Hydrological Effects on the Reproduction of the Giant South American River Turtle Podocnemis expansa (Testudines: Podocnemididae)
Understanding the relationship between species and the environment is crucial to predicting their responses to human-induced global changes, i.e., habitat conversion, biological invasions, and global warming. Precipitation and river level are relevant factors that regulate the populations of aquatic organisms. We used long-term data to assess the effects of climate on nest number, clutch size (number of eggs per nest), hatching success, and unviable eggs of the Giant South American River Turtle (Podocnemis expansa) in a protected area of Brazilian Amazonia. We found a positive relationship between the number of nests and precipitation on headwaters in May. We also observed that clutch size increased when the local river level rose; hatching success increased with rising local river level, mainly during October and November; and egg failure increased with rising headwater river level. We show how precipitation and river level (at local and headwater) can influence reproductive success in P. expansa, highlighting the perils of human-induced environmental changes.
Reproductive parameters (nest quantity, clutch size, hatching success probability, death probability, and unviable eggs probability) variation among the years of Podocnemis expansa from five beaches (Canguçu, Coco, Comprida, Goiaba, and Jaburu) in the Javaés River, Pium, Tocantins, Brazil. Black lines and gray shades represent local polynomial regression (LOESS) estimates and standard errors.
Biplot of the most important principal components (PCs) that explained nest quantity (number of females) of Podocnemis expansa in the Javaés River, Pium, Tocantins, Brazil. The PCs come from the first principal component analysis (PCA1) that summarized the spatio-temporal envelope of the hydroclimate. The red circles depict the most important hydroclimate variables from each PC that best explained nest quantity. Avg = average; Min = minimum; Max = maximum; Apr = April; Hdwt = headwaters; Precip = precipitation; RivLev = river level.
Relationship between nest quantity of Podocnemis expansa from five beaches (Canguçu, Coco, Comprida, Goiaba, and Jaburu) and precipitation variables in the Javaés River, Pium, Tocantins, Brazil. Black lines represent the regression estimates from the generalized linear mixed-effects model. Marginal R2 refers to fixed-effects only and conditional R2 to fixed- and random-effects.
Biplot of the most important principal components (PCs) that explained average clutch size of Podocnemis expansa in the Javaés River, Pium, Tocantins, Brazil. The PCs come from the first principal component analysis (PCA1) that summarized the spatio-temporal envelope of the hydroclimate. The red circles depict the most important hydroclimate variables from each PC that best explained average clutch size. Avg = average; Min = minimum; Max = maximum; Apr = April; Hdwt = headwaters; Precip = precipitation; RivLev = river level.
Relationship between average clutch size of Podocnemis expansa from five beaches (Canguçu, Coco, Comprida, Goiaba, and Jaburu) and precipitation and river level variables in the Javaés River, Pium, Tocantins, Brazil. Black lines represent the regression estimates from the linear mixed-effects model. Marginal R2 refers to fixed-effects only and conditional R2 to fixed- and random-effects.
Biplot of the most important principal components (PCs) that explained hatching success and unviable egg probabilities of Podocnemis expansa in the Javaés River, Pium, Tocantins, Brazil. The PCs come from the second principal component analysis (PCA2) that summarized the spatio-temporal envelope of the hydroclimate. The red circles depict the most important hydroclimate variables from each PC that best explained hatching success and unviable eggs probabilities. Avg = average; Min = minimum; Max = maximum; Sep = September; Oct = October; Nov = November; Hdwt = headwaters; Precip = precipitation; RivLev = river level.
Relationship between hatching success of Podocnemis expansa from five beaches (Canguçu, Coco, Comprida, Goiaba, and Jaburu) and river level variables in the Javaés River, Pium, Tocantins, Brazil. Black lines represent the regression estimates from the generalized linear mixed-effects model. Marginal R2 refers to fixed-effects only and conditional R2 to fixed- and random-effects.
Relationship between unviable eggs probability of Podocnemis expansa from five beaches (Canguçu, Coco, Comprida, Goiaba, and Jaburu) and river level variables in the Javaés River, Pium, Tocantins, Brazil. Black lines represent the regression estimates from the generalized linear mixed-effects model. Marginal R2 refers to fixed-effects only and conditional R2 to fixed- and random-effects.
Contributor Notes
Associate Editors: J. M. Davenport and J. D. Litzgus.