Pathogens may use different routes of transmission to maximize their spread among host populations. Theoretical and empirical work conducted on directly-transmitted diseases suggest that horizontal (i.e., through host contacts) and vertical (i.e., from mother to offspring) transmission modes trade off, on the ground that highly virulent pathogens, which produce larger parasite loads, are more efficiently transmitted horizontally, and that less virulent pathogens, which impair host fitness less significantly, are better transmitted vertically. Other factors than virulence such as host density could also select for different transmission modes, but they have barely been studied. In vector-borne diseases, pathogen transmission rate is strongly affected by host-vector relative densities and by processes of saturation in contacts between hosts and vectors. The parasite Trypanosoma cruzi which is transmitted by triatomine bugs to several vertebrate hosts is responsible for Chagas' disease in Latin America. It is also widespread in sylvatic cycles in the southeastern U.S. in which it typically induces no mortality costs to its customary hosts. Besides classical transmission via vector bites, alternative ways to generate infections in hosts such as vertical and oral transmission (via the consumption of vectors by hosts) have been reported in these cycles. The two major T. cruzi strains occurring in the U.S. seem to exhibit differential efficiencies at vertical and classical horizontal transmissions. We investigated whether the vector-host ratio affects the outcome of the competition between the two parasite strains using an epidemiological two-strain model considering all possible transmission routes for sylvatic T. cruzi. We were able to show that the vector-host ratio influences the evolution of transmission modes providing that oral transmission is included in the model as a possible transmission mode, that oral and classical transmissions saturate at different vector-host ratios and that the vector-host ratio is between the two saturation thresholds. Even if data on parasite strategies and demography of hosts and vectors in the field are crucially lacking to test to what extent the conditions needed for the vector-host ratio to influence evolution of transmission modes are plausible, our results open new perspectives for understanding the specialization of the two major T. cruzi strains occurring in the U.S. Our work also provides an original theoretical framework to investigate the evolution of alternative transmission modes in vector-borne diseases.