Survival of Juvenile Acer grandidentatum Nutt. (Bigtooth Maple, Aceraceae) in Central Texas Woodlands

Populations of Acer grandidentatum Nutt. (Bigtooth maple, Aceraceae = Sapindaceae) in central Texas are mostly found in isolated, deep, relatively remote, limestone canyons. Acer grandidentatum is found with a few other mostly deciduous species. Recruitment of juveniles has been reported to be lacking. One population of A. grandidentatum juveniles was found in a limestone canyon in a State Natural Area in Central Texas. Fifty juveniles were located. Wire enclosures were placed around half of the seedlings with half left in the open. In an adjacent canyon, 50 juvenile seedlings were planted in a similar habitat with adult A. grandidentatum trees nearby. Half were in enclosures and half in the open. Plant survival was followed for four growing seasons until November 2019. At the end of that time when survivals were compared between plants in enclosures and those in the open in both canyon communities, there was a significant difference in survival in both communities (χ, P < 0.001). Survival in the planted population in enclosures was 92% with 52% in the open. In the natural population in enclosures 68% survived, with 32% survivals in the open. When population extinction was compared, extinction for planted juveniles in enclosures using linear regression was 60.5 yrs and it was 11.9 yrs in the open. For native juveniles, it was 12.6 yrs in enclosures and 5.9 yrs in the open. The cause of most mortalities in the open seemed to be herbivory by white-tailed deer (Odocoleus virginianus).

cannot be understood unless light, nutrient and water levels are understood in addition to effects of competitors, herbivores and carnivores [9] [10] [11] [12].
Changes in species composition have been reported for many North American woodland and forest communities and are widespread and species independent including lack of replacement of mature species [11] [13] [14] [15]. Lack of replacement is not caused by lack of germination but occurs at later growth stages and was not a problem prior to Europeans arriving in North America [16] [17].
The topic of the current study, A. grandidentatum, is widespread in the mountains in New Mexico, western Texas and other mountains of the western United States [18] [19] [20] [21]. It is also found in central Texas and southwestern Oklahoma [20] [21] [22]. Anecdotal reports suggest populations of A. grandidentatum in central Texas are not being replaced [23] [24]. There is one limited experiment showing complete loss of non-protected A. grandidentatum juveniles due to large animal herbivory in one growing season [12]. A structural study demonstrated that A. grandidentatum populations had inverse quadratic size distributions with few juveniles and very little possible recruitment over the 20 years prior to the study [25]. Acer grandidentatum has high seed germination, grows best at medium light levels and can carry out CO 2 uptake at low light levels (shade) [12]. However, we have not found other specific studies concerning population dynamics of A. grandidentatum.
There are studies that have shown lack of recruitment of the mature forest trees in eastern North America [13] [26] [27] [28], in the mountains of western North America [11] [29] [30], and in woodlands of Texas and the mountains of southwestern

