. .586-(iE3.
on, every B.petidulu they come across. At the same time, Jermy, T. (IYM) Evolution of inscctlhost plant rcliitionships.
B . p i r l x w m s is so abundant, and not poor enough as a American Natiualist. 124, 609-630.
host. that it may pay females not to discriminate against all Larscn, T.B. (1974) Butterfies of Lehunon. National Council for
Brtlllu.
Scicntific Rcscarch, Beirut.
Nylin, S. (1988) Host plant specialization and seasonality in
The time hypothesis states that there may not always
a polyphagous butterfly, Po/.vgoniu c-ulhrim (Lcpidoptcra:
have been time enough for perfect preference-performance Nymphalidac). Oiko.s, 53. 381 -386.
correlations to evolve. In the case of P.c-album, however, Pcnz, C.M. & Araujo. A . M . (1991) Intcraction bctwccn Pupilio
phylogenetic analysis suggests that the species and its hectorides (Papilionidae) and four host plants (Pipcraccac.
ancestors have used all of the studied plants, or their Rutaccac) i n a southern Brazilian population. ./oiunul of
ancestors, for a very long time. This opinion is based Research on the Lepidopteru, 29. 161- 17 I .
on the fact that the same range of plant genera (Urtica, Pratt, C. (1086) A history and invcstigation i n t o the Iluctuatiom
Ulmus, Humulus, Sulix, Ribes, Corylus, Betula) are of Polygonia c-album L: thc comma butterfly. F i t o r ~ i o l o ~ i . s t ' . ~
important hosts for all Polygonia and also for the related Record, 98, 197-203, 244-250.
Pratt, C. (1987) A history and investigation into the lluctuatioiis
Nymphulis, Agluis and Iriuchis (Table 2; e.g. Larsen, 1974;
of Polygonia c-album L: the comma huttcrfly. E t ~ t o t ~ ~ o l ~ ~ g i . s t ' . s
Scott. 1086; Ackery, 1988). Note from Table 2 especially Record. 99, 21-27, 69-80.
t h a t P.[Ult~iit.S, the probable sister-species to P.c-album Rcavcy, D. & Lawton, J.H. (1991) Larval contrihution t o titncw
(Scuddcr, 1889). feed on the same range of plants, including in leaf-cating insccts. Reproductive Behuviour of Iti.si,crs; lndi-
Betulu. These plant genera are neither very closely related, viduals and popularions (cd. by W . J . Bailey and J . Kidsdill-
nor an assortment of the most abundant plants in habitats Smith), pp. 293-329. Chapman & Hall, London.
of Polygoriiu. Therefore the only plausible explanation for Ronquist, F. & Nylin, S. (1990) Process and pattern i n the
this pattern is a long history of associations with these evolution of specics associations. .Svstetnuric~ Z o o l o p , 39.
genera, reaching back at least to the time before the latest 323-344.
several speciations in this butterfly group. Nearer to the Roskain, J.C. (1985) Evolutionary pattcrns in gall midge-host
plant associations (Diptera. Cccidomyiidac). Tijdschrift voor
present, the hosts in Fagales may well have been more
Entornologie, 128, I93 - 2 13.
important as hosts for P.c-ulbum in prehistoric days, when Scott, J.A. (1986) The Rurterflies of North Amerinr. Stanford
Hetitlu and C.avellunu were highly abundant in comparison University Prcss.
with U.gluhru (Southwood, 1961), whereas the 'follower Scudder, S.H. (1889) The Butrerflies of /he Eastern Unired State.s
o f man' U.dioicu and the generally cultivated Humulus and Canadu with Special Reference to New Englutid. Published
Iiq~ulusmust have been much rarer than today. by the author, Cambridge.
Seppancn. E.J. (1970) The Food Plants of the l.arvue of /he
Macrolepidopteru of Finlund. Werner Siiderstriim Osakcyhtiti.
Acknowledgments Helsinki.
Southwood, T.R.E. (1961) The number of species of insects
associatcd with various trccs. Journal of Animul E c d o g v .
We thank Nina Wedell for her help with the rearings. The
30, 1-8.
constructive comments of J. H. Lawton improved the Strong, D.R. (cd.) (1988) Insect host range. Ecologv, 69.
manuscript. This work was supported by grants from the 885-91.5.
Swedish Natural Science Research Council to S.N. Thompson, J.N. (1988a) Evolutionary ccology of the relationship
bctwcen oviposition prefcrence and pcrformancc of offspring i n
phytophagous insects. Enromologia E.tperimentali.s el Rpplicutci.
References 41. 3- 14.
Thompson, J.N. (1988b) Variation i n prefcrcncc and specificity
Ackcry, P.R. (1988) Hostplants and classification: a rcview of i n monophagous and polyphagous swallowtail butterflies.
nymphalid buttcrflics. BiologicalJournal of the Linnean Society. Evolution, 42, 118-128.
33. 9.5-203. Thompson, J.N. & Pcllmyr, 0. (1991) Evolution of oviposition
Atsatt, P.R. (1981) Lycacnid buttcrflics and ants: sclcction for behavior and host preference i n Lcpidoptcra. Anniuil Review
enemy-free space. American Natiirulist, 188, 638-654. of Entomology, 36. 65-89,
Ucrnays, E. & Graham, M. (1988) On thc cvolution of host Valladarcs, G. & Lawton, J.H. (1991) Host-plant sclcction in thc
specificity i n phytophagous arthropods. Ecology, 69, 886-892. holly leaf-miner: docs mother know best? Journul of Animul
Chcw. F.S. (1977) Coevolution of pierid butterflies and thcir E c o l ~ g y60,
, 227-240.
crucifcrous foodplants. 11. The distribution of cggs on potcntial Wiklund, C. (1975) The evolutionary relationship bctwccii adult
foodplmits. Evolution, 31. 568-579. oviposition preferences and larval host plant range i n Popilio
Claridge, M.F. & Nixon. G . A . (1986) Oncopsis flavicollis (L.) machaon. Oecologiu, 18, 18.5- 197.
associatcd with tree birches (Betula): a cornplcx of biological Wiklund, C. (1981) Generalist vs. specialist oviposition bchaviour
species o r a host plant utilization polymorphism? Riologicul in Papilio machaon and functional aspects on the hicr;irchy o f
Journal o f the Linnean Sociery, 27, 381 -397. oviposition prcfcrcnccs. C)iko.s, 36. 163- 170.
Courtney, S.P. ( 1081) Coevolution of pierid butterflies and their Williams, K.S. ( 1983) The cocvolution o f Euphytlrvus chulcedonti
crucifcrous foodplants. 111. An/hocaris curchfmines. Survival, butterflies and thcir larval host plants. 111. Oviposition hchnvior
clrvclopmcnt and oviposition on diffcrcnt host plants. Oecdogia, and host plant quality. Oecologiu. 56. 330-340.
51. 91 -Y6.
Ehrlich. P.R. & Kavcn. P.H. (1964) Butterflies and plants: ii Accepted I June 1993You can also read
