Common facultative endosymbionts do not influence sensitivity of cereal aphids to pyrethroids

1. Cereal aphids, including the bird cherry-oat aphid, Rhopalosiphum padi, and the grain aphid, Sitobion avenae, can transmit viruses that significantly reduce crop yields. To mitigate against yield losses, insecticides are routinely used to manage aphid populations.
2. Aphids can form relationships with endosymbionts that confer fitness benefits or consequences to the aphid. Recent artificial inoculation experiments indicate that endosymbionts could increase aphid susceptibility to insecticides, but this has not been explored using aphid populations naturally infected with endosymbionts.
3. Here, we sampled aphids from an important cereal production region in Lower Saxony, Germany. We characterized the endosymbiont profile of these aphid populations and conducted pyrethroid dose–response assays to test the hypothesis that facultative endosymbionts increase aphid susceptibility to insecticides.
4. We find that the level of insecticide susceptibility is highly variable in S. avenae and we identify populations that are sensitive and tolerant to pyrethroids, including populations collected from the same field. For R. padi, we find evidence for decreased sensitivity to pyrethroids, representing the first report of reduced sensitivity to pyrethroids in R. padi sampled from Central Europe.
5. We detected high endosymbiont infection frequencies in the aphid populations. 84% of aphids carry one facultative endosymbiont and 9% of aphids carry two facultative endosymbionts. We detected associations with Regiella insecticola, Fukatsia symbiotica, and Hamiltonella defensa. However, we do not identify a link between endosymbiont infection and insecticide susceptibility, indicating that other factors may govern the development of insecticide resistance and the need for alternative management strategies.

     

Defence variation within a guild of aphid-tending ants explains aphid population growth

• Mutualism studies often focus on the service provided by single species, while variation in benefits provided by multiple partners is less understood. Ant-aphid food-for-protection mutualisms are suitable to study the implications of intra-guild service variation because they often involve several ant species with varying levels of aggressiveness.
• We studied an aphid species and its associated ant guild to address whether intra-guild defence variation against aphid natural enemies explains aphid performance on plants (thistles). We surveyed plants with natural abundances of aphids associated with different ant species and estimated aphid population growth. We conducted confrontation experiments between ant species and aphid natural enemies (ladybugs and hoverfly larvae). In plants patrolled by the most aggressive ant species, we determined the ant's probability of expelling aphid natural enemies and tested whether ant exclusion affects the abundance of aphids and their natural enemies.
• The ant Dorymyrmex tener was the most abundant and frequent species on plants and the most aggressive against aphid natural enemies. Aphid populations grew faster on plants patrolled by D. tener compared to plants patrolled by Camponotus distinguendus or D. richteri. Field experiments confirmed that D. tener effectively expels aphid natural enemies from plants. When D. tener was excluded, the density of aphids decreased, while the abundance of aphid natural enemies increased.
• The disruption of aphid predation by aggressive and numerically dominant ant species is a determinant of aphid population dynamics. This study illustrates the importance of considering intra-guild service variation to better understand multi-partner mutualisms.

     

Overdispersion of Allothrombium pulvinum larvae (Acari: Trombidiidae) parasitic on Aphis gossypii (Homoptera: Aphididae) in cotton fields

Abstract.

• 1 Dispersion patterns of the protelean parasite, Allothrombium pulvinum Ewing, among individuals of an aphid host, Aphis gossypii Glover, were examined during spring 1991 in several cotton fields in Jiangsu Province, China.
• 2 The variance-to-mean ratios (i.e. dispersion index) of larval mites per host were greater than 1, indicating that the mite parasites were overdispersed among aphid hosts. The variance increased with the mean according to the power law, variance = 1.51 mean¹⁰⁶, which explained 99.7% of the variation in the data.
• 3 The negative binomial distribution adequately describes the patterns of larval mite dispersion among aphid hosts in eight out of ten populations. The degree of clumping (1/k) decreased curvilinearly with parasite density (mites per host).
• 4 Mites were more clumped among adult aphids than among immature ones.
• 5 Ecological and evolutionary consequences of mite overdispersion within host populations are discussed. The role of Allothrombium in pest control is also discussed.