Materials & Methods
Site description. This venture was carried out on the 1520 ha (= 3757 ac) Albert and Bessie Kronkosky State Natural Area which is in Kendall and Bandera counties, central Texas from April 2016 through November 2019. This area is in the Edwards Plateau Physiographic region of central Texas (Figure 1, approximately 29˚44'25''N, 98˚50'18''W) [42]. The study locations were specifically in the "Tin Cup Canyon" and the canyon that is the source of water for Bessie's pond [43].  [44]. Acer grandidentatum had the highest density and relative density (52%) of overstory plants in the canyon communities [25]. Other overstory species found in the canyon communities in descending order of relative density included Juniperus ashei (a gymnosperm), Prunus serotina (black cherry), Quercus laceyi (Lacey oak), Q. buckleyi (Texas red oak), Tilia caroliniana (Carolina basswood), and a few other mostly deciduous species. The approximate mean annual temperature in the study area is 18.3˚C, and ranges from near 0.7˚C in January to 34.1˚C in August, but is highly variable. The mean annual precipitation is extremely variable as well, but roughly 72.4 cm/yr with very little in July and August with May and September being wettest [45].
Plants studied. In spring 2016 two areas of the ABK Natural Area were surveyed for newly emerged A. grandidentatum Nutt (bigtooth maple, Aceraceae = Sapindaceae) [21], seedlings. The two areas were called Tin Cup Canyon and the canyon that is the source of water for Bessie's Pond. A number of newly emerged A. grandidentatum seedlings were found in the Tin Cup Canyon. They could be told as newly emerged seedlings because the two cotyledons (1 -2 cm long, narrow, green, first leaf like structures) were still attached to the stem. Each was flagged and a numbered tag was placed in the soil 2 -4 cm from the base of the stem. The bottom of the canyon, near adult A. grandidentatum trees, was searched until 50 seedlings were found. The Bessie's Pond Canyon was also searched for seedlings, but unfortunately few were found. We obtained 65 newly emerged seedlings from a local grower. The seedlings were from seed locally collected the previous fall and germinated over the previous winter. These seedlings were in five centimeter peat cups, were approximately seven cm tall with 2 -6 new leaves. Seedlings were planted in 5 -7 cm deep holes dug in the soil in the A. grandidentatum community, in the canyon upstream from Bessie's Pond in the ABK natural area. The seedlings were flagged, tagged and watered immediately and for the next four weeks before the start of the experiment.
Enclosures and Analyses. While the plants were establishing and growing, a series of 50 "rabbit wire", 19-gauge, 1.27 cm 2 galvanized hardware cloth enclosures were constructed. The enclosures were 12 cm in diameter, 48 cm in height and covered with a cap of the same type of wire. The enclosures were transported to the two field sites and then one enclosure was placed over each one of the 25 seedlings to be part of the enclosure treatment. Two seedlings were selected in relatively close proximity and one was haphazardly chosen to be enclosed. The enclosures were next secured to the ground surface using three 60 cm long, 1.5 cm diameter iron rebars pounded into the ground. The plants in the Tin Cup Canyon were inspected once per week for the first month of the study.
The plants in Bessie's Pond Canyon were inspected and watered once a week for the first month of the study to insure they had survived the transplant. Next, the plants were inspected approximately once per month to determine survival. For the next three years the plants were examined intermittently and approximately once per month, except during winter when they were not checked or examined for survival or mortality. In several cases a plant was considered dead (no leaves) early in the growing season but found alive (with leaves) at a later check. When this occurred, the number of mortalities previously recorded was adjusted. The data was summarized in the fall of 2019, but flags, tags and cages were left in place.
Total survival in both locations and both treatments were statistically compared using a χ 2 analysis. Seedling survival in time was examined using a linear O. W. Van Auken, D. L. Taylor regression in spreadsheet software, but there was considerable variability. Consequentially, the data for both sites and treatments were summed over each year and reanalyzed. Annual survival of the seedlings in the four treatments was evaluated for temporal changes with Kaplan-Meier survival analysis [34] [46]. Subsequently, time of population extinction for each treatment population was estimated using simple linear regression analyses.

Results
The number of survivors and mortalities for the plants transplanted into the Bessie's Pond area are shown (Table 1), with the percent survivors, that can also be seen in Figure 2. There were significantly more dead plants in the open treatment compared to those in the enclosure treatment (χ 2 , p < 0.001, Table 1). One plant in the enclosure died (4%) in the first 17 weeks of the experiment vs.    Results were re-examined looking at percent survival per year. Results presented are best fit inverse linear plots (Table 2). Linear regressions were used to estimate population extinctions ( Table 2). The regression equation and R 2 values  are presented ( Table 2) (Table 2).

Discussion
Recruitment of juveniles for a number of woody species has diminished in woodlands and forests in many parts of North America [13] [47]. Juveniles or small plants appear to be inhibited or growth suppressed and not entering the adult population. However, the basis of the mortalities in most cases is only suspected. One study showed 33% survival of planted A. grandidentatum in enclosures for one year when compared to no survivors when plants were in the open, suggesting large animal herbivory [12]. However, the sample size was small. In another project with other species, changes in diameter were demonstrated to be associated with community succession and included high seedlings mortality [48] Acer grandidentatum populations have also been shown to have unimodal size and age distributions with a positive skew and only a very few individuals in the smallest size or age class, suggesting little or no recruitment into the mature population [25]. The juveniles in the enclosures had a high survival and low rate of mortality for the entire study period. The high juvenile survival in the current study, compared to previous studies, is probably due to the lower density of the large herbivores in the natural area studied. In other places in central Texas, the density of Odocoleus virginianus (white-tailed deer) was quite high at 1 deer /5 ha which is as high or higher than anywhere else in North America [49] [50] [51].
In other areas of western North America, the density of Cervus elaphus (red deer or elk) was high causing similar problems [11]. However, in the Albert and Bes-  [53]. Another study showed the western United States, including much of Texas, is in the midst of an unprecedented, long term drought [54] that certainly could have caused limited and sporadic recruitment in some of these populations.
There doesn't appear to be enough fuel in these central Texas canyon A. grandidentatum communities to support fires, although the canyon communities are surrounded by Juniper ashei woodlands that will burn, but only when it is very dry, temperatures are high and wