Can Horton hear the whos? The importance of scale in mosquito-borne disease.

The epidemiology of vector-borne pathogens is determined by mechanisms and interactions at different scales of biological organization, from individual-level cellular processes to community interactions between species and with the environment. Most research, however, focuses on one scale or level with little integration between scales or levels within scales. Understanding the interactions between levels and how they influence our perception of vector-borne pathogens is critical. Here two examples of biological scales (pathogen transmission and mosquito mortality) are presented to illustrate some of the issues of scale and to explore how processes on different levels may interact to influence mosquito-borne pathogen transmission cycles. Individual variation in survival, vector competence, and other traits affect population abundance, transmission potential, and community structure. Community structure affects interactions between individuals such as competition and predation, and thus influences the individual-level dynamics and transmission potential. Modeling is a valuable tool to assess interactions between scales and how processes at different levels can affect transmission dynamics. We expand an existing model to illustrate the types of studies needed, showing that individual-level variation in viral dose acquired or needed for infection can influence the number of infectious vectors. It is critical that interactions within and among biological scales and levels of biological organization are understood for greater understanding of pathogen transmission with the ultimate goal of improving control of vector-borne pathogens.

Food availability alters the effects of larval temperature on Aedes aegypti growth.

Variation in temperature and food availability in larval habitats can influence the abundance, body size, and vector competence of the mosquito Aedes aegypti. Although increased temperature has energetic costs for growing larvae, how food resources influence the developmental response of this mosquito species to thermal conditions is unknown. We explored how rearing temperature and food affect allometric scaling between wing size and epidermal cell size in Ae. aegypti. Mosquitoes were reared at 22 and 28 degrees C across a gradient of field-collected detritus designed to simulate commonly observed natural larval food resources. Overall, reduced temperature and increased food level increased wing size, but only temperature affected cell size. Females fed the least food had the longest time to maturation, and their increases in wing size induced by cold temperature were associated with larger, rather than more, cells. By contrast, males fed the most food had the shortest time to maturation, and their increases in wing size induced by cold temperature were associated with more, rather than larger, cells. Therefore, food levels can alter the underlying physiological mechanisms generating temperature-size patterns in mosquitoes, suggesting that the control of development is sensitive to the combination of nutrient and thermal conditions, rather than each independently. Conditions prolonging development time may favor increased cell division over growth. We suggest that understanding the effects of climate change on Ae. aegypti vectorial capacity requires an improved knowledge of how water temperature interacts with limited food resources and competition in aquatic container habitats.

Temperature induces trade-offs between development and starvation resistance in Aedes aegypti (L.) larvae.

Heightened temperature increases the development rate of mosquitoes. However, in Aedes aegypti (Diptera: Culicidae), the larvae of which commonly experience limited access to food in urban habitats, temperature effects on adult production may also be influenced by changes in the capacity of larvae to survive without food. We carried out experiments to investigate the effects of temperatures increasing at intervals of 2 °C from 20 °C to 30 °C on the growth, maturation rate and longevity of optimally fed larvae placed in starvation. Overall, both growth rate and starvation resistance were lower in the first three larval instars (L1-L3) compared with L4, in which growth of >75% occurred. Although increasing the temperature reduced the duration of each instar, it had a U-shaped impact in terms of the effect of initial growth on starvation resistance, which increased from L1 to L2 at 20 °C and 30 °C, remained constant at 22 °C and 28 °C, and decreased at 24 °C and 26 °C. Growth from L2 to L3 significantly increased starvation resistance only from 26 °C to 30 °C. Increased temperature (>22 °C) consistently reduced starvation resistance in L1. In L2-L4, increments of 2 °C decreased starvation resistance between 20 °C and 24 °C, but had weaker and instar-specific effects at >24 °C. These data show that starvation resistance in Ae. aegypti depends on both instar and temperature, indicating a trade-off between increased development rate and reduced starvation survival of early-instar larvae, particularly in the lower and middle temperatures of the dengue-endemic range of 20-30 °C. We suggest that anabolic and catabolic processes in larvae have distinct temperature dependencies, which may ultimately cause temperature to modify the density regulation of Ae. aegypti populations.