     

Two‐year oscillation cycle in abundance of soybean aphid in Indiana

• 1 The present study evaluated the population dynamics of the heteroecious soybean aphid Aphis glycines Matsumura (Hemiptera: Aphididae) during an 8‐year period in Indiana, shortly after its detection in North America. Sampling conducted at multiple locations revealed that A. glycines exhibited a 2‐year oscillation cycle that repeated itself four times between 2001 and 2008: years of low aphid abundance were consistently followed by years of high aphid abundance.
• 2 Similar patterns of abundance of A. glycines and coccinellids (Coleoptera: Coccinellidae) in soybean fields, both within and between‐years, suggest that late season predation by coccinellids plays a role in the oscillatory cycle of aphids. Insidious flower bugs Orius insidiosus (Say) (Hemiptera: Anthocoridae) were numerically more abundant than coccinellids, although the lack of synchrony between aphids and predatory bugs suggests that O. insidiosus has a limited influence on between‐year variations in aphid density.
• 3 The inverse relationship between aphid densities before and after the start of the autumn migratory period changes direction in alternate years. High aphid density on soybean in the summer is associated with a reduced number of alate migrants produced in the autumn. Conversely, years with low density aphids on soybean in the summer are characterized by high numbers of alates that migrate to the primary host in the autumn.
• 4 From a pest management perspective, the 2‐year oscillation cycle of A. glycines is a desirable attribute with respect to population dynamics because it implies that aphids cause significant economic damage only in alternate years (as opposed to every year). Cultural practices enhancing the conservation biological control of Coccinellidae may help to preserve the periodicity of aphid infestation and restrict the pest status of A. glycines.

     

Temporal and spatial variability of rosy apple aphid Dysaphis plantaginea populations: is there a role of the alternate host plant Plantago major?

• 1 The rosy apple aphid Dysaphis plantaginea (Passerini) (Homoptera: Aphididae) is a pest of economic importance to the apple industry worldwide, particularly in organic apple orchards where no acceptable controls are available. In the Similkameen Valley of British Columbia, Canada, the rosy apple aphid population size varies widely between orchards and between years. To explain this variation, potential environmental correlates of aphid density were evaluated. The architecture of the alternate host was also evaluated for its effect on rosy apple aphid summer survival and reproduction.
• 2 The percentage of trees infested by rosy apple aphids among orchards was in the range 8–94% for trees having at least one cluster with more than ten aphids in 2007 and in the range 0–39% in 2008.
• 3 A general linear model correlating aphid densities to the environmental variables of abundance of the alternate host (plantain Plantago spp.), foliar nitrogen, tree age and planting density, and reduced by stepwise regression, indicated that foliar nitrogen and tree age explained 33% of the variation. Abundance of the summer, alternate food plant, plantain, was not related to later aphid densities on apple trees.
• 4 Plantain architecture, however, influenced aphid numbers and 25‐fold more aphids were found on low‐lying plantain leaves than on more upright leaves. Experimental manipulation of leaf angle and leaf size showed that significantly more aphids occurred on low angle, large leaves. Finally, mowing that encouraged low lying plants prior to spring aphid migration was associated with a four‐fold greater number of both winged and wingless aphids on the plantain.

     

Biology and trophic interactions of lucerne aphids

相似文献   

Nematode–citrus plant interactions: host preference,damage rate and molecular characterization of Citrus root nematode Tylenchulus semipenetrans

• Citrus plants are host to several plant parasitic nematodes (PPNs), which are microscopic organisms. Among PPNs, the citrus root nematode, T. semipenetrans (Cobb 1913) (Tylenchida: Tylenchulidae), causes significant damage to citrus plantations worldwide. Understanding citrus nematode populations, precise identification, host preference among citrus species, and damage threshold are crucial to control T. semipenetrans. The minutiae of citrus plant–nematode interactions, nematode density and molecular nematode identification are not well understood. In this study, nematode species and density in citrus orchards, host specialization, molecular and morphological characteristics of nematodes were assessed.
• Molecular and morphological methods, host–nematode interactions, host (citrus species) preference, damage economic threshold (ET), and economic injury level (EIL) were determined using laboratory methods and field sampling. Citrus plantations in different provinces in the Mediterranean region of Turkey were investigated.
• Nematode species were identified molecularly and morphologically. ITS sequences revealed that samples were infected by citrus root nematode T. semipenetrans. The lowest nematode density was in C. reticulata in Mersin (53 2nd stage juveniles (J2s) 100 g⁻¹ soil), while the highest density was from Hatay in C. sinensis (12173 J2s 100 g⁻¹ soil). Highest citrus nematode population density was on roots of C. reticulata, followed by C. sinensis, C. limon, and C. paradisi.
• The citrus nematode is more common than was thought and population fluctuations change according to specific citrus species. Environmental conditions, host and ecological factors, such as temperature, soil pH, and soil nutrients, might influence nematode populations in citrus orchards. Investigating nematode density in diverse soil ecologies and the responses of different resistant/tolerant citrus species and cultivars to nematode populations is essential in future studies.

