Attraction of mosquitoes to domestic cats in a heartworm enzootic region

Abstract

Heartworm disease is caused by a mosquito-borne parasite that can affect many different mammalian species and has worldwide distribution. The agent, Dirofilaria immitis (Leidy 1856), infect mainly dogs but feline infection have been frequently reported in the last decade. Feline heartworm infection is difficult to detect, therefore, low reported prevalence could reflect true low prevalence or poor diagnostic efficiency. As mosquitoes are known to be attracted differently by different mammalian species, mosquitoes were collected from both a cattery and a contiguous home located in a canine heartworm enzootic area in Niterói, state of Rio de Janeiro, Brazil. For 14 months, mosquitoes were collected weekly for genus identification, speciation when possible, and for individual blood meal identification. Culex species mosquitoes were the most captured and those most frequently found with feline blood meal, followed by Aedes species that, although captured in lower numbers, also fed on feline blood. While Culex species mosquitoes have been reported as potential secondary heartworm vectors for dogs and primary vectors for cats, the present results suggest that Aedes species mosquitoes may also be involved in feline heartworm transmission in a larger proportion than previously thought.

Dirofilaria immitis (Leidy 1856) is a widespread mosquito-borne nematode parasite of dogs that can infect cats, among other hosts (Guerrero et al 1992a). Feline heartworm infection parallels that of dogs in a given area, although at lower rates (Dillon 1986). In Italy, the prevalence of heartworm infection in cats from highly endemic areas (prevalence in dogs >70%) was low (<5%) (Genchi and Cancrini 1992). Experimentally, infection rates and the average number of live worms found per cat are low compared with those in dogs (Mansour et al 1995). Feline heartworm infection seems to be increasing although this could be due to better diagnostic skills and growing veterinary awareness (Guerrero et al 1992b, McCall et al 1994, Brown et al 1999, Knight et al 1999).

It is well known that different hosts attract mosquito species differently; Aedes albopictus collected in the United States, for example, contained blood meal extracts of avian, mammalian and reptile species when tested using precipitin and enzyme-linked immunosorbent assay (ELISA) techniques (Niebylski et al 1994). A shift in mosquito feeding habits could also play a role in changing feline heartworm disease prevalence. Some species will feed only on certain hosts, while others will feed according to attraction and host abundance (Deane 1951, Forattini et al 1987). Large animals might attract more mosquitoes in paired-host tests, particularly when mosquitoes are initially released (Edman et al 1974). To serve as a vector for D immitis, mosquitoes must survive the infection and live long enough to allow filarial development. Some anophelines, for example, kill most microfilariae ingested by the lethal effects of their fore-gut armatures, thereby precluding them as an effective vector (Zahedi and White 1994). Also, flight range and host seeking habits can also influence vector capability (Ludlam et al 1970, Otto and Jachowski 1989). To vector D immitis to cats requires that a species must feed both on dogs and cats because canidae are the principal reservoir host for transmission of the parasites to vector mosquitoes (McCall et al 1992).

In the state of Rio de Janeiro, canine heartworm infections are known to be frequent, especially in the county of Niterói (Labarthe et al 1997a). Although cats are susceptible to heartworm infection, infected cats are rarely found in Rio de Janeiro (McCall et al 1992). The highest feline infection rate reported in Rio de Janeiro is 1.6% (Labarthe et al 1997b) in areas where the reported canine prevalence was 30% (Labarthe et al 1997a). Vectors' feeding habits are seen as a reason for feline low infection rates. Wild hemisynanthropic mosquito species Aedes scapularis (Rondani) and Aedes taeniorhyncus (Wiedmann) have been shown to be the principal vectors of D immitis for dogs, while the endophilic Culex quiquefasciatus seems to be of secondary importance as a vector for dogs. It is, however, the only mosquito species commonly associated with cats (Genchi et al 1992, Labarthe 1998).

To determine the willingness of wild hemisynanthropic mosquito species to feed on cats when they are abundant and exposed, captures were performed at an outdoor cattery.

Materials and methods

The survey was performed at a section of the City of Niterói, where 26% of the canine population is known to harbor Dirofilaria immitis with microfilaremia, although no cat has been reported to be antibody positive (Gomes et al 2001a). The section is known as Engenho do Mato (22°56′17″S; 43°00′21″W), where the climate is subtropical and secondary rain forest is preserved on the hillside (Lacerda 1984).

Mosquito captures were performed weekly from the interior walls (December 1998 through January 2000) of an unscreened cattery that sheltered approximately 90 cats. All cats were kept together and not allowed to roam in the neighborhood. Outside the cattery wall, two dogs were kept free in the garden between the cattery and the house. Mosquitoes were also captured from the exterior walls of a contiguous home where three people lived (100 m away from the cattery) starting 1 month later. Captures lasted for 1 h, starting 1 h after sunset, and were performed concomitantly at the two sites by two volunteers using aspirators. Mosquitoes were immediately frozen (−20°C) until they could be identified using taxonomic keys of Lane (1953) and Consoli and Lourenço De Oliveira (1994). Immediately after taxonomic identification, their digestive contents were taken and diluted at 1:10 in phosphate-buffered saline to identify their blood meal source by precipitin test as feline, canine or human (Gomes et al 2001b).

