Parasitoid suppression and life-history modifications in a wolf spider following infection by larvae of an acrocerid fly

2012·4.5 MB·English

by  Søren Toft

—
2012. The Journal ofArachnology 40:13-17
Parasitoid suppression and life-history modifications in a wolf spider following infection by larvae of an
acrocerid fly
Soren Toft. Boy Overgaard Nielsen and Peter Fundi: Department of Bioscience. Aarhus University. Ny Munkegade 1 16,
DK-8000 Aarhus C, Denmark. E-mail: [email protected]
Abstract. Flies ofthe family Acroceridae are specialized internal parasitoids ofspiders. We infected hatchlings ofwolf
spiders Pardosaprativaga (L. Koch 1870) (Araneae: Lycosidae) with larvae ofAcrocera orbiculus (Fabricius 1787); most
hostswereinfected byasinglelarva, butothersendured multiple infectionsofuptoeight larvae. Theinfected spidersanda
group of uninfected control spiders were raised in the laboratory for up to 23 weeks. We found that most (81%) spiders
infected by only one larva were able to suppress the infection, whereas most multiple infections (73%) were “successful”
(i.e., a larva emerged or was recovered by dissection, perhaps from a prematurely dead spider). Infected spiders had their
survival reduced in proportion to the infection load, but the reduction was not significant ifthe infection was suppressed.
Infected spiders had higher growth rates than uninfected, and growth stimulation was proportional to the number of
initiallyinfectinglarvaeand independent ofwhetherthe larva wassuppressedornot. Dueto thesepatterns, wesuggest that
growth enhancement results from the spider's mobilization of extra resources for combating the infection rather than
parasitoid manipulation ofspider growth. Spiders with multiple infections took longer to mature than uninfected spiders,
and the pattern of instar durations was changed compared with that ofcontrol and singly infected spiders. As multiple
infections were important for the parasitoid’s success, we suggest that the parasitoid fly’s habit of laying eggs in large
clumps may be an adaptation to increase the chance ofsuccess via multiple infections.
Keywords: Acroceridae, life history, Lycosidae
The interplay between a parasitoid and its host is intriguing. allowed us to obtain eggs and first-instar larvae (planidia)
On one hand, a parasitoid’s survival and growth depend from a few mated females of the species Acrocera orbiculus
entirely on the successful survival and growth of its host (Fabricius 1787) from the field. In the first place this allowed
(Godfray 1994; Brodeur & Boivin 2004). On the other hand, a us to observe its unique mode ofentrance into the spider host
host in excellent condition may more easily avoid a parasitoid (Nielsen et al. 1999); subsequently, we infected a large number
infection in the first place, but also be more effective at of wolf spider hatchlings and thus were able to get some
suppressing a current parasitoid infection. Since a successful information on the fate of the parasitoids after infection, as
parasitoid is deadly, it is always in the host’s interest to get well as on how the life history parameters of the spider are
rid ofit as early as possible. The parasitoid should be virulent modified by the infection. Since some spiders were singly and
enough to avoid suppression from the host, but at the same others multiply infected, we were also able to analyse how
time minimize the physiological costs to the host and parasitoid success and spider life history depend on the in-
selectively manipulate specific aspects of host physiology and fection load. We were interested in how effective the spider
behavior (Godfray 1994; Brodeur & Boivin 2004). Though we host may be in suppressing parasitoid infections, and whether
expect a parasite infection to becostly to the host, thereshould the parasitoid isable to manipulate the spider’s growth pattern
be strong selection on the parasitoid to reduce the negative for its own benefit. Considerable attention has been devoted to
impact of the infection until it finally kills the host (Slansky modifications of host behavior (Godfray 1994; Thomas et al.
1986). If possible, it might even be advantageous for the 2005), also in spiders (Eberhard 2000, 2010), but less attention
parasitoid to stimulate the growth of the host. This would has been paid to how spider life histories are molded by
either provide more resources for the parasitoid that may parasitoid infection and the relative role ofparasitoid and host
obtain a larger body size, or reduce the time of development strategies in these modifications.
and thus enhance the chances of survival. Growth enhance-
METHODS
ment is common for hosts of gregarious parasitoids, whereas
growth inhibition and reduced activity is the rule for hosts of The parasitoid. Acrocera orbiculus has a wide distribution
solitary parasitoids (Slansky 1986; Harvey et al. 1999; Harvey covering both the Palaearctic and the Nearctic (Nartshuk
2000), reflecting a difference in nutritional demands of mul- 2010). Adult females lay clumps ofseveral hundred eggs in the
tiple versus single parasitoids, though this may also depend on vegetation, and the tiny, hatched planidia larvae (0.3-0.4 mm)
the size of the host (Harvey et al. 2010). move about actively searching for a potential spider host.
Flies ofthe family Acroceridae are specialized parasitoids of Having found a potential host, the larva first attaches itself
spiders, mostly cursorial species (Schlinger 1987). Most pub- by the mouthparts, usually to a leg, to first become an ecto-
lished accounts of these animals are lists of parasite-host parasite. It then becomes an internal parasite after the first
relationships and data on spider infection rates based on molt when theamoeboid second instarlarva enters the spider’s
rearings of field-collected spiders. Information on the interac- haemolymph via the tiny hole where the planidium larva’s
tions between the spider host and the developing parasitoid mouthparts are attached (Nielsen et al. 1999). This second
is very limited (Schlinger 1987) because the adult Hies are instar larva enters the abdomen (via legs, prosoma and
difficult to maintain in the laboratory. Chance events have pedicel) and takes up a position near the booklungs. Here it
13
14 THE JOURNAL OF ARACHNOLOGY
grows to its final size, mainly during the third and fourth they had been able to suppress the parasitoid infection. Most
instar. The fully grown larva exits the dying spider to pupate ofthe spiders originally infected with only one larva (30 out of
outside the ca—rcass. 37 = 81%) got rid of it, while this was the case for only a
Procedure. Two females ofA. orbiculus were collected live minority (4 out of 15 = 27%) ofmultiply infected spiders (y2
at Mols, Denmark, and brought to the laboratory. During = 13.96, P < 0.0002). All four spiders originally infected with1
subsequent days they laid a large number of eggs in the 4-8 planidia had larvae in their bodieswhen dissected; ofthose
collecting vials. After approximately two weeks at 21°C the infected with 2-3 planidia, 4 out of 1 1 spiders had suppressed
eggs hatched into active planidia larvae. Approximately 100 the infection. Of the three emerging larvae, two were from
hatchlings ofthe wolfspiderPardosaprativaga(L. Koch 1870) singly infected and one from a doubly infected spider.
(Araneae: Lycosidae) were released into the vials and left Spider survival.- Survival of the spiderlings was higher in
there overnight to allow the planidia to infect the spiders. The the control group than among infected spiders (Wilcoxon test,
spiderlings were inspected for attached larvae under a bin- X22 = 23.0, P < 0.0001; Fig. 1) and directly related to the
ocular microscope and transferred to individual breeding vials number of infecting larvae (Acr = 0 vs. Acr = 1: y2 = 4.44,
with a plaster bottom. The number of attached larvae varied P - 0.0351; Acr = 1 vs. Acr = 2+: x2\ = 8.08, P =i0.0045).
