Multilocus Oxford Nanopore sequencing reveals genetic diversity of Giardia duodenalis and Cryptosporidium spp. in surface and wastewater in Southwestern Colombia

Abstract Background Giardia duodenalis and Cryptosporidium spp. are globally significant enteric protozoa responsible for waterborne diarrheal disease with important public health and zoonotic implications. Conventional genotyping approaches based on Sanger sequencing frequently underestimate mixed infections and minority variants, limiting the resolution of environmental molecular surveillance. This study aimed to characterize the presence and genetic diversity of both pathogens in surface water and wastewater samples from the Río Pasto basin (Nariño, southwestern Colombia) using a multilocus amplicon-based Oxford Nanopore Technologies (ONT) sequencing strategy. Methods A total of 102 water samples were collected over seven months (August 2022–March 2023) from three surface water sites representing the upper, middle, and lower Río Pasto basin, and one wastewater collection point. Novel primers targeting the Cryptosporidium 18S rRNA gene were designed and validated alongside previously published primer sets for three G. duodenalis loci (glutamate dehydrogenase, triose phosphate isomerase, beta giardin). PCR-positive amplicons were sequenced on a MinION Mk1C device for 72 h. Reads were quality-filtered (Q ≥ 10) and taxonomically assigned using Centrifuge against curated reference databases; assignments were considered valid at ≥ 1000 reads per taxon per sample. Results For G. duodenalis , 49 of 102 samples (48.0%) produced at least one PCR-positive amplicon. Successful taxonomic assignment after ONT sequencing and quality filtering was achieved in 26/32 (81.3%), 35/40 (87.5%), and 27/33 (81.8%) samples for the bg , gdh , and tpi loci, respectively. In surface water, assemblage A predominated across loci [mean relative read abundance: 80.7% (bg), 53.8% (gdh)]; in wastewater, assemblage B was more prevalent at the gdh locus (mean 51.5%) while assemblage A remained dominant at bg (mean 66.7%). Assemblages C–E were detected at low relative abundances, predominantly via gdh. At the sub-assemblage level, AII was the most consistently dominant lineage across both matrices and loci; co-occurrence of multiple sub-assemblages within a single sample was detected in subsets of samples across all three markers. For Cryptosporidium spp., 67 of 102 samples (65.7%) were PCR-positive, of which 49 (73.1%) yielded valid species-level assignments. C. parvum was the most abundant species (mean relative read abundance: 40.2% in surface water, 32.9% in wastewater), followed by C. andersoni and C. canis . Co-detection of two or more species within a single sample was observed in 47 of 49 assigned samples (95.9%), most frequently involving the combination of C. parvum , C. andersoni , and C. canis (36.7%); this high co-detection rate may partly reflect the limited taxonomic resolution of single-locus 18S rRNA genotyping combined with the sensitivity of high-depth long-read sequencing, and should be interpreted with caution. Concurrent detection of at least one Cryptosporidium species and at least one G. duodenalis assemblage was observed in 27 of 69 assigned samples (39.1%). Conclusions The integration of newly designed primers with multilocus ONT amplicon sequencing provides improved resolution for detecting co-occurring genotypes and species of waterborne protozoa in complex environmental matrices compared to conventional single-locus Sanger approaches. These findings document broad circulation of diverse G. duodenalis assemblages and Cryptosporidium species in the Río Pasto basin, contributing to the genomic surveillance of these pathogens in Latin America. The presence and genetic diversity data reported here may support water quality monitoring and inform risk management strategies under a One Health framework; however, conclusions regarding quantitative exposure risk require complementary data on parasite viability and concentration.