     

The population dynamics of the stinging nettle aphid, Microlophium carnosum (Bukt.)

• 1 Populations of Microlophium carnosum on patches of perennial stinging nettle, Urtica dioica increased rapidly during April and May to reach their peaks in June. The decline in numbers was equally rapid and very small populations persisted into autumn; parthenogenetic over-wintering was recorded.
• 2 Suboptimal ambient temperature and mortality due to natural enemies contributed mainly to the post-peak development of aphid populations.
• 3 The combined effects of intraspecific competition and a deterioration in the food quality of the host plant appeared to be the major factors determining the temporal pattern of aphid abundance.
• 4 Variation in the size of populations between the three sampled sites was correlated with differences in the food quality of nettles as indicated by aphid mean relative growth rate.
• 5 Each nettle patch has a particular ‘carrying capacity’ for aphids, within which a biennial fluctuation between relatively large and small aphid populations appeared to be emerging at most of the sites investigated. A natural or aphid-induced cycle in host plant quality, or alternatively, the persistent effects of intraspecific competition over several generations, may explain this fluctuation.

     

Aphid population dynamics: does resistance to parasitism influence population size?

1. Insect population size is regulated by both intrinsic traits of organisms and extrinsic factors. The impacts of natural enemies are typically considered to be extrinsic factors, however insects have traits that affect their vulnerability to attack by natural enemies, and thus intrinsic and extrinsic factors can interact in their effects on population size. 2. Pea aphids Acyrthosiphon pisum Harris (Hemiptera: Aphididae) in New York and Maryland that are specialised on alfalfa are approximately two times more physiologically resistant to parasitism by Aphidius ervi Haliday (Hymenoptera: Braconidae) than pea aphids specialised on clover. To assess the potential influence of this genetically based difference in resistance to parasitism on pea aphid population dynamics, pea aphids, A. ervi, and other natural enemies of aphids in clover and alfalfa fields were sampled. 3. Rates of successful parasitism by A. ervi were higher and pea aphid population sizes were lower in clover, where the aphids are less resistant to parasitism. In contrast, mortality due to a fungal pathogen of pea aphids was higher in alfalfa. Generalist aphid predators did not differ significantly in density between the crops. 4. To explore whether intrinsic resistance to parasitism influences field dynamics, the relationship between resistance and successful field parasitism in 12 populations was analysed. The average level of resistance of a population strongly predicts rates of successful parasitism in the field. The ability of the parasitoid to regulate the aphid may vary among pea aphid populations of different levels of resistance.     

The influence of nitrogen fertiliser on the population development of the cereal aphids Sitobion avenae (F.) and Metopolophium dirhodum (Wlk.) on field grown winter wheat

The effect of nitrogen fertiliser on populations of the cereal aphids Sitobion avenae and Metopolophium dirhodum on winter wheat was investigated in a three year field experiment. Naturally occurring aphid populations were monitored on three nitrogen treatments; none, nitrogen application using Canopy Management guidelines (130–210 kg ha^-1) and conventional practice (190 kg ha^-1). Inoculations of laboratory reared S. avenae were used to enhance field populations on half the plots. Natural populations of M. dirhodum remained below the current UK spray threshold level of two-thirds of shoots infested at the start of flowering, or five aphids per shoot in all years, whilst populations of S. avenae exceeded the threshold in all years. The response of the two species to nitrogen differed. Significantly higher populations of M. dirhodum were recorded in both treatments which received nitrogen in all years, whilst the response of S. avenae varied between years. In 1994 and 1995 when environmental conditions favoured aphid development, higher populations were recorded in the two treatments which received nitrogen. In 1993 when high rainfall created unfavourable conditions, higher populations were recorded in the plots receiving no nitrogen. Differences in peak density and cumulative aphid index of S. avenae resulted from differences in the rate of population increase between ear emergence and peak density on the different treatments. Populations prior to ear emergence were higher in the plots which received nitrogen but the differences were not statistically significant. There was no evidence of a difference in the timing of population decline in the different treatments. In 1993 higher levels of infection by entomopathogenic fungi were observed in all treatments. Significantly higher levels of infection were recorded in the treatments receiving nitrogen, which may have accounted for the lower S. avenae populations recorded. It is possible that the larger canopies recorded in these treatments produced conditions which favoured infection by fungi, thereby limiting aphid population growth. The results indicate that application of nitrogen increases natural populations of M. dirhodum, and under favourable conditions, populations of S. avenae. However, in suboptimal climatic conditions, the application of nitrogen fertiliser can lead to lower populations of 5. avenae. The data also suggest that there is no consistent difference between a conventional and Canopy Managed approach to nitrogen fertiliser use in terms of the risk of infestation by cereal aphids.     

Preference,performance, and impact of the water-lily aphid on multiple species of duckweed

1. The role of herbivores in driving the structure of freshwater macrophyte communities remains poorly understood in comparison with terrestrial ecosystems. For instance, although duckweed (subfamily Lemnoideae) are globally distributed, can be locally highly abundant and ecologically dominant, and are of growing economic importance, their interactions with herbivores remain understudied.
2. To address how herbivores may impact duckweed species composition, we here experimentally quantify the preference and performance of a common duckweed herbivore, the water-lily aphid (Rhopalosiphum nymphaeae) on four widespread duckweed species.
3. Our two-way choice experiments reveal that aphids display a preference for Spirodela polyrhiza > Landoltia punctata = Lemna minor > > Wolffia brasiliensis. These results are rarely influenced by natal host species.
4. By evaluating the growth of aphid populations on each duckweed species, we find that preference may be adaptive in certain ecological conditions.
5. Quantifying the population growth rate of duckweed growing in the presence and absence of aphids revealed differential tolerance of herbivory across duckweed species.
6. This study shows that aphids, through preferential feeding and significant differential effects on duckweed growth, can have a significant impact on duckweed population dynamics and potentially community composition.

     

Predation and avoidance behaviour in aphid‐ladybird interactions of native and invasive ladybirds in Europe

1. While detrimental effects of invasive predators on native species are well documented, we often lack a mechanistic understanding of the invasion success. Lack of prey avoidance behaviour can lead to higher consumption rates by invasive predators compared to native predators. This competitive advantage is expected to contribute to the invasion success of non‐native predators.
2. We compared aphid consumption and cue avoidance behaviour of aphids between four native ladybird species (Coccinella septempunctata, Adalia bipunctata, Propylea quatuordecimpunctata, and Hippodamia variegata) and the invasive Asian ladybird Harmonia axyridis.
3. The invasive H. axyridis and the native C. septempunctata consumed more aphids than the three smaller native ladybird species. In line with our expectations, aphids avoided leaves bearing cues of most native ladybird species but not of the invasive H. axyridis.
4. Our results indicate that body size rather than ladybird origin determined aphid predation rates. The lack of aphid avoidance behaviour towards cues of H. axyridis indicates that they were not able to recognise the chemical cues of the invasive predator.
5. Relatively large body size and the absence of cue avoidance in aphids might benefit the invasive H. axyridis, particularly in comparison to smaller native ladybird species. The absence of avoidance behaviour in aphids might lead to even higher predation rates of H. axyridis under more natural conditions.

     