between 0 and 8 per spider; thus the parasitoid load was not Among the singly infected spiders, survival was much better in
strictly controlled. We had 39 uninfected spiders, 37 infected those that suppressed the infection than in those that did not
wlairtvhae1.lTarhvea,sp7iwdiertlhin2,gs4wweirtehr3,ais1ewditihn 4t,he2vwiiatlhs a5t, a2n5d°C1 waintdh a8 (iynf2e1c=ted7.s7p,idPer=s i0n.0s0p5i4t)e,oafndlotwhesasmapmlee wsaizse t(ryue\ f=or4m.7u,ltPipl=y
16L:8D photoperiod, being fed fruit flies ad libitum and 0.0296). Survival ofsingly infected spiders that suppressed the
watered 2-3 times per week. The non-infected spiderlings infection was intermediate between control spiders and all
served as a control group. The fruit flies used were nutrient- singly infected spiders and did not differ from that ofcontrol
enriched by being raised in a mixed medium of Carolina spiders (x“i = 1.0, P = 0.31). Thus, survival seems to be
Biological Supply Drosophila medium (Formula 4-24) and determined by whether or not the larvae survive and grow in
crushed dog food. This mixed medium is nutritionally optimal the spider’s body. Therefore, overall survival ofsingly infected
for survival and growth of the spiders (Mayntz & Toft 2001). spiders was only slightly reduced because most spiders
The spiders were weighed weekly from the start of the suppressed the infection, whereas multiply infected spiders
experiment and up to 23 weeks. At this time some spiders had (with 2-8 larvae initially), of which fewer succeeded in
died; acrocerid larvae had emerged from others in order to suppressing the infection, had more strongly reduced survival
pupate; the remaining spiders were killed at this time. Spiders (Fig. 1). —
that died, as well as those that remained at the end, were Spider growth. Spiders infected with acrocerid larvae had
preserved in alcohol and subsequently dissected to check for higher growth rates than control spiders, and the more larvae
presence and number of acrocerid larvae in the abdomen. initially attached themselves, the higher the growth stimula-
Statistical analyses. JMP 8 was used for the statistical tion (repeated-measures ANOVA, Wilk’s 2 = 0.3, P < 0.0001;
analysis. We distinguished three groups of spiders according Fig. 2). Growth stimulation was significant in the spiders that
to the initial infection load (”# infecting larvae”): control (Act suppressed the infection, both as regards the singly (contrast:
= 0, non-infected), single infection (Acr = 1), and multiple F — 2.39, P = 0.0126) and the multiply infected (contrast: F
infection (Acr = 2+, 2-8 larvae). Spiders were also grouped 4.4!. p < 0.0001) spiders. There was no difference in growth
according to the success ofthe parasitoid, defined by the result stimulation, however, between the spiders that remained in-
of the dissections: no larva (L = 0) or one or more larvae fected and those that suppressed the infection (singly infected
present (L = 1+). Survivorship was analyzed by a Wilcoxon spiders: F = 1.59, P = 0.18); multiply infected spiders could
test. The growth curves were analyzed by repeated measures not be tested due to small sample sizes, but show a trend of
ANOVA of the weekly weight measurements. The time* reduced growth stimulation in spiders with a growing larva
parasite load (# initially infecting larvae or # surviving (Fig. 2). —
larvae) interaction terms were used to indicate significantly Spider development. For individuals that completed devel-
different growth patterns. Due to mortality, meaningful tests opment to the adult stage, total development time was
could only be made on data up to week 15. Developmental independent of the number of initially infecting acrocerid
parameters were analysed by ANOVA and /-tests; the data larvae (ANOVA, F2a9 = 0.38; P = 0.69) and of spider sex
were checked for homogeneity of variances (Levene’s test); (Welch /-test, / = 1.85; P — 0.0734). Spiders with a suc-
when this criterion could not be met, the Welch /-test was cessfully developing acrocerid larva took longer to reach
used. maturity than uninfected spiders (142.0 vs. 122.8 days; /-test,
RESULTS / = 4.82, P = 0.0019, n = 52). The first three instars were
rather short (1-2 weeks) in the normal pattern of juvenile
—
Parasitoid suppression. Only three fully developed acro- development unaffected by parasitoids, and were followed by
cerid larvae emerged from their spider hosts. However, several two very long (5-6 weeks) and an intermediate (ca. 3 weeks)
spiders that died during the experiment or were killed when instar. This pattern was repeated in singly infected spiders
the experiment ended turned out by dissection to contain one (most of which suppressed the parasitoid), but not in the
or two acrocerid larvae of intermediate or large (i.e., close to multiply infected spiders (Fig. 3). In the latter, the fourth
fully grown) size in their abdomens. These are here considered instar was as short as the three previous ones, and the fifth and
successful. However, many of the originally infected spiders sixth instars were prolonged. Spiders matured after 5-7
turned out to have no acrocerid larva inside, suggesting that juvenile instars; the maturation instar was independent of
TOFT ET AL.—LIFE HISTORY OF PARASITOID-INFECTED SPIDER 15
No. infecting/surviving
Weeks
—
Figure 1. Survivorshipcurvesforgroupsofspiderssubjectedtoinfectionbylarvaeoftheacrocerid flyAcroceraorbiculus: uninfected(control)
spiders(marked0); spidersoriginallyinfectedwith one(marked 1), splitinto thosethat suppressed the infection(1/0)and thosethatdid not(1/1);
spiders originally infected with 2-8 (marked 2+) larvae, split into those that suppressed the infection (2+/0) and those that did not (2+/1+).
