FunctionalEcology200519, 625–631© 2

Functional
Ecology
2005
19
, 625–631
© 2005 British
Ecological Society
625
Blackwell Publishing, Ltd.
Is a starving host tastier? Reproduction in fleas
parasitizing food-limited rodents
B. R. KRASNOV,*† I. S. KHOKHLOVA,‡ M. S. ARAKELYAN*§ and
A. A. DEGEN‡
*
Ramon Science Center and Mitrani Department of Desert Ecology, Jacob Blaustein Institutes for Desert Research,
Ben-Gurion University of the Negev, Mizpe Ramon, Israel,
‡
The Wyler Department of Dryland Agriculture, Jacob
Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, Israel, and
§
Department of Zoology, Yerevan State University, Yerevan, Armenia
Summary
1.
We hypothesized that food availability, and therefore body condition, of a rodent
host,
Meriones crassus
, affects egg production and survival, and development time of
preimago and adults of the first generation of the flea
Xenopsylla ramesis
.
2.
Egg production was significantly higher in fleas parasitizing underfed than control
animals.
3.
Food availability for hosts affected survival of eggs and larvae produced by fleas on
these rodents, but did not affect survival of pupae. More than twice the number of eggs
from fleas on food-limited hosts survived than those from fleas on control rodents.
Highest larval survival was recorded in fleas on rodents with 30% of maintenance
energy intake.
4.
Survival of new generation imagos was lowest in fleas from parents on hosts with
the highest food limitation. By contrast, survival of parent fleas was highest on hosts
offered 30% of maintenance energy intake.
5.
Time of egg and larval development was longest on hosts consuming 30% of energy
requirements for maintenance. By contrast, there was no difference in time to emergence
in pupae from flea females on rodents from different treatments. Survival time
under starvation of imago of the first generation was shortest in offspring of fleas that
parasitized
M. crassus
offered the minimal amount of food. The opposite was true for
parent fleas.
6.
The results suggest nutritional and/or energetic cost of host resistance, measured
as host-mediated parasite fitness loss, as well as possible adaptive stress-induced
immunosuppression.
Key-words
: Egg production, flea, food limitation, rodent host, survival
Functional Ecology
(2005)
19
, 625–631
doi: 10.1111/j.1365-2435.2005.01015.x
Introduction
Distribution of parasite individuals across a host
population is characterized by their aggregation. Most
individuals of the parasite occur on a few host individuals,
while most host individuals have only a few, if any,
parasites (Anderson & May 1978). The general cause
of the heterogeneous distribution of parasites is due to
host heterogeneity of the rate of gain or loss of parasites
(Poulin 1998). In other words, if some hosts have more
parasites than would be predicted by chance, it suggests
that they offer the parasites a better quality habitat. In
particular, intraspecific host variation in suitability for
a parasite can be caused by the variation in the pattern
of resource acquisition by a parasite such as intrahost
variation in defences.
Hosts defend themselves against parasites using
specific behavioural, physiological and/or immunological
mechanisms that can result in loss of fitness in the
parasites. This host-mediated loss of fitness in a parasite
is considered to be host resistance (Poulin 1998). The
defence against parasites can be costly for a host. For
example, activation of an immune response and even
maintenance of a competent immune system is an energetically
demanding process that requires trade-off decisions
among competing energy demands for growth,
†Author to whom correspondence should be addressed.
E-mail: [e-mail]
626
B. R. Krasnov
et al.
© 2005 British
Ecological Society,
Functional Ecology
,
19
, 625–631
reproduction, thermoregulation, work and immunity
(Sheldon & Verhulst 1996). Empirical evidence suggests
that such costs can be high (e.g. Moret & Schmid-
Hempel 2000).
The trade-off between the advantage of resistance
and its cost should be most critical for hosts that face
energy limitations. Therefore, energy-deprived hosts
might be less resistant and, thus, represent better patches
for parasites. Intraspecific host variation in energy reserves
can arise due to a variety of reasons (Kam & Degen 1993;
Cumming & Bernard 1997). As a result, in many nontropical
vertebrate animals, disease prevalence is increased
during periods of food shortage compared with periods
when food is readily available (Lochmiller, Vestey &
McMurray 1994). However, resources that parasites
extract from their hosts (e.g. blood) can be of a lower
quality in energy-deprived hosts (De Pedro
et al
. 2003).
Thus, hosts in good condition can be a better food source
than hosts in poor condition (Dawson & Bortolotti 1997).
