Individual　movements　within　transport　networks　create　activity　spaces　that　shape　travel　communities.　However,　few　studies　have　examined　the　spatial　structures　within　bus　travel　and　the　associated　factors　across　different　geographic　areas,　which　may　overlook　the　underlying　travel　patterns　and　characteristics　within　these　communities.　Taking　one-month　bus　smart　card　data　in　Beijing,　China　as　a　case　study,　we　first　build　spatial　interaction　networks　for　older　and　younger　adults,　and　conduct　analysis　on　various　network　measures.　Then　we　detect　travel　communities　using　Leiden　algorithm　and　further　investigate　the　determinants　for　bus　flows　across　different　communities　based　on　Poisson　regression　models.　The　findings　indicate　that　older　adults　have　a　shorter　peak　interval,　more　localized　activity　spaces,　lower　network　connectivity,　and　weaker　interaction　strength,　suggesting　limited　mobility　in　bus　travel　compared　to　younger　adults.　The　study　highlights　that　travel　duration　and　land　use　mix　are　important　predictors　for　both　groups　regardless　of　geographic　areas,　and　there　are　also　differences　in　the　factors　influencing　bus　travel　across　various　regional　communities.　The　results　of　this　study　could　better　portray　the　mobility　patterns,　travel　networks,　activity　structures,　and　determinants　impacting　bus　travel　flows　among　older　and　younger　adults,　thereby　providing　nuanced　and　efficient　strategic　support　for　urban　transportation　development.