Ecological links between water storage behaviors and Aedes aegypti production: implications for dengue vector control in variable climates.

Understanding linkages between household behavior and Aedes aegypti (L.) larval ecology is essential for community-based dengue mitigation. Here we associate water storage behaviors with the rate of A. aegypti pupal production in three dengue-endemic Colombian cities with different mean temperatures. Qualitative, semi-structured interviews and pupal counts were conducted over a 7-15-day period in 235 households containing a water storage vessel infested with larvae. Emptying vessels more often than every 7 days strongly reduced pupal production in all three cities. Emptying every 7-15 days reduced production by a similar magnitude as emptying <7 days in Armenia (21.9 degrees C), has a threefold smaller reduction as compared to <7 days in Bucaramanga (23.9 degrees C), and did not reduce production in Barranquilla (29.0 degrees C). Lidding vessels reduced mosquito production and was most feasible in Barranquilla because of container structure. Vessel emptying strongly correlated with usage in Barranquilla, where many households stored water in case of interruptions in piped service rather than for regular use. In the cooler cities, >90% of households regularly used stored water for washing clothes, generating a weaker correlation between emptying and usage. Emptying was less frequent in the households surveyed in the dry season in all three cities. These results show that A. aegypti production and human behaviors are coupled in a temperature-dependent manner. In addition to biological effects on aquatic stages, climate change may impact A. aegypti production through human behavioral adaptations. Vector control programs should account for geographic variation in temperature and water usage behaviors in designing targeted interventions.

Transmission patterns of African horse sickness and equine encephalosis viruses in South African donkeys.

African horse sickness (AHS) and equine encephalosis (EE) viruses are endemic to southern Africa. AHS virus causes severe epidemics when introduced to naive equine populations, resulting in severe restrictions on the movement of equines between AHS-positive and negative countries. Recent zoning of South Africa has created an AHS-free zone to facilitate equine movement, but the transmission dynamics of these viruses are not fully understood. Here, we present further analyses of serosurveys of donkeys in South Africa conducted in 1983-5 and in 1993-5. Age-prevalence data are used to derive estimates of the force of infection, A. For both viruses, A was highest in the northeastern part of the country and declined towards the southwest. In most of the country, EE virus had a higher transmission rate than AHS. The force of infection increased for EE virus between 1985 and 1993, but decreased for AHS virus. Both viruses showed high levels of variation in transmission between districts within the same province, particularly in areas of intermediate transmission. These data emphasize the focal nature of these viruses, and indicate areas where further data will assist in understanding the geographical variation in transmission.

Simulation studies of St. Louis encephalitis virus in south Florida.

Two simulation models were used to investigate the epidemiology of St. Louis encephalitis virus (SLEv) in south Florida, one including sentinel hosts (chickens) and amplification hosts (wild birds), while the other one included age structure in the amplification host population. The overall population size of the vector, Culex nigripalpus, was a major factor in the likelihood of epizootics for both models, but the seasonal dynamics of the vector alone did not explain variation in transmission. Interactions between seasonal factors in the mosquito and reproduction in the wild amplification avian hosts appeared to be important in the likelihood of epizootics. Biased feeding between sentinel and amplification hosts affected the time course of virus prevalence and may have implications for the interpretation of sentinel data. The time of virus introduction strongly affected the timing of outbreaks but did not affect the likelihood of epizootics. In most cases, the outbreak occurred immediately after virus introduction; however, in some cases the outbreak was delayed until the mosquito population increased. This has implications for the timing of control strategies directed against mosquito populations.

Simulation studies of St. Louis encephalitis and West Nile viruses: the impact of bird mortality.

West Nile virus (WNv) has spread through much of the eastern United States following its introduction in 1999, and arrived in Florida in 2001. Prior to its arrival, we anticipated that its transmission cycle was likely to be similar to that of St. Louis encephalitis virus (SLEv). However, high levels of avian mortality have been reported for WNv in the northeastern United States, and it was unknown how this would impact the transmission dynamics of WNv. Simulation models were used to compare the two viruses by considering the impact of bird mortality on the transmission dynamics of arboviruses in south Florida. Transmission models without disease-induced mortality (SLEv) were compared with models including disease-induced mortality (WNv). Disease-induced mortality depressed transmission, eliminating epizootics in two of 14 simulations that were epizootic without the additional mortality. In both models, the most important factor in the likelihood of epizootics was mosquito population size; the mosquito mortality rate was also important. The additional avian mortality altered the factors most important in the size and timing of epizootics, although it did not always directly affect the outcome of the simulations. In some cases, low-level transmission occurred prior to the epizootic peak. When disease-induced avian mortality was included in the simulations, appreciable numbers of dead birds occurred prior to high levels of infection in mosquitoes. This has implications for the use of dead birds as a surveillance tool monitoring the spread and transmission of WNv.

First record of Amblyomma auricularium (Acari: Ixodidae) in the United States.

Amblyomma auricularium (Conil) was collected from an armadillo in Glades County, FL. Two females and 11 males were collected. This is the first record of this species in the United States.

Attractiveness of chickens and bobwhite quail for Culex nigripalpus.

An experiment to determine if wild Culex nigripalpus preferentially enter lard-can traps baited with chickens or bobwhite quail was conducted. A strong preference for the chicken-baited trap was observed. Further development of transmission models for St. Louis encephalitis virus in southern Florida should consider increased biting rates on sentinel chickens and the consequences for the interpretation of sentinel seroconversion data.

The relationship between size and fecundity in Aedes albopictus.

Recent use of correlates of the intrinsic rate of increase and models to study population dynamics in mosquitoes has emphasized the need for species specific data on the relationship of morphological factors such as body size to life history parameters such as fecundity. Aedes albopictus is of concern because of its continued spread and population growth in the U.S. and the potential for virus transmission. We examined the relationships between pupal wet weight and adult female wing length and between female wing length and fecundity in the first gonotrophic cycle in U.S. strains of Ae. albopictus. There was a significant positive relationship between pupal wet weight and adult wing length. The relationship between wing length and fecundity was also significant and positive, but with lower correlation coefficients and higher variance indicating that other factors also affect fecundity. The regression coefficients for the fecundity-wing length equation did not differ between two strains of Ae. albopictus.

Estimation of survival rates in haematophagous insects.

Estimation of the survival rate through a gonotrophic cycle is an important factor in determining the vectorial capacity of a population of haematophagous insects in a disease cycle. Most methods used to calculate survival rates make stringent assumptions which may not be valid for all species. Birley and colleagues used a time series analysis of samples collected over several consecutive days, the lagged parous rate. Here, we use a simulation model to investigate (i) the length of data series needed and (ii) the consequences of failures in the assumptions of this method for the estimated survival rate. The accuracy of the estimated survival rate per cycle was high with sample periods of 10-100 days. The standard deviation (a measure of precision) decreased with the length of the sample period. When random sampling efficiency was included, the accuracy remained high but the estimates were less precise (larger standard deviations). If the sampling was biased in favour of either nulliparous or parous females, estimates of the survival rate were not accurate. The relationship between estimated survival rate, bias in collection, and true survival rate was non-linear. Thus, correction for the bias requires (i) prior knowledge of the direction and the severity of the bias and (ii) an independent estimate of the survival rate. This method of estimating survival rates is less accurate when the collection method is biased for or against parous females, although robust to other assumptions.

Aggregation and distribution of strains in microparasites.