Week
—
Figure 2. Growth curves forcontrol spiders and spiders infected with planidia larvae ofan acrocerid fly. Acr: number ofinitially infecting
larvae (0, 1 or >2); L: number ofgrowing larvae in the spiders’ body (0, 1 or >2). For clarity, error bars (SE) indicated only at weeks 0, 5, 10
and 15.
THE JOURNAL OF ARACHNOLOGY
16
60
1 2 3 4 5 6
Instar
Figure 3- Duration(mean+SE)ofjuvenileinstarsofthewolfspiderPardosaprativagadependingon initial infection load by planidialarvae
of an acrocerid lly. Acr = 0: uninfected control spiders; Acr = 1: spiders infected with I larva; Acr = 2+: spiders infected with 2-8 larvae.
Infection happened at the start ofinstar 1.
sex (~/J2 = 0.11, P = 0.95), number ofinitially infecting larvae howA. orbiculussynchronizesitslifecyclewith thatofitsspider
(X24 = 5.62, P = 0.23) and whether there was a surviving larva hosts.
(X22 = 0.96, P = 0.62). Multiple infections had higher costs for the spiders in terms
of reduced survival, but were much more successful from the
DISCUSSION
fly’s point of view than single infections. More larvae
Most spiders infected with a single parasitoid succeeded in developed from multiply than from singly infected spiders in
suppressingthe infection and in obtainingaclose to normal life spite ofmany fewer multiply infected spiders. Superparasitism
in terms ofsurvival and growth. It is unknown to what extent thus seems advantageous, probably because it may help to
the spiders’ high success of parasitoid suppression will apply break the host’s resistance, even in a situation where only one
also to natural conditions. Thead libitum feedingofthespiders parasitoid per host can complete development, as known also
with nutrient-enriched Hies may have strengthened the spiders’ from insects (Blumberg & Luck 1990; Khafagi & Hegazi
immune system (Slansky 1986) to more effectively combat the 2008). The adult fly is of approximately the same size as the
infection. Several environmental conditions in the laboratory adult spider, probably eliminating the possibility that more
also differed from those in the field, but it cannot be decided than one fly can successfully emerge. It may therefore be
whether these have benefitted the spider or the parasitoid. The hypothesized that the females’ habit of laying eggs in large
wolfspidermayalso bean inferiorhost forA. orbiculus. Species clumps on the vegetation is an adaptation that increases the
of Acrocera have a wide host range, including seven spider chance ofmultiple host infection. Multiple infections will lead
families (Schlinger 1987), but the relative suitability of these to strong competition between the larvae with only one of
potential hosts is unknown. The greatly increased mortality of them being successful. However, when they occur, multiple
successfully infected spiders compared with control spiders infections are likely to be by larvae from the same batch of
and unsuccessfully infected spiders may indicate a possibly low eggs; i.e., they will be kin (at least half-sibs), and even the
suitability of P. prativaga. The fact that we only allowed larvae that succumb will gain an inclusive fitness benefit. In
infection of spider hatchlings may also have been of impor- some cases two acrocerid larvae were found in a spider’s
tance, as the instar infected is known to affect the suitability of abdomen, but in no case were these of the full size ready for
insect hosts of parasitoids (Harvey 2000; Harvey et al. 2010; pupation, as was common with single larvae. It is therefore
Khafagi & Hegazi 2008). It prevented the acrocerid larva from doubtful ifenduring multiple infections ever lead to pupation.