Consequently, parasites face a trade-off between the
choice to attack less defended but lower-quality,
vs
more
defended but higher-quality, hosts. From the evolutionary
perspective, the strategy of parasitizing either
energy deprived or energy-rich hosts would depend on
relative fitness rewards from exploiting these hosts.
This reasoning leads to two alternative predictions
regarding the effect of host energy deprivation on
reproductive patterns of a parasite. Reproductive output
of a parasite will be higher when exploiting energy
deprived hosts if the fitness increment due to reduced
host defences is higher than the fitness decline due to
lower quality of resources extracted from a host.
Alternatively, this reproductive output will be lower
when exploiting energy deprived hosts if the fitness
increment due to reduced host defences is lower than
the fitness decline due to lower quality of extracted
resources.
The effect of body condition of hosts on their
resistance to parasites has been studied in a variety of
animals by manipulating the body condition of hosts
(Oppliger, Christe & Richner 1996; Brown, Loosli &
Schmid-Hempel 2000; Jokela
et al
. 2005). However, all
of these studies have investigated this phenomenon from
the host perspective, whereas the effect of host condition
on parasite parameters (apart for abundance) has largely
been neglected.
Here, we studied the effect of host energy intake on
parasite fitness. We hypothesized that food limitation
of a rodent host,
Meriones crassus
, affects reproductive
potential (in terms of egg production) and quality of
offspring (in terms of survival and development time
of preimaginal stages and adults of the first generation)
of its characteristic flea parasite
Xenopsylla ramesis
.
Fleas (Siphonaptera) are parasites of higher vertebrates,
being most abundant and diverse on small mammals.
In most cases, preimaginal development is entirely
off-host. Larvae of almost all flea species are not
parasitic and feed on organic debris in the burrow and/
or nest of the host.
Materials and methods
    
Fleas and rodents (27 adult males) were obtained from
our laboratory colonies started from field-collected
individuals. Details on the flea and rodent-rearing
procedures can be found elsewhere (Krasnov
et al
.
2001a,b, 2004). In this study we used only newly emerged
fleas, 2 days of age, which did not feed from emergence
until experimental treatments. The study was conducted
under permits from the Israel Nature and National
Parks Protection Authority and Ben-Gurion University
Committee for the Ethical Care and Use of Animals in
Experiments.
  
Prior to experimental trials with fleas, the rodents were
placed in individual plastic cages (20
×
40
×
15 cm
3
with
3 mm clean sand as a substrate) and divided randomly
into three groups (nine individuals per group). One group
(control group, C) was offered millet seeds equivalent to
approximately 100% of maintenance energy requirements,
whereas the two other groups were offered
60% (group T1) and 30% (group T2), respectively, of
maintenance requirements. In addition, rodents were
offered 3 g fresh alfalfa leaves, which provided their water
needs but with minimal additional energy (I. Khokhlova,
unpublished data). Energy requirements for maintenance
of
M. crassus
were taken as 8·3 kJ day
−
1
(Khokhlova, Degen & Kam 1995) and calculated for
each animal based on its body mass (
m
b
). Animals were
weighed daily at 09.00 h to 0·01 g (Ohaus CT200-S
electronic balance, Ohaus Corporation, Pine Brook, NJ,
USA). Food offered was weighed to 0·0001 g (Mettler-
Toledo AB, Dietikon, Switzerland). Animals from the
control group and animals from the T1 treatment
maintained their body mass throughout the experimental
period; their body mass after the first week was
98·9
±
2·1% and 97·3
±
1·7%, respectively, of initial
body mass. However, body mass of rodents from the
T2 group decreased after the first week to 81·1
±
1·8%
of initial body mass.
