Disentangling complex Disentangling complex phenotype-environment phenotype-environment relationships— relationships— R. Brian Langerhans R. Brian Langerhans 1 1 , , Lauren J. Chapman Lauren J. Chapman 2 2 , Thomas J. , Thomas J. DeWitt DeWitt 3 3 Diversification of the African cyprinid Diversification of the African cyprinid Barbus neumayeri Barbus neumayeri across water flow and oxygen across water flow and oxygen gradients gradients Funded by EPA STAR and Society of Wetland Scientists (RBL) Funded by EPA STAR and Society of Wetland Scientists (RBL) NSF (IBN0094393 to LJC; DEB0344488 to TJD), Wildlife Conservation Socie NSF (IBN0094393 to LJC; DEB0344488 to TJD), Wildlife Conservation Socie (LJC). (LJC). 2. Concepts 2. Concepts Environmental factors influence phenotypes directly, as Environmental factors influence phenotypes directly, as well as indirectly. Indirect effects manifest due to trait well as indirectly. Indirect effects manifest due to trait correlations and interactions with other environmental correlations and interactions with other environmental factors. Often phenotype-environment correlations are factors. Often phenotype-environment correlations are adaptive. Yet it is reasonable to expect adaptation for one adaptive. Yet it is reasonable to expect adaptation for one environmental factor may constrain or be variously environmental factor may constrain or be variously correlated with other factors, or their phenotypic effects. correlated with other factors, or their phenotypic effects. Such intercorrelations make it difficult to interpret Such intercorrelations make it difficult to interpret simple correlations between trait and environment.. For simple correlations between trait and environment.. For example, one might expect a relationship between example, one might expect a relationship between WF and and body shape due to adaptation for hydrodynamic efficiency. body shape due to adaptation for hydrodynamic efficiency. Ultimately however, the story was more complex in our Ultimately however, the story was more complex in our study: we found that, study: we found that, WF ’s direct effect on body shape was ’s direct effect on body shape was equally countered by a chain of indirect effects, where equally countered by a chain of indirect effects, where increased increased WF increased increased DO , leading to decreased gill size, , leading to decreased gill size, and reduced head size. This resulted in and reduced head size. This resulted in no total effect of no total effect of WF on on body shape body shape . . In the present study (published in J Evol Biol 62:1243- In the present study (published in J Evol Biol 62:1243- 1251), we employed path analysis to examine direct, 1251), we employed path analysis to examine direct, indirect and total effects of indirect and total effects of WF and and DO on morphological on morphological traits of the barb, traits of the barb, B. neumayeri B. neumayeri . . 3. Methods 3. Methods We collected fish from nine populations ( We collected fish from nine populations (map ) that varied in ) that varied in DO and and WF, for which we have monthly data over multiple years). We measured specimen morphology using monthly data over multiple years). We measured specimen morphology using geometric and traditional morphometrics ( geometric and traditional morphometrics (picture ) and performed path analysis to ) and performed path analysis to determine interrelationships among environmental and ecomorphological variables. determine interrelationships among environmental and ecomorphological variables. Player: Player: 5. Conclusions 5. Conclusions WF and and DO influenced relative gill size, body shape and influenced relative gill size, body shape and caudal fin shape in manners consistent with caudal fin shape in manners consistent with a priori a priori predictions. Indirect effects were also noted: (1) strong, predictions. Indirect effects were also noted: (1) strong, oppositely signed direct and indirect effects of oppositely signed direct and indirect effects of WF on body on body shape resulted in a nonsignificant total effect; (2) shape resulted in a nonsignificant total effect; (2) DO had had no direct effect on body shape, but a strong total effect no direct effect on body shape, but a strong total effect via indirect effects on gill size; (3) via indirect effects on gill size; (3) WF indirectly indirectly influenced gill size via effects on influenced gill size via effects on DO . Only through . Only through examination of multiple environmental parameters and examination of multiple environmental parameters and multiple traits can we hope to understand complex multiple traits can we hope to understand complex relationships between environment and phenotype. relationships between environment and phenotype. 1. Natural history 1. Natural history This barb inhabits a range of systems This barb inhabits a range of systems within the African within the African rift lake basins, from from dense swamps to fast flowing rivers, dense swamps to fast flowing rivers, habitats that vary dramatically in water habitats that vary dramatically in water flow ( flow (WF ) and dissolved oxygen ( ) and dissolved oxygen (DO ). ). Because Because DO is generally positively is generally positively related to related to WF , it can be difficult to , it can be difficult to separate effects of the two variables as separate effects of the two variables as they impact aquatic organisms. However, they impact aquatic organisms. However, in our study system, much of the water in our study system, much of the water flows through papyrus swamps that due to flows through papyrus swamps that due to heavy metabolic oxygen demand, can strip heavy metabolic oxygen demand, can strip oxygen from even fast flowing water. This oxygen from even fast flowing water. This effect creates a factorial combination of effect creates a factorial combination of WF and and DO . . Water velocity U U 0.72 - 0.78 1.69 1.08 - 0.01 U 0.60 0.95 U Dissolved oxygen Gill size Body shape Caudal fin shape 1 1 Dept. Biology, Washington University (presently at U. Oklahoma) Dept. Biology, Washington University (presently at U. Oklahoma) 2 2 Biology Dept., McGill University Biology Dept., McGill University 3 3 Dept. Wildlife & Fisheries Sciences, Texas A&M University Dept. Wildlife & Fisheries Sciences, Texas A&M University 4. Results 4. Results The principle body shape variation we found involved The principle body shape variation we found involved a shift from low-bodied, small-headed fish to the a shift from low-bodied, small-headed fish to the opposite morphology ( opposite morphology (figure ). ). Papyrus Papyrus swamp swamp Swamp/ Swamp/ river river ecotone ecotone Rive Rive r r Barbus neumayeri Barbus neumayeri When DO is low, my gills grow large, and I breathe like this.