immediate growth and development and enforced upon it a The benefit may be restricted to cases where competition
period ofdevelopmental arrest, waiting for the spider to grow between the larvae results in an early winner.
sufficiently for the parasitoid to complete its development. We Contrary to most solitary hymenopterous parasitoids of
have no information on the optimal host size for infection, or insects (Slansky 1986; Harvey et al. 1999), acrocerid infection
TOFT ET AL.—LIFE HISTORY OF PARASITOID-INFECTED SPIDER 17
ACKNOWLEDGMENTS
caused a stimulation of spider growth and thus an increased
rate of feeding. Multiply infected spiders also had enhanced ST was supported by grants from the Danish Research
growth compared with singly infected spiders; i.e., the Council (FNU) and the Carlsberg Foundation. We are
stimulation depended on the infection load. Since the body indebted to two reviewers for their comments.
mass measurements included the combined masses of the
spider and the parasitoid larva, the enhanced total growth LITERATURE CITED
might result from simply adding the growth of the larva to Blumberg, D. & R.F. Luck. 1990. Differences in the rates of
that ofthe spider. Enhanced growth ofinfected hosts may be superparasitism between two strains of Comperiella bifasciata (How-
interpreted as a result of parasitoid manipulation through ard) (Hymenoptera: Encyrtidae) parasitizing California red scale
which the parasitoid adjusts the resources available from the (Homoptera: Diaspididae): an adaptation to circumvent encapsula-
hHoasrtvefoyreotptali.mi(z2i0n1g0)itfsoouwnnd gtrhoatwtpha.raTshiutso,idHianrfveecytio(n20r0e0d)uacnedd Brotdieoun?r,AnJn.al&s oGf.theBoEinvtion.mol2o0g0i4c.alFuSoncciteitoynaolfAemceorliocgay8o3:f59i1m-m59a7t.ure
parasitoids. Annual Review ofEntomology 49:27M9.
the growth oflarge, but stimulated that ofsmall, host species. Eberhard, W.G. 2000. Spider manipulation by a wasp larva. Nature
Stimulation of growth in P. prativaga might thus be due to 406:255-256.
infection of the hatchling instar. However, the data do not Eberhard,W.G. 2010. Recoveryofspidersfrom theeffectsofparasitic
support this view: growth enhancement occurred as a result wasps: implications for fine-tuned mechanisms of manipulation.
of the original infection, independently of the fate of the Animal Behaviour 79:375-383.
parasitoid; it was seen also in the singly infected spiders that Godfray, H.C.J. 1994. Parasitoids: Behavioral and Evolutionary
suppressed the parasitoid, and parasitoid larvae successfully Ecology. Princeton University Press, Princeton, New Jersey.
developing in the spider’s body did not further increase the Harvey,J.A.2000. Dynamiceffectsofparasitismbyanendoparasitoid
spider’s growth rate. This indicates that growth enhancement wasp on thedevelopment oftwo host species: implications forhost
might be part of the process by which the spider fights Harqvueayli,tyJ.aA.n,dMp.aAra.siJteorviids,fiGt.neJs.sZ..EGcoollso,giNc.alJiEanntgom&olLo.gE.yM2.5:V2e6t7.-2197989..
the parasite. Our experimental spider, Pardosa prativaga,
Development of the parasitoid, Cotesia rubecula (Hymenoptera:
is known to be able to quickly catch up with a setback in Braconidae) in Pieris rapae and P. brassicae (Lepidoptera:
growth and development following periods of food stress Pieridae): evidence for host regulation. Journal of Insect Physiol-
(toxic or nutrient deficient food) by compensatory growth ogy 45:173-182.