Nitrogen effects on an interaction chain in a salt marsh community

Nutrients can structure communities by influencing both plant interactions and plant herbivore interactions, though rarely do studies integrate these processes. In this study we examined how nitrogen fertilization influenced (1) the positive interaction between the marsh elder, Iva frutescens, and the black rush, Juncusgerardi, and (2) the quality of Iva as a host plant for the aphid, Uroleuconambrosiae. Previous studies have shown that by mitigating soil salt accumulation and hypoxia, Juncus is essential to the survival of Iva and its aphid herbivore at mid-marsh elevations. To address the effects of nitrogen on this interaction, we compared fertilized and unfertilized Iva plants subject to Juncus removal and control treatments in the field. Additionally, we measured the monthly population growth rates of aphids transplanted onto these Iva plants. Iva leaf biomass and flower number results indicated that fertilizing Iva eliminated its dependence upon Juncus, such that fertilized plants grown without Juncus were not different from unmanipulated plants. Aphid monthly population growth rates through mid-summer revealed that fertilization also eliminated the indirect dependency of aphids on Juncus, so that aphid growth rates on fertilized Iva without Juncus neighbors were similar to rates on unmanipulated Iva. Results also indicated that fertilizing Iva grown with Juncus increased Iva size, potentially enabling these plants to support larger aphid populations. Our results suggest that only under conditions of nitrogen limitation are the positive effects of Juncus essential to the mid-marsh persistence of Iva and its aphid herbivore. Furthermore, we found that nitrogen effects on aphid populations may arise not only from a direct effect of nutrients on Iva size but also through the indirect effects of nitrogen on the interaction between Juncus and Iva. We argue that studies integrating processes occurring both within and between trophic levels, are important to fully understanding the community-wide effects of nutrients. Received: 14 November 1997 / Accepted: 11 May 1998     

Density-dependent parasitism of cowpea aphid,Aphis craccivora by Lysiphlebus fabarum,Binodoxys acalephae,and Aphidius matricariae (Hymenoptera: Braconidae: Aphidiinae)

Biological control, as a major component of pest management strategies, uses natural biological agents to reduce pest populations. Studying the interaction among Aphis craccivora and its parasitoids including, Lysiphlebus fabarum, Binodoxys acalephae, and Aphidius matricariae in 2016 and 2017 in Tehran Parke-Shahr, showed positive, significant correlations in all cases between the densities of three parasitoid species and that of aphid nymphs and adults. The density of the parasitoids increased by increasing the density of the aphids. The parasitoids showed aggregative behavior in response to different densities of the host. There was a positive density-dependent correlation between the density of A. craccivora and rate of parasitism. Parasitism rates of nymphs and adult aphids by L. fabarum, B. acalephae, and A. matricariae increased or decreased along with decline or increase in the population of the aphid host. In 2016 spring, the highest rates of parasitism on aphid nymphs by L. fabarum, B. acalephae, and A. matricariae were 46.82, 23.09, and 17.16%, respectively. In 2017 spring, the highest rates of parasitism on aphid nymphs by L. fabarum, B. acalephae, and A. matricariae were 48.97, 21.77, and 15.06%, respectively. So, given the accordance between changes in aphid population and that of parasitoids, and parasitoids’ efficacy in Tehran’s polluted air, they can be used as biological agents in the management of A. craccivora population.     

Fly disturbance suppresses aphid population growth

1. The presence of predators is known to suppress prey populations not only by consumptive but also by non-consumptive effects, as it stresses the prey inducing costly changes to behaviour and physiology. However, there is recent evidence that disturbance from non-predacious, non-competing commensals can also negatively affect herbivore performance. 2. Populations of cherry-oat aphids (Rhopalosiphum padi L.) were initiated with adult aphids in mesocosms containing seedling wheat grass. Following aphid establishment, fruit flies (Drosophila melanogaster Meigen) were added at a relatively high density to half of the mesocosms and the aphids were left for another 5 days to reproduce. The experiment was performed over two blocks at 24 ± 4 °C. It was hypothesized that a relatively high density of commensals would stress the aphids and reduce their fitness, causing lower population growth and resulting in lower population sizes. 3. Aphid numbers were significantly lower in mesocosms with commensal flies after 5 days of fly presence across the two experimental blocks, documenting that fly disturbance suppresses aphid fitness and population growth. 4. The negative effect of fruit flies on aphid population growth must have come from the disturbance that flies imposed on the aphids in their search for food, indicating that the flies stressed the aphids. Thus, our study indicates that commensals may stress herbivores that do not distinguish between enemies and other active species in their environment, adding in the overall herbivore top-down control through fitness costs.     

Field manipulation of populations of individual staphylinid species in cereals and their impact on aphid populations

Abstract.