The frequency distribution was used to analyze the discrete data and the Z test used to compare two proportions (Rodrigues 1993).

Results

A total of 227 female mosquitoes of different genera were collected. Due to freezing damage, species identification was possible only in Aedes individuals. Among Aedes species, the most common was Aedes scapularis (39/67). The most common genus was Culex (139/227) (Table 1). Blood meal analysis was performed in 109 individuals and of those, 20 mosquitoes did not react with either of the test antisera. The precipitin test showed that most Culex species mosquitoes fed on cats (79%) and that Aedes species mosquitoes (65%) also fed on cats, although the total population of Culex species was greater (Table 2). A higher proportion of Aedes species mosquitoes (30%) than Culex species individuals (25%) had both canine and feline blood in their guts. The other four genera (Coquillettidia, Mansonia, Psorophora and Wyeomyia) were also shown to feed on cats (Table 2).

Table 1.

Frequency of female mosquitoes captured in a cattery and contiguous house ^* in Niterói, RJ, Brazil, from December 1998 through January 2000

Species	House	Cattery	Total	%
Aedes albopictus Skuse	4	2	6	2.6
Aedes scapularis (Rondani)	6	33	39	17.2
Aedes taeniorhynchus Wiedemann	4	8	12	5.3
Aedes species †	6	4	10	4.4
Combined	20	47	67	29.5
Genus Culex †	58	81	139	61.2
Coquillettidia species	8	1	9	4
Mansonia species	1	1	2	0.9
Psorophora species	3	3	6	2.6
Wyeomyia species	2	1	3	1.3
Undetermined †	0	1	1	0.4
Total	92	135	227	100

^*100 m away from cattery.

^†Speciation not possible due to freezing damage.

Table 2.

Number of female mosquitoes captured in Niterói, RJ, Brazil according to blood meal source identified by the precipitin test, from December 1998 through January 2000

Mosquitoes	N	Feline	Canine	Feline+canine	Human	No reaction
Aedes species	37	13 (35.13%)	1 (2.7%)	11 (29.72%)	2 (5.40%)	10 (27.02%)
Culex species	61	33 (54.09%)	2 (3.27%)	15 (24.59%)	2 (3.27%)	9 (14.75%)
Coquillettidia species	4	0 (–)	0 (–)	3 (75%)	0 (–)	1 (25%)
Mansonia species	1	0 (–)	0 (–)	1 (100%)	0 (–)	0 (–)
Psorophora species	4	3 (75%)	0 (–)	1 (25%)	0 (–)	0 (–)
Wyeomyia species	2	2 (100%)	0 (–)	0 (–)	0 (–)	0 (–)
Total	109	51 (46.78%)	3 (2.75%)	31 (28.44%)	4 (3.66%)	20 (18.34%)

Aedes species were more prevalent at the cattery (47/135) than at the neighboring home (20/92; P=0.034, χ²=4.48, odds ratio (OR)=1.92, OR 95% confidence interval=1.00–3.70) and these were predominately Aedes scapularis.

Discussion

The apparently low density of mosquitoes at the site shows that captures were performed after the peak feeding time of mosquitoes, therefore, the interference of volunteers in the mosquitoes feeding habits was minimized. Some mosquitoes like Ae aegypti are assumed to be anthropophilic (Christophers 1960), and studies in Africa and Florida have shown that the percentage of mosquitoes feeding on humans is typically high (Scott et al 1993). Despite the low number of mosquitoes found to have fed on human blood, the present study allows no comment on mosquito preference for different host species to be made as the mosquitoes did not have equal access to all hosts studied. The most frequent mosquito genera captured were Culex and Aedes, as had been observed before, in reverse order, in a nearby area using different capture methods (Labarthe et al 1998a, Souza 1999). In the present survey Culex species (61.2%) were more frequently captured than Aedes species (29.5%) suggesting that cats are possibly less attractive to Aedes than other host species. It is known that when Aedes genus mosquitoes find canine and feline hosts side by side they are more attracted to dogs than to cats (Labarthe et al 1998a), nevertheless, when cats are abundant and exposed day and night, they may become an important blood meal source for these eclectic genera of mosquitoes. In the present survey, most mosquitoes of both genera (Culex and Aedes) were found with feline blood in the digestive tract (72/109). The other genera (Coquillettidia, Mansonia, Psorophora and Wyeomyia) although not previously considered as potential vectors of heartworm in the area for dogs or cats, were captured in low numbers and were also shown to have fed on feline blood (Labarthe et al 1998b). The small number of mosquitoes captured made it impossible to analyze the ability of mosquitoes to transmit heartworm.

As canidae are the known microfilaria source for mosquitoes, in order to guarantee transmission of Dirofilaria immitis to cats, mosquitoes must feed on dogs and cats. Therefore, the finding that Aedes (Aedes albopictus, Aedes scapularis, Aedes taeniorhynchus), Culex, Coquillettidia, Psorophora and Mansonia individuals had fed on canine and feline blood makes them potential vectors for feline heartworm as well as any other mosquito-borne disease passing between dogs and cats. Consequently, veterinarians of the area must address chemoprophylaxis for cats and should include heartworm in their differential diagnoses when cats exhibit appropriate clinical signs.