To mitigate the biodiversity crisis and human-land conflicts during rapid urbanization, the construction of "urban forests" and other ecological landscape has been adopted as a common approach for decades, which emphasizes their role of urban green space as species habitat or recreation sites in densely-populated regions. In the research and planning of urban forests and other ecological landscapes, the landscape connectivity associated with the behaviors of certain species (users) has increasingly attracted attentions. However, there are still many problems unsolved in terms of the modeling and the analysis of behavior-integrating landscape structure, as well as the meaningful ways to parameterize the real "landscape connectivity". Based on the blueprint-background transposition idea of negative planning, this paper aims to introduce a new approach to the modeling and the analysis of functional landscape connectivity. Space syntax, a widely accepted principle in the field of architecture and urban planning, has the potentials to facilitate the development of new models on the landscape connectivity integrated with species behaviors. Driven by the theory of space syntax on current stage, this paper introduced the language of artificial intelligence into landscape ecology research, based on the hypothesis that the principle of space syntax still works after replacing the study objects in human settlement with those in ecological landscape. A case study was used to demonstrate the advantages of this new method of studying landscape connectivity. In view of the similarity in spatial configuration between the landscapes inhabited by avian edge species and the objects of axial map study, we used the axial map approach of space syntax to study the landscape connectivity based on the behaviors of avian edge species. Firstly, this paper introduced the traditional connectivity quantification methods and species migration models, summarized their limitations based on literature review, and identified the need of improvements on the individual based models to reflect weighted choice among landscape elements and to predict the messages of interaction between species and landscape in upper scale. Secondly, we put forward the idea of introducing the space syntax model into the field of "edge species-habitat" in landscape ecological research, inspired by the idea of negative planning. Finally, as a case study, we applied the axial map approach of space syntax in the planning research of urban forests in Hezhou city, made a primary test on the "free space" of landscape inhabited by avian edge species and drew the conclusion on the index of integration and intelligibility accordingly. Simulation results were to demonstrate that the model of axial map in the field of behavior-integrating landscape connectivity could reflect the species' migration choice among differently-shaped landscape elements, which has different weights in direction and distance, and could predict the multi-scale behaviors of the species based on the messages they perceived in vicinity.