Recent research has shown that many parasite populations are made up of a number of epidemiologically distinct strains or genotypes. The implications of strain structure or genetic diversity for parasite population dynamics are still uncertain, partly because there is no coherent framework for the interpretation of field data. Here, we present an analysis of four published data sets for vector-borne microparasite infections where strains or genotypes have been distinguished: serotypes of African horse sickness (AHS) in zebra; types of Nannomonas trypanosomes in tsetse flies; parasite-induced erythrocyte surface antigen (PIESA) based isolates of Plasmodium falciparum malaria in humans, and the merozoite surface protein 2 gene (MSP-2) alleles of P. falciparum in humans and in anopheline mosquitoes. For each data set we consider the distribution of strains or types among hosts and any pairwise associations between strains or types. Where host age data are available we also compare age-prevalence relationships and estimates of the force of infection. Multiple infections of hosts are common and for most data sets infections have an aggregated distribution among hosts with a tendency towards positive associations between certain strains or types. These patterns could result from interactions (facilitation) between strains or types, or they could reflect patterns of contact between hosts and vectors. We use a mathematical model to illustrate the impact of host-vector contact patterns, finding that even if contact is random there may still be significant aggregation in parasite distributions. This effect is enhanced if there is non-random contact or other heterogeneities between hosts, vectors or parasites. In practice, different strains or types also have different forces of infection. We anticipate that aggregated distributions and positive associations between microparasite strains or types will be extremely common.

Transmission and distribution of African horse sickness virus serotypes in South African zebra.

The prevalences of African horse sickness (AHS) virus serotypes in zebra foals from the Kruger National Park, South Africa were examined for possible associations between serotypes. Serotypes known to cross-react were combined for analysis. The distributions of serotypes between zebra were not always independent; in 7-8 month old zebra positive pairwise associations were observed between 3 serotypes. This could be generated by biological interactions between serotypes or heterogeneity in host-vector transmission. The data were also used to estimate the basic reproduction number, R0. For AHS virus overall, estimates of R0 ranged from 31-68. This underlines the need for a better understanding of serotype transmission and interactions in AHS.

Simulation studies of African horse sickness in Spain.

Factors affecting epidemics of African horse sickness in Spain were studied using a mathematical model. The model examined the likelihood of an epidemic after the introduction of the virus, and the effectiveness of vaccination strategies. Two host species (horses and donkeys) and one vector species (the biting midge Culicoides imicola) were included. A stratified random sampling method (Latin hypercube sampling) was used for sensitivity analysis of the likelihood of an epidemic. Systematic variation of vaccination parameters was used to consider alternative control strategies. In general, when an epidemic occurred most potential hosts became infected. The peak prevalence in C. imicola was low, and never exceeded 3%. The most significant factors in the likelihood of an epidemic were vector population size, the recovery rate in horses and the time of year when the virus was introduced. The lag between virus introduction and protection, the proportion of hosts vaccinated, and including donkeys in vaccination programmes where the factors that most strongly affected the success of different vaccination strategies. These factors should be priorities for empirical research, and should be considered in the design of control strategies in areas at risk of virus introduction.

Density dependence in larval Aedes albopictus (Diptera: Culicidae).

Aedes albopictus (Skuse) is expanding its distribution in the United States and elsewhere, and a better understanding of its population regulation is needed. A field experiment under seminatural conditions determined that density had a negative effect and food level a positive effect on immature survival, duration of development, and female size at emergence. A summary index (r') indicated that population growth also depended on density and food availability. These data can be used to estimate the relationships needed in the development of mathematical models for Ae. albopictus.

Simulation studies of vaccination strategies in African horse sickness.

A simulation model including two hosts (horses and donkeys) and one vector (Culicoides imicola) for African horse sickness in Spain is extended to consider vaccination strategies. If hosts were protected prior to virus introduction, elimination of simulated epidemics was related nonlinearly to the fraction protected. Protecting donkeys as well as horses increased the effectiveness of vaccination. Prevention of 50% of epidemics required 75% coverage of horses and donkeys or 90% coverage of horses only. Protection after the introduction of the virus was rarely successful in preventing outbreaks. If horses alone were protected, the number of donkeys was the most significant factor determining the level of protection needed to prevent an epidemic. If both hosts were protected, the abundance of other hosts for vector blood meals was the most significant factor. These results suggest that prophylactic vaccination of both horses and donkeys with high coverage is necessary to prevent outbreaks of African horse sickness in Spain.