(Jespersen & Toft 2003). We suggest that a similar process of Harvey, J.A., T. Sano & T. Tanaka. 2010. Differential host growth
enhanced resource acquisition and mobilization is induced in regulation bythesolitaryendoparasitoid, Meteoruspulchricornisin
response to parasitoid infection, perhaps induced already by two hosts of greatly differing mass. Journal of Insect Physiology
fthreompiWaneiedkia5la(rFviag..T2h),ewehnihcahncisedprgorboawbtlhy imsuocbhvieoaursliearlrtehaadny Har5v6e:y1,178J-.1A.1,83.D.J. Thompson & T.J. Heyes. 1996. Reciprocal
influences and costs ofparasitism on the development of Corcyra
tmhaeypenroitodsoofmiuntcehnsebegraowrtehduocftithoenpianrarseistoouirdcelasrvfaoer. Tithsuso,wnit gcieaphEaxlpoenriicmaeantnadliistseetndAopppalriacsaittaoi8d1:3V9e-n4t5u.ria canescens. Entomolo-
growth that stimulates the spider, but rather the increased Higgins, L.E. &M.A. Rankin. 2001. Mortalityriskofrapidgrowthin
nutritional demands of the spider’s immune system during the spider Nephila clavipes. Functional Ecology 15:24-28.
encapsulation (i.e., resisting a parasitoid infection) of the Jespersen, L.B. & S. Toft. 2003. Compensatory growth following
parasitoid (Slansky 1986; Strand & Pech 1995). Since there early nutritional stress in the wolf spider Pardosa prativaga.
may be a cost of increased growth per se (Higgins & Rankin Functional Ecology 17:737-746.
2001), the reduced survival of infected spiders may be Khafagi,W.E.&E.M. Hegazi.2008.Doessuperparasitismimprovehost
explained not only by a negative “health effect” due to the suitabilityforparasitoiddevelopment?AcasestudyintheMicroplitis
infection but also by this growth cost, or both. This line of Mayrnuftizv,enDtr.is&-SSp.odToopftte.ra20l0it1t.oraNluitsrsiyesnttemc.omBpioosCionttiroon!o5f3:th4e27p-r4e3y8’.s diet
argument interprets the enhanced growth rate of the spiders affectsgrowth andsurvivorshipofageneralist predator. Oecologia
as a costly response likely to be paid later in terms ofreduced 127:207-213.
reproductive success (Metcalfe & Monaghan 2001). Encap- Metcalfe, N.B. & P. Monaghan. 2001. Compensation fora bad start:
sulation was also found to have costs in a pyralid moth grownow, paylater?Trendsin Ecologyand Evolution 16:254-260.
(Harvey et al. 1996). Nartshuk, E.P. 2010. Fauna Europaea: Acroceridae. In Fauna
The study has revealed an impressive ability of the wolf Europaea: Acroceridae. Fauna Europaea version 2.3. (T. Pape &
spiderto suppress an infection by theacrocerid fly. This comes P. Beuk, eds.). Online at http://www.faunaeur.org
on top of previously reported pre-infection attacks on the Nielsen, B.O., P. Funch & S. Toft. 1999. Self-injection ofa dipteran
larvae as prey, and early post-infection ability to mechanically parasitoid into a spider. Naturwissenschaften 86:530-532.
Schlinger, E.I. 1987. The biology of Acroceridae (Diptera): true
free itself from attached larvae (Nielsen et ai. 1999). endoparasitoids of spiders. Pp. 319-327. In Ecophysiology of
Physiological and life history costs to spiders that suppressed spiders. (W. Nentwig, ed.). Springer, Berlin.
theirinfection seemed minimal. Unfortunately, wedid notalso Slansky, Jr., F. 1986. Nutritional ecology ofendoparasitic insectsand
have the opportunity to test whether parasitoid suppression their hosts: an overview. Journal ofInsect Physiology 32:255-261.
modified the spider’ssubsequent reproductive life. Atthesame Strand, M.R. & L.L. Pech. 1995. Immunological basis for compat-
time the parasitoid had a heavy impact on the spiders in terms ibility in parasitoid-host relationships. Annual Review of Ento-
of higher mortality. It remains to be established whether A. mology 40:31-56.
orbiculus hasmore suitable hoststhan P. prativaga available in Thomas, F.,S.Adamo&J. Moore.2005. Parasiticmanipulation:where
the northern European part ofits distribution. Ifnot, this may are we and where should we go? Behavioural Processes 68:185-199.
explain its rarity here. Manuscript received24 March 2011. revised28 November 2011.