Home | Business News | Browse by Publication | G | Georgia Journal of Science

Preliminary investigation into the use of logistic regression to predict parasite intermediate hosts; Case study: Dujardinascaris waltoni (Nematoda: Ascarididae) in the American Alligator (Alligator Mississippiensis).

Article Excerpt
ABSTRACT

We present here a case study that illustrates the potential for the use of logistic regression to facilitate identification of a parasite's intermediate host. We used a full factorial logistic regression to estimate the probability that adult American alligators (Alligator mississippiensis) were infected with the ascarid nematode Dujardinascaris waltoni given the alligators' physical characteristics and stomach contents. The strength of association between a fish-based diet and infection predict fish as the intermediate host taxon. Significant but weaker associations with reptile-mammal and reptile-crustacean diets are likely due to the high percentage of alligator remains present in the stomachs examined, with infection likely the result of vertical transmission of the parasites, not the ingestion of these other prey categories. This case study strongly suggests logistic regression has the potential to determine parasite intermediate hosts if definitive host stomach contents can be routinely and adequately sampled and the food items identified to taxa.

Key words: Alligator mississippiensis, American alligator, ascarid nematode, diet, Dujardinascaris waltoni, gastrointestinal parasite, intermediate host, logistic regression.

INTRODUCTION

The study of parasitic nematodes is complicated by the appearance of heteroxeny (use of intermediate hosts). Knowledge of a complete life cycle is necessary in order to analyze parasite population structure. Furthermore, population structure is of significant importance in regards to control of specific parasite infections because of the ability to generate a theoretical frequency distribution of the parasite among hosts. Without knowledge of intermediate hosts, strategies for interrupting the life cycle are incomplete and may be based on subjective rather than objective data. Because mathematical modeling has been used to formulate quantitative descriptions of parasite life cycles in regards to population structure (1), we postulated that other mathematical techniques, such as statistical applications, could help in elucidating further aspects of population structure. To test this hypothesis, we selected the American alligator (Alligator mississippiensis) and its nematode parasite Dujardinascaris waltoni (Phylum Nematoda, Family Ascarididae) as a case study in the predictive abilities of statistics to identify the unknown intermediate host.

Numerous accounts have documented the helminth communities of the intestinal tracts of American alligators (2-6) and other crocodilians (7-13). Most of these studies are systematic or taxonomic in nature and provide little insight into the factors that influence the likelihood of infection or the source of the infection. Clarification of relationships between infection intensity and the infected animal's physical characters (sex, age) (5) or diet (9) may be useful in identifying which individuals within a host population are most at risk of infection and subsequently predicting the population-level impacts of the infection. In addition, because infection with gastrointestinal parasites is primarily through consumption of infected prey, a relationship between presence of infection and diet may aid...