  
We selected 945 newly emerged female and 270 newly
emerged male
X. ramesis
and assigned them randomly
to the three experimental treatments that differed in
food availability to a rodent host. We placed 35 female and
10 male fleas on each rodent a week after the beginning
of experiments. Four days after the fleas were in the
rodent cage, we collected the fleas by brushing the hair
of the rodent with a toothbrush and sieving the cage
substrate until no fleas could be recovered during
10 min of either brushing or sieving. No behavioural
difference during manipulations was recorded among
rodents of the three experimental groups. In total, we
mb
−0⋅54
627
Reproduction in
fleas exploiting a
food-limited host
© 2005 British
Ecological Society,
Functional Ecology
,
19
, 625–631
recovered 804 female and 156 male fleas. Each female
flea was examined under a light microscope, and the
degree of egg development (early, middle or late stage)
was determined visually (see Krasnov
et al
. 2002 for
details). Fleas with egg

Results (Indonesian) 1:

FunctionalEcology200519, 625–631© 2005 BritishEcological Society625Blackwell Publishing, Ltd.Is a starving host tastier? Reproduction in fleasparasitizing food-limited rodentsB. R. KRASNOV,*† I. S. KHOKHLOVA,‡ M. S. ARAKELYAN*§ andA. A. DEGEN‡*Ramon Science Center and Mitrani Department of Desert Ecology, Jacob Blaustein Institutes for Desert Research,Ben-Gurion University of the Negev, Mizpe Ramon, Israel,‡The Wyler Department of Dryland Agriculture, JacobBlaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, Israel, and§Department of Zoology, Yerevan State University, Yerevan, ArmeniaSummary1.We hypothesized that food availability, and therefore body condition, of a rodenthost,Meriones crassus, affects egg production and survival, and development time ofpreimago and adults of the first generation of the fleaXenopsylla ramesis.2.Egg production was significantly higher in fleas parasitizing underfed than controlanimals.3.Food availability for hosts affected survival of eggs and larvae produced by fleas onthese rodents, but did not affect survival of pupae. More than twice the number of eggsfrom fleas on food-limited hosts survived than those from fleas on control rodents.Highest larval survival was recorded in fleas on rodents with 30% of maintenanceenergy intake.4.Survival of new generation imagos was lowest in fleas from parents on hosts withbatasan makanan yang tertinggi. Sebaliknya, kelangsungan hidup kutu orangtua adalah tertinggi pada hostditawarkan 30% dari asupan energi pemeliharaan.5.Waktu telur dan perkembangan larva adalah terpanjang di host mengkonsumsi 30% dari energipersyaratan untuk pemeliharaan. Sebaliknya, ada tidak ada perbedaan dalam waktu untuk timbulnyadalam kepompong dari loak betina pada tikus dari perawatan yang berbeda. Waktu kelangsungan hidupdi bawah kelaparan imago generasi pertama adalah terpendek dalam keturunan kutu yangparasitizedM. crassusminimal jumlah makanan yang ditawarkan. Sebaliknya adalah benar untukorangtua kutu.6.Hasilnya menunjukkan biaya gizi dan/atau energik resistensi host, diukursebagai tuan rumah-dimediasi parasit kebugaran kerugian, serta mungkin adaptif diinduksi stresimunosupresi.Kata kunci: Telur produksi, kutu, batasan makanan, hewan pengerat host, kelangsungan hidupEkologi fungsional(2005)19625-631Doi: 10.1111/j.1365-2435.2005.01015.xPendahuluanDistribusi parasit individu di seluruh hostpopulasi dicirikan oleh agregasi mereka. Sebagianindividu parasit terjadi pada beberapa host individuSementara kebanyakan host individu memiliki hanya sedikit, jika ada,parasit (Anderson & Mei 1978). Penyebab umumdistribusi heterogen parasit adalah karenahost heterogenitas tingkat keuntungan atau kerugian parasit(Poulin 1998). Dengan kata lain, jika beberapa host memiliki lebihparasit daripada akan dapat diprediksi secara kebetulan, itu menyarankanbahwa mereka menawarkan parasit habitat kualitas lebih baik. Dalamparticular, intraspecific host variation in suitability fora parasite can be caused by the variation in the patternof resource acquisition by a parasite such as intrahostvariation in defences.Hosts defend themselves against parasites usingspecific behavioural, physiological and/or immunologicalmechanisms that can result in loss of fitness in theparasites. This host-mediated loss of fitness in a parasiteis considered to be host resistance (Poulin 1998). Thedefence against parasites can be costly for a host. Forexample, activation of an immune response and evenmaintenance of a competent immune system is an energeticallydemanding process that requires trade-off decisionsamong competing energy demands for growth,†Author to whom correspondence should be addressed.E-mail: [e-mail]626B. R. Krasnovet al.