• 1 Four species of staphylinids were caged separately on wheat and their effect on the population development of the grain aphid, Sitobion avenae (F.), was investigated over two seasons.
• 2 A novel field cage design was used to exclude all but the staphylinid species under investigation.
• 3 Predator effects were compared with total-exclusion cages in which all predators and parasitoids were kept out.
• 4 During a high-density phase of aphid population development only Philonthus cognutus (Stephens) significantly influenced aphid population numbers.
• 5 At low aphid densities, Tachyporus obtusus (L.), T.chrysomelinus (L.) and P.cognatus reduced aphid population levels.
• 6 The experiments showed that Tachyporus spp. can reduce the numbers of cereal aphids prior to the exponential phase of aphid population increase and that P.cognatus causes some reduction at both the stages of aphid population growth investigated.

     

Patterns of host selection by four species of aphidiid (Hymenoptera) parasitoids: influence of host switching

Abstract.

• 1 We tested switching behaviour in four species of aphidiid parasitoids, using a two-aphid experimental system consisting of second-instar nymphs of pea aphid (Acyrthosiphon pisum (Harris)) and alfalfa aphid (Macrosiphum creelii Davis) feeding on broad beans in the laboratory.
• 2 Aphidius ervi Haliday, A.pisivorus Smith, A.smithi Sharma & Subba Rao, and Pram pequodorum Viereck showed an innate preference for pea aphid when both host species were provided in equal numbers.
• 3 Wasps encountered both aphid species equally but differed in their acceptance of alfalfa aphid. Females of A.pisivorus and P.pequodorum accepted alfalfa aphids when few pea aphids were available, but A. smithi always concentrated attacks on pea aphid. Aphidius ervi super-parasitized an increasing proportion of pea aphids as their availability declined.
• 4 Switching to the alfalfa aphid occurred in A.ervi and P.pequodorum (but not in A.pisivorus and A.smithi) under the condition of a 1:3 ratio of pea aphids:alfalfa aphids. Wasps did not switch when more pea aphids than alfalfa aphids were provided (3:1 ratio).
• 5 Alfalfa aphids were more likely than pea aphids to escape from parasitoid attack.
• 6 Switching to the most abundant host may not be adaptive in these four species of aphid parasitoids. A foraging wasp incurs a potentially higher cost in lost opportunity time when attacking (and failing to oviposit in) alfalfa aphids. In addition, alfalfa aphids may have lower host quality than pea aphids, a difference that could influence offspring fitness.

     

Population density of aphids at the time of settling and ovariole maturation inCoccinella septempunctata [Col.: Coccinellidae]

In 1978 and 1979 in Central Bohemia the census of aphid populations on bean, sugar beet, cereals, maize, and alfalfa were taken at the time when overwinteredCoccinella septempunctata L. (1) settled and (2) the ovarioles of females ripened. Settling occurred immediately after aphid immigration. Minimum aphid density required for settling was estimated to about 10 aphids per 1 m² of field area. The aphid density at the time of ovariole ripening equaled 1 aphid per 200–400 cm² of leaf area.     

Aphids decelerate litter nitrogen mineralisation through changes in litter quality

相似文献   

Patch size of gall-forming aphids: deme formation revisited

Plants of the genus Pistacia (Anacardiaceae) serve as obligate hosts for a group of specialized gall-forming aphids (Homoptera: Fordinae). The aphids regularly migrate between the Pistacia (primary) host plants and the roots of non-specific grasses and cereals (secondary hosts). Gall density varies considerably between trees and sites. The intimate relationships between the aphids and their primary host, the natural variation of host susceptibility, and the heterogeneous geographical environment may promote local adaptation and deme formation in the aphid populations. Indeed, previous analyses of the genetic structure of the aphid Baizongia pistaciae, which forms large galls on the deciduous P. palaestina trees, suggested deme formation (Martinez et al. 2005). In this study, we analyzed the genetic structure of the B. pistaciae population at eight sites and 78 trees throughout Israel and a single population in Turkey, using two molecular markers (AFLP fingerprints and COI sequencing). The genetic distance between the Israeli populations was found to be low (D = 0.01–0.02), and there was no genetic differentiation found between any population pairs. In five of the Israeli populations, we also compared the genetic identity between aphids forming galls on the same tree and between galls on neighboring trees. The analysis indicated that the genetic identity of different galls within a tree resembles the correspondence between trees within a population. Our results showed no indication of deme formation or any hierarchical genetic substructuring within B. pistaciae populations in Israel. The extensive gene flow between aphid colonies may be explained by their dispersal abilities and the potential bridging role of the secondary hosts.