Hypotheses, Settings and Workplan

Back to top

The Settings and Factors of the Epidemic and subsequent Endemism

The epidemiology of malaria results from complex interactions between the human host, the mosquito vector, and the Plasmodium parasite. Each of these constituent components can be influenced, positively or negatively, by changes in ecological conditions. According to the World Health Organization (WHO), an estimated 300-500 million people are infected with malaria each year, causing 1-2 million deaths annually, and it is estimated that 40% of the global population is at risk of contracting malaria (WHO, 2004). The populations at risk are located primarily in tropical regions, where most of the world’s poorest countries are located. Malaria is a debilitating disease that has been shown to be a significant economic burden by preventing affected individuals from working or carrying out regular tasks in communities with high levels of infection (Sachs, 2002). Thus, malaria further compounds the socioeconomic disparities in these regions by disrupting vital subsistence activities among both individuals and communities. Increases in malaria incidence have been attributed to climate change, vector resistance to insecticides, altered land-use by humans, and human migration (Martens & Hall 2000; Patz et al. 1996; Walsh et al. 1993).  It is critical to understand the pattern of migration, resettlement and mobility.

Back to top

a. Mobility

In order for a disease such as malaria to become established and spread into new areas, it must first be introduced into those new habitats. The current speed and frequency of global human travel has facilitated introductions of species and infectious diseases. Thus human migration, especially in mass, has been recognized as a key factor in the spread of malaria (Walsh et al., 1993; Patz et al., 2000; Martens and Hall, 2000; Conn et al., 2002). Malaria, in relation to migration, can be spread in at least two ways. First, migrants can import the disease endemic to their native locality into non-immune communities. Migrations of this sort have brought malaria to novel regions, and have also been known to move malaria back into regions where it was once eradicated (Martens and Hall, 2000). Secondly, malaria can be harbored in communities that have acquired natural levels of immunity and are asymptomatic. Migrants who colonize and co-inhabit an area are then susceptible to contracting the disease (Patz et al., 2000). The movement of malaria along with human migration has occurred through government-sponsored settlement projects (Castro and Singer, 2003, 2005), economic fluctuations (Epstein and Selber, 2002; Witzig and Ascencios, 1999), and through development projects promoted by governmental and non-governmental organizations (Lambin et al., 2001). The vast majority of studies looking at the effect of migration on increased prevalence of malaria focus on epidemics arising from rural to urban migration. We are interested in focusing on the peri-urban settlements, which usually show a higher level of malaria prevalence, compared with both the rural or urban areas, which can be attributed to rapid changes in land-use, unplanned settlements, and the high mobility of populations that inhabit such communities (Singer and Castro, 2001; Castro and Singer, 2003).

Populations in Amazonia have long been highly mobile, with not only members of households, but even whole communities occasionally moving from place to place. Padoch and de Jong (1992), for instance, found that the community of Santa Rosa moved five times within 50 years. Pinedo-Vasquez et al (2002), recounted the fate of villages that disappeared while others were formed due to the lateral erosion and sedimentation of river channels in the last 20 years in the area of Muyuy, not far from Iquitos.  A considerable portion of the population of Muyuy moved to settlements along the Iquitos-Nauta road in the 1980s and 1990s (Barletti, personal communication). 

Back to top

b. Development - the Iquitos Nauta Road

Landscape development, such as road building, may be seen as a form of progress within rural communities, but negative consequences often accompany it. A road transforms the physical conditions of the land it is constructed on, as well as the areas adjacent to it (Trombulak and Frissell, 2000). Hydrological disruption associated with road building may negatively impact some species; Anopheles darlingi populations, however, are known to have increased in the Amazon Basin following the construction of the Iquitos-Nauta road, which created prime breeding sites (Conn et al., 2002). A further negative aspect of road construction is its incremental impact on forest cover (Maki et al., 2001). Deforestation first occurs as a direct result of road construction, but more detrimental deforestation occurs after it is constructed. A road allows humans to access resources in areas that were formerly remote. This highlights the fact that development has the potential to increase malaria by combining aggravating features, such as deforestation, hydrological shifts, and a large migratory work force (Wolfe et al., 2000; Patz et al., 2000).

The residents of the two urban centers connected by the Iquitos-Nauta road considered its construction a priority, as an important advance contributing to improved economic opportunities (Maki et al., 2001). The construction process was slow and steady for the first decade, but combined multiple factors for increased risk of malaria. During this time active settlement was promoted along the road, through low-interest agricultural loans. Without an organized and effective plan for colonization, however, the opening of roads is synonymous with deforestation, increased  migration, pressure on natural resources, disease outbreaks, and poverty (Castro and Singer, 2003). In addition, an important governmental stipulation to acquiring property required land clearing as evidence of development intentions. This led to deforestation in some areas where agricultural development was never the intended outcome. The economic instability of the 1980s created a collapse in financial support for newly acquired lands, and resulted in eventual abandonment of these cleared areas (Maki et al., 2001). The deforested plots consequently produced extensive breeding habitat for mosquitoes. Construction of the road was inconsistent, with recurring periods of stagnation and delays. Uncertain circumstances, lack of services, and poor health conditions led to repeated occupancy and vacancy of land; each time the road extended it was accompanied by extraction, deforestation, development and abandonment.