© 2005 BritishEcological Society,Functional Ecology,19, 625–631reproduction, thermoregulation, work and immunity(Sheldon & Verhulst 1996). Empirical evidence suggeststhat such costs can be high (e.g. Moret & Schmid-Hempel 2000).The trade-off between the advantage of resistanceand its cost should be most critical for hosts that faceenergy limitations. Therefore, energy-deprived hostsmight be less resistant and, thus, represent better patchesfor parasites. Intraspecific host variation in energy reservescan arise due to a variety of reasons (Kam & Degen 1993;Cumming & Bernard 1997). As a result, in many nontropicalvertebrate animals, disease prevalence is increasedduring periods of food shortage compared with periodswhen food is readily available (Lochmiller, Vestey &McMurray 1994). However, resources that parasitesextract from their hosts (e.g. blood) can be of a lowerquality in energy-deprived hosts (De Pedroet al. 2003).Thus, hosts in good condition can be a better food sourcethan hosts in poor condition (Dawson & Bortolotti 1997).Consequently, parasites face a trade-off between thechoice to attack less defended but lower-quality,vsmoredefended but higher-quality, hosts. From the evolutionaryperspective, the strategy of parasitizing eitherenergy deprived or energy-rich hosts would depend onrelative fitness rewards from exploiting these hosts.This reasoning leads to two alternative predictionsregarding the effect of host energy deprivation onreproductive patterns of a parasite. Reproductive outputof a parasite will be higher when exploiting energydeprived hosts if the fitness increment due to reducedhost defences is higher than the fitness decline due tolower quality of resources extracted from a host.Alternatively, this reproductive output will be lowerwhen exploiting energy deprived hosts if the fitnessincrement due to reduced host defences is lower thanthe fitness decline due to lower quality of extractedresources.The effect of body condition of hosts on theirresistance to parasites has been studied in a variety ofanimals by manipulating the body condition of hosts(Oppliger, Christe & Richner 1996; Brown, Loosli &Schmid-Hempel 2000; Jokelaet al. 2005). However, allof these studies have investigated this phenomenon fromthe host perspective, whereas the effect of host conditionon parasite parameters (apart for abundance) has largelybeen neglected.Here, we studied the effect of host energy intake onparasite fitness. We hypothesized that food limitationof a rodent host,Meriones crassus, affects reproductivepotential (in terms of egg production) and quality ofoffspring (in terms of survival and development timeof preimaginal stages and adults of the first generation)of its characteristic flea parasiteXenopsylla ramesis.Fleas (Siphonaptera) are parasites of higher vertebrates,being most abundant and diverse on small mammals.In most cases, preimaginal development is entirelyoff-host. Larvae of almost all flea species are notparasitic and feed on organic debris in the burrow and/or nest of the host.Materials and methods    Fleas and rodents (27 adult males) were obtained fromour laboratory colonies started from field-collectedindividuals. Details on the flea and rodent-rearingprocedures can be found elsewhere (Krasnovet al.2001a,b, 2004). In this study we used only newly emergedfleas, 2 days of age, which did not feed from emergenceuntil experimental treatments. The study was conductedunder permits from the Israel Nature and NationalParks Protection Authority and Ben-Gurion UniversityCommittee for the Ethical Care and Use of Animals inExperiments.  Prior to experimental trials with fleas, the rodents wereplaced in individual plastic cages (20×40×15 cm3with3 mm clean sand as a substrate) and divided randomlyinto three groups (nine individuals per group). One group(control group, C) was offered millet seeds equivalent toapproximately 100% of maintenance energy requirements,whereas the two other groups were offered60% (group T1) and 30% (group T2), respectively, ofmaintenance requirements. In addition, rodents wereoffered 3 g fresh alfalfa leaves, which provided their waterneeds but with minimal additional energy (I. Khokhlova,unpublished data). Energy requirements for maintenanceofM. crassuswere taken as 8·3 kJ day−1(Khokhlova, Degen & Kam 1995) and calculated foreach animal based on its body mass (mb). Animals wereweighed daily at 09.00 h to 0·01 g (Ohaus CT200-Selectronic balance, Ohaus Corporation, Pine Brook, NJ,USA). Food offered was weighed to 0·0001 g (Mettler-Toledo AB, Dietikon, Switzerland). Animals from thecontrol group and animals from the T1 treatmentmaintained their body mass throughout the experimentalperiod; their body mass after the first week was98·9±2·1% and 97·3±1·7%, respectively, of initialbody mass. However, body mass of rodents from theT2 group decreased after the first week to 81·1±1·8%of initial body mass.  We selected 945 newly emerged female and 270 newlyemerged maleX. ramesisand assigned them randomlyto the three experimental treatments that differed infood availability to a rodent host. We placed 35 female and10 male fleas on each rodent a week after the beginningof experiments. Four days after the fleas were in therodent cage, we collected the fleas by brushing the hairof the rodent with a toothbrush and sieving the cagesubstrate until no fleas could be recovered during10 min of either brushing or sieving. No behaviouraldifference during manipulations was recorded amongrodents of the three experimental groups. In total, wemb−0⋅54627Reproduction infleas exploiting afood-limited host© 2005 BritishEcological Society,Functional Ecology,19, 625–631recovered 804 female and 156 male fleas. Each femaleflea was examined under a light microscope, and thedegree of egg development (early, middle or late stage)was determined visually (see Krasnovet al. 2002 fordetails). Fleas with egg