Between species mating trial data This file contains the data for time spent with conspecific (Con) or heterospecific (Het) mates when focal fish had the choice between both. Time measured in seconds (out of 600 second trials). Allopatric and Sympatric focal fish compared. Data for Figure 2 of the paper. claremalevf3Rev.txt Between population female mating trial data This file contains association time (in seconds) for female L. parva given the choice between males of their own population or males of a foreign population. Foreign populations could either be from the same salinity environment (ecology) or different salinity. Females came from populations that were either sympatric with L. goodei or allopatric. Data for Figure 3 and Figure 5. GRdata3Rev.txt Between population male mating trial data This file contains association time (in seconds) for L. parva males choosing between females from their own population or a foreign population. Focal males were from populations sympatric with L. goodei or allopatric. Data from Figure 4, Figure 5, and Figure A2. amymRed.txt Between population egg laying data 2 hours Number of eggs female L. parva laid with males from their own or foreign populations over a 2-hour period with each male. Data for Figure 6. eggs.txt Between population egg laying data 10 days Number of eggs laid over 10 days for sympatric Indian River (IR) females and allopatric Texas (TX) females in no-choice trials. Data for Figure A3. IRTX_premate2.txt Male morphology Standard length and total anal fin area as measured from pictures of L. parva (IR,SJ,GC,LB) and L. goodei (UB) males. Data for Figure A4. parvamales2.txt Latency to mate between species Latency to mate (in days) when Indian River L. parva (IR-sympatric) were paired with Blue Springs L. goodei (BS-allopatric). Data for Figure A5. BSIRlatency.txt

Reinforcement occurs when behavioral isolation is strengthened between species due to selection against hybridization in sympatry. Mate preferences and their target traits may change in sympatry as a consequence of reinforcement. This can potentially generate further behavioral isolation within species if sympatric populations evolve extreme preferences or traits that cause them to reject individuals from foreign populations as mates or be rejected as mates. This process is known as cascade reinforcement. We measured behavioral isolation between sympatric and allopatric populations of Lucania killifish to determine whether isolation evolves due to reinforcement between species and whether reinforcement affects preferences within species, consistent with the cascade reinforcement hypothesis. We measured mate preferences in both sexes between species (Lucania parva vs. Lucania goodei) and within species (among populations of L. parva). Between species, both male and female preferences for conspecifics were highest in sympatric populations. Within species, L. parva females from sympatric populations preferred their own native males over foreign males. Allopatric L. parva females and all L. parva males showed no preferences within species. Within species, behavioral isolation showed no association with ecological variables, such as salinity. Thus, reinforcement is a primary factor generating behavioral isolation in Lucania killifish, creating strong preferences in both sexes among species and leading to cascade reinforcement of female mate preference within species.