Abstract: 【Aim】 The purpose of this study is to identify the autotomy sites in the German cockroach (Blattella germanica), and to explore the relationship between autotomy and leg regeneration of B. germanica, so as to provide a theoretical basis for the study of insect regeneration. 【Methods】 The healthy nymphs of B. germanica at the 3rd to 6th instars were selected and amputated separately at 11 sites of the right hindleg, including the proximal 1st segment of tarsus, proximal 2nd segment of tarsus, joint site of tarsus and tibia, one-third, one half and two-thirds from distal tibia, joint site of tibia and femur, one half of femur, joint site of femur and trochanter, joint site of trochanter and coxa, and base of coxa. The treated nymphs of B. germanica were observed daily to record whether the phenomenon of autotomy occurred. And the time and sites of autotomy, and regeneration or not after molting were also recorded. By using the length of the unamputated left hindleg as the control, the differences in the regenerated legs between autotomy and without autotomy in B. germanica were analyzed and compared, and the relationship between autotomy and leg regeneration was analyzed.【Results】Two autotomy sites were recorded in all the 11 amputation sites of B. germanica nymphs. One autotomy site was at the end of trochanter when amputation was performed at different parts of tibia, the joint site of tibia and femur, and one half of femur, the other was at the end of tibia when amputation was performed at the 1st and 2nd segments of proximal tarsus. There was no autotomy detected in leg amputation treatments at the other sites. The autotomy sites were decided by amputation sites but not affected by nymphal instars. At different amputation sites with the same site of autotomy, there was a positive correlation between the degree of amputation and the probability of autotomy in the same nymphal instar. While there was a negative correlation between the nymphal instars and the probability of autotomy when leg amputation was performed at the same site. Autotomy did not affect whether regeneration happened or not but influenced the sites of regeneration. When autotomy occurred at the end of trochanter or at the end of tibia, a completly new leg was regenerated or tarsus was regenerated at the end of tibia. When autotomy did not happen, regeneration occurred at the amputation sites. At the same time, the length of the regenerated leg of the individuals under autotomy was significantly longer than that of the individuals without autotomy, this phenomenon was more obvious when the leg was amputated at the joint site of tibia and femur and at two-thirds from distal tibia. It was more coordinated for the proportion of regenerated legs under autotomy than that without autotomy, and the length of sensilla on regenerated legs under autotomy was much longer than that without autotomy. 【Conclusion】B. germanica can optimize the regeneration by autotomy. There are two autotomy sites at the end of trochanter and the end of tibia, respectively, and the regenerative ability is strong at these two autotomy sites. B. germanica faces a choice between autotomy and limb salvage when its leg is amputated: when limb autotomy can optimize the length and sensilla integrity of the regenerated legs, B. germanica prefers to choose autotomy, while when the regeneration can not be optimized by autotomy, no autotomy will happen.

Key words: Blattella germanica, leg amputation, autotomy, regeneration, optimization