Back to top

c. Aggravating factor in the Research Area - Anopheles darlingi

The aggravating factor contributing to a rise in malaria prevalence around Iquitos, apart from road construction, deforestation and human migration, includes the introduction of a new and more efficient mosquito vector. The most important vector transmitting malaria in Amazon regions of South America is Anopheles darlingi (Singer and Castro, 2001). Between 1988 and 1991, however, mosquito collections were made at three separate locations near Iquitos and of 35,000 mosquitoes collected, representing 13 genera and 25 species, A. darlingi was absent (Schoeler et al., 2003). A similar study was conducted near the Ecuadorian border in which 4,000 anopheline mosquitoes were collected, and again, no A. darlingi were detected (Schoeler et al., 2003). Larvae of this invasive species are thought to have been introduced to the Iquitos region by boats transporting tropical fish for the aquaculture industry (Lounibos, 2002), and the increased spread of malaria, particularly P. falciparum, in the Peruvian Amazon has been attributed to its introduction (Aramburu et al., 1999; Conn et al., 2002; Povoa et al., 2003; Schoeler et al., 2003). A. darlingi is considered to be an efficient vector because it is highly anthropophilic, readily enters homes to feed, has a wide range of feeding times, and is susceptible to both P. vivax and P. falciparum (Flores-Mendoza et al., 2004). Since its introduction in the Departmento of Loreto A. darlingi has continued to invade previously uninfested areas. It is now the most abundant mosquito species during the wettest parts of the year, making up nearly 90% of the mosquito population, and still remains a major anopheline species through the dry season as well (Roshanravan et al., 2003; Aramburu et al., 1999).  A. darlingi has a heterogenous distribution, so while it is not the most abundant species at all sites within Loreto, it has been found to be the most widespread (Schoeler et al., 2003; Flores-Mendoza et al., 2004). A. darlingi breeds in partially shaded, neutral, unpolluted, relatively still waters (Walsh et al., 1993). Suitable breeding sites include shallow pools left after floods recede, borrow pits after road construction, fish farms, edges of small rivers, swamps, and small pools on cleared land (Walsh et al., 1993; Aramburu Guarda et al., 1999; Roshanravan et al., 2003).  The diversity of suitable habitat also demonstrates the species’ ability to colonize an area even in the absence of deforestation.

Non-producing fish pond

Back to top

d. Aquaculture

Non-governmental organizations promoted aquaculture as a sustainable form of economic development benefiting poor communities, with minimal environmental degradation. Although the increase in malaria infection rates that has followed this development intervention was neither expected nor intended, it could have been anticipated by examining the ecological and epidemiological conditions of the Iquitos region (or those of almost any tropical region). Vectors of many infectious diseases, including all mosquito species, require aquatic habitat for egg-laying and larval development. Water projects such as construction of reservoirs, canals, dams and irrigation systems have been closely linked to infectious disease by causing shifts in parasite or vector populations (Fraire, 1975; Patz et al. 2000; Vasconcelos and Novo, 2004; Bartram and Carr, 2003; Keiser et al., 2005). Such water projects can either increase an endemic rate of malaria, or create a new transmission area. Water projects, like any other development regime, are implemented with beneficial outcomes in mind. Governments have initiated programs to provide drinking water, irrigation, or individual and community poverty alleviation (Keiser et al., 2005; Ghebreyesus et al., 1999; Sissoko et al., 2004; Baudon et al., 1986). But poorly planned and unintentional outcomes can occur through a division of socioeconomic abilities and environmental costs. Health risks are particularly insidious, as malaria and other infectious diseases can have enormous economic impacts (Hunter et al., 1992; Sachs, 2002; Ghebreyesus et al., 1999). The promotion of aquaculture ponds possesses particular risk since they are small in size and rapidly produced, primarily on private landholder properties. This increases the amount of surface water near a home, thus bringing mosquito habitat closer to the home and altering the human-vector dynamics, resulting in more bites and more transmission opportunities. Human residences provide proper conditions for certain mosquito species to flourish by providing shelter, a guaranteed blood meal, and a more constant climate. The magnitude of malaria prevalence is related to the human exposure to infected mosquitoes (Roper et al., 2000). Given these two facts, one can predict an increase of malaria accompanying aquaculture. Additionally, aquaculture farmers working and maintaining the pond can have associated occupational risks. Furthermore, if ponds are abandoned, the lack of predatory fish preying on larvae can lead to dramatic disease outbreaks along with seasonal weather fluctuations, or poor condition of pond maintenance.

Back to top

Workplan

Region-level analysis is defined by municipal boundaries, encompassing all three units (urban, rural and peri-urban). Attributes at this level include demographic, economic, and geophysical events such as agrarian colonization, development projects (building fish ponds to supplement income), and migration forced by catastrophic floods.
Unit-level analysis is defined by land use and human population density. Attributes in the dataset for this level include number of households and fishponds, household composition, actual and proportional extent of malaria infection, and types of malaria.
Individual-level analysis is defined by the individual household. Attributes at this level include cultural background, economic activities over time, land tenure condition, mobility, and available capital resources.

Data collection scheme. Three spatial scales, A) individual, B) unit, and C) region; with four types of data (white squares). The units correspond to, from left to right, urban, peri-urban, and rural types of settlements.

Four kinds of data will be collected at the different scales: human demographic, socioeconomic, epidemiology, and landscape (mosquito breeding sites, fish ponds, floods, land cover). There have been prior studies on the relationship between abandoned ponds and malaria transmission in Iquitos (Aramburu et al., 1999; Alcantar et al., 2001), but none have combined the kinds of data that we propose to investigate, or integrated the entire range of human settlement, migration, displacement, and mobility in disease dynamics.

All four data sets will be compiled in a geographic information system (GIS) to create an epidemiological map and will be analyzed using two methods. These methods are: (1) multivariate analyses of factors influencing rate, extent, and direction of malaria dispersal within each unit, and (2) agent-based models to simulate patterns of malaria dispersal within and across units.