Invasive non-native trees cause structural and functional changes in plant communities, which tend to increase over time since invasion. Native vegetation responses after control operations provide important information for restoration.

We evaluated the effects of time since invasion and of pine control on plant community structure and on functional traits in a coastal open ecosystem in southern Brazil. We compared richness, diversity, abundance and cover of woody and non-woody native plant species, as well as species composition and community-weighted means (CWM) based on functional traits (dispersal syndrome, fruit type, maximum height, and shade tolerance) of plant communities, in four conditions: a non-invaded area, an area where pines were controlled (managed area), an area of recent invasion, and an area invaded longer ago.

Woody species abundance, richness and diversity declined over time since invasion. However, while abundance recovered to the point of not differing from the non-invaded condition in areas where pines were controlled, species diversity and richness were lower in the managed area than in the area that was never invaded. The effects of pine invasion on richness and diversity of non-woody plants did not increase over time, but plant cover progressively diminished.

Woody and non-woody species composition varied between the four conditions. Species composition similarity was lower between conditions for non-woody than for woody species. CWM differed between the older invasion and the other conditions, determined especially by native plant height and shade tolerance. Taller plants and more shade tolerant native species were exclusively sampled in the older invasion.

Synthesis and application. Pine invasion reduced species abundance, plant cover, richness, and diversity, altering the composition of plant community. The escalation of negative temporal effects of pine invasion was observed on the composition of woody and non-woody species and on functional traits. Although pine control favored the natural regeneration of non-woody species, diversity of woody species in the area submitted to pine control was lower than in the non-invaded condition. Restoration activities are therefore required to increase woody species diversity. These results provide relevant guidance for the restoration of coastal ecosystems following invasive pine control.

Data collection was conducted between August and October, 2018, in four areas between sand dunes with similar environmental characteristics. Four conditions were evaluated: A) area not invaded by P. eliottii, defined as non-invaded (NI – 0.075 ha); B) area previously invaded, where pines were managed in 2013, defined as managed area (MA – 0.075 ha); C) area invaded more recently, defined as recent invasion (RI – 0.025 ha); and D) area invaded for a longer time, defined as older invasion (OI – 0.050 hectares) (Figure 1). The total area of plots was 0.225 hectares. All plots and subplots were set up in areas between sand dunes where the original vegetation was characterized by herbs and shrubs (Guimarães, 2006).

The NI condition represents the control area – i.e., what the other areas would look like if pines had not invaded (Figure 1A). In the MA condition, seedlings were pulled out (< 50 cm height) and juveniles (> 50 cm height) and adults cut down in 2013 - in other words, every pine tree or seedling was removed (Dechoum et al., 2019). All residue of control was left in the area to degrade (Figure 1B). The estimated number of pines eliminated, including adults, juveniles, and seedlings in the area, was 16,000 (ca. 114 pines/ha) (Dechoum et al., 2019). In the RI condition, the herb-shrub physiognomy is still dominant, and the majority of pines consist of seedlings and small to medium size juveniles, as well as some scattered adults (Figure 1C). The forest physiognomy in the OI condition is dominated by adult pines, with scattered native shrubs and low herb cover (Figure 1D). There were no pine seedlings, only a few juvenile trees. None of the invaded areas (RI and OI) had been subjected to previous conversion and/or other management intervention. We postulate that the difference in time since invasion is a consequence of density and age of adult pines planted in private properties in the park surroundings (see subitem 2.1). In other words, there was a higher density of larger/older adult trees in private areas closer to the OI condition compared with the RI condition.

Ninety 5 x 5 m plots were set up in the four conditions. Conditions NI and MA comprised 30 plots each, 10 plots were set up in RI and 20 in OI. The minimum distance between plots was 20 meters. The number of plots varied due to the size of the areas in each of the four conditions.

All native woody plants above 1 meter in height were identified and had their height measured in each plot. Four subplots measuring 1 x 1 m were set up at the vertices of each plot, totaling 360 subplots. All plants between 0.1 and 1 m in height were identified at the species level (whenever possible) and categorized as “woody” or “non-woody”. Plants not identified in the field were collected for later identification with identification keys and taxonomic references, or with support from experts. Among the specimens not identified at the genus or species level, most are in family Poaceae (grasses), which are very hard to distinguish if not fertile, or in families Myrtaceae and Lauraceae, which are two of the richest woody species families along the Brazilian coast, therefore often hard to identify from sterile material (see Appendix 1).

Percentage of plant cover by woody and non-woody species and class of soil exposure were also measured in the subplots. The proportion of soil without live plant cover in the subplots was classified as exposed soil. Percentage of cover was divided in the following classes: Class 1: 0 a 5% (2.5%); Class 2: 5 to 15% (10%); Class 3: 15 to 25% (20%); Class 4: 25 to 50% (37.5%); Class 5: 50 to 75% (62.5%) e Class 6: 75 to 100% (87.5%) (Assumpção & Nascimento, 2000). Median values were used in statistical analyses.

All pines in the RI and OI plots were counted and the perimeter at ground level (PGL) of all trees with PGL ≥ 25 cm was measured. All the stumps remaining after pine control in MA were counted and classified in two size classes: trees with PGL ≥ 25 cm (adults) and trees with PGL < 25 cm (juveniles).

All woody species taller than 1 m in the plots were classified according to four functional traits: (1) dispersal syndrome: anemochory (wind), zoochory (animals) or autochory (self-dispersed); (2) fruit type: dry dehiscent or indehiscent, fleshy dehiscent or indehiscent; (3) maximum height of woody plants measured in the plots; and (4) shade tolerance: tolerant or intolerant. These four functional traits were selected from scientific literature (Reitz, 1965; van der Pijl, 1982; Carvalho, 2003, 2008, 2010; Lorenzi, 2009; Pires et al., 2009; Seubert et al., 2017; Flora do Brasil, 2020).