Transmission and distribution of virus serotypes: African horse sickness in zebra.

The prevalence of African horse sickness (AHS) serotypes in zebra foals from the Kruger National Park, South Africa was examined for possible associations between serotypes and to estimate the basic reproduction number, R0. The distributions of serotypes between zebra were not independent in the 6- and 7-8-month-old age classes (P < 0.005). This does not necessarily imply biological interactions between serotypes, as heterogeneity in host-vector transmission rates can generate non-independent distributions of serotypes. Both age and month of capture were significant factors in the number of serotypes infecting each zebra (P < 0.0001). Pairwise, positive associations between non-cross-reacting serotypes were found in the 7-8-month-old class only. For AHS overall, estimates of R0 ranged from 31-68. Assuming serotypes are transmitted independently, estimates of R0 for individual serotypes ranged from 10 for serotype 1 to 23 for serotype 6. The wide range of estimates emphasizes the need for a better understanding of serotype transmission and interactions in AHS.

Simulation studies of African horse sickness and Culicoides imicola (Diptera:Ceratopogonidae).

A simulation model of African horse sickness in Spain was developed to investigate what factors affect the likelihood of an epidemic after the introduction of the virus. The model included 2 host species (horses and donkeys) and 1 vector species (Culicoides imicola Kieffer). Latin hypercube sampling was used for sensitivity analysis of the model, to include uncertainty in parameter estimates. In general, if an epidemic occurred most hosts were infected. The peak prevalence in midges was low, and never exceeded 3%. Midge population size, the recovery rate in horses, and the time of year when the virus was introduced were the most significant factors in determining whether or not an epidemic occurred. The uncertainty in interbloodmeal interval, removal rate (mortality and recovery) of infectious horses, midge population size, and transmission rates were significant factors in the size of the epidemic. These factors should be priorities for empirical research, and should be considered in the design of control strategies in areas at risk of virus introduction.

Vector-borne diseases and the basic reproduction number: a case study of African horse sickness.

The basic reproduction number, R0, can be used to determine factors important in the ability of a disease to invade or persist. We show how this number can be derived or estimated for vector-borne diseases with different complicating factors. African horse sickness is a viral disease transmitted mainly by the midge Culicoides imicola. We use this as an example of such a vector-transmitted disease where latent periods, seasonality in vector populations, and multiple host types may be important. The effect of vector population dynamics which are dependent on either host or vector density are also addressed. If density-dependent constraints on vector population density are less severe, R0 is more sensitive to vector mortality and the virus development rate. Host-dependent vector dynamics change the relationship between R0 and host population size. Seasonality can either increase or decrease the estimate of R0, depending on the lag between the peak of the midge population and the infective host population. The relative abundance of two host types is a factor in the ability of a disease to invade, but the strength of this factor depends on the differences between the hosts in recovery from infection, mortality and transmission. Removal of a reservoir host may increase R0.

Seasonal activity of nymphal Ixodes scapularis (Acari: Ixodidae) in different habitats in New Jersey.

Activity patterns of nymphal Ixodes scapularis Say were compared between habitat types (dominant tree types: mixed deciduous, oak, white pine, red cedar, sassafras, and spicebush). Both the time of peak abundance and the relative abundance of questing nymphs at the peak were compared. Several smoothing algorithms were tested with the data to determine if they could be used to estimate the time of peak abundance more accurately. Determination of the time of peak abundance using the raw data or simple moving averages was susceptible to outliers. Weighted averages were less susceptible to outliers. The seasonal pattern of nymphal abundance was similar in all habitat types. Variation in the time of peak abundance between habitats was low. Peak densities were lower in deciduous habitats (0.24 +/- 0.05 nymphs per square meter) than in nondeciduous habitats (0.85 +/- 0.15 nymphs per square meter); this could have resulted from higher host use of the nondeciduous areas. These data suggest that there are differences in the population dynamics of nymphs found in different habitats.