This work was designed to study the dynamic evolution of forest landscape patterns in Mao'er Mountain Forest Farm, analyze its changing laws and the driving factors, and further propose reasonable and effective improvement measures which can provide guidance for the sustainable management of forests.

Based on the data of the four forest inventories in 1983,1993, 2004 and 2016, this research used the methods of area transfer matrix, landscape pattern index and structural equation model to analyze the dynamic changes and influencing factors of the forest landscape in Mao'er Mountain Forest Farm.

1) In 2016, the forest land accounted for 86.77% of the whole region, which was the base of the forest farm landscape. The forest land landscape was dominated by soft and hard broad-leaved mixed forest, the forest land landscape generally showed an increasing trend. From 1983 to 1993, the landscape area of forest land decreased by 378.6 hm², and the area increased by 1 730.8 hm² and 499.4 hm² during 1993-2004 and 2004-2016, respectively. Due to the influence of the natural environment and human disturbance, various types of vegetation in the forest land also changed to varying degrees. Compared with 1983, in 2016, the area of soft broad-leaved mixed forest decreased by 1 183.20 hm², the poplar forest area increased by 1 618.89 hm², the hard broad-leaved mixed forest area increased by 1 457.07 hm², and the coniferous broad-leaved mixed forest area increased by 767.87 hm². 2) The increased forest land mainly came from 1 280.84 hm² of cultivated land, 1 050.46 hm² of afforestation land, 779.14 hm² of sparse forest land and 619.30 hm² of unused land, while the proportion of cultivated land was the largest. 3) At the landscape level, the numbers of patch, patch density, patch shape index, Shannon diversity index and Shannon evenness index all decreased, while the patch association index and the maximum patch index increased, which indicated that the landscape fragmentation decreased and the distribution tended to gather. 4) From Phase I to Phase Ⅲ, the overall effects of natural factors (slope, slope position, and elevation) in the structural equation model were 0.319 6, 0.168 5, and 0.109 2, respectively, and the distance factors (distance from roads, rivers and settlements) were 0.271 2, 0.222 6 and 0.201 0, and the management measures were 0.280 0, 0.030 0 and 0.020 0, respectively. The distance factor had a significant and stable influence on stand types, while the natural factors and management measures had a significant impact in the early stage, which was gradually weakened in the latter two periods.

The landscape area of forest land increases steadily. The landscape distribution becomes more concentrated and the landscape diversity decreases. Natural and distance factors have a significant impact on forest types, and human factors promote the positive succession of forests.

Fraxinus mandshurica, Phellodendron amurense and Tilia amurensis are the main sub top-level community building tree species in the Xiaoxing'an mountains region. The purpose of this research is to analyze the potential distribution areas of F. mandshurica, P. amurense and T. amurensis in the context of global climate change, reveal the impact of climate change on plant distribution patterns, and provide a basis for precise improvement of forest quality.

Based on 19 climate factors, soil, and altitude data, the MaxEnt model was used to predict the potential distribution of three protected tree species in the Xiaoxing'an mountains region under four climate change scenarios (ssp126, ssp245, ssp370, and ssp585): currently and in the future (2021-2040, 2041-2060, 2061-2080, and 2081-2100).

1) The MaxEnt model had a good fitting effect, with AUC values greater than 0.7; 2) The important environmental factors of F. mandshurica in sequence were the annual precipitation, the mean temperature of warmest quarter, and the precipitation of wettest month. The important environmental factors of P. amurense were the annual precipitation, the average temperature of coldest season, and the precipitation of coldest season. The important environmental factors of T. amurensis were the annual mean temperature, the mean temperature of coldest season and the annual precipitation; 3) Under the conditions of future climate change, the area of suitable habitat for F. mandshurica fluctuated significantly; Under the ssp370 climate scenario, the area of suitable growing areas for yellow pineapple decreased the most, while in other cases, the area of suitable growing areas tended to stabilize; The suitable growth area of T. amurensis ranged from 21 769.87 to 25 579.46 km², with a small range of variation and a decreasing trend. The central elevation of the core area of the three tree species increased by 85-183 m under four climate changes.

In order to improve the forest quality in the Xiaoxing'anling area, based on the principle of suitable land and trees, P. amurense can be vigorously planted in Yichun city, and purple linden can be vigorously planted in the southern part of Heihe city and the northern part of Yichun city. Climate change has a significant impact on F. mandshurica, with a small area of suitable areas and a large area of reduced levels of suitable areas. It is recommended to increase the protection of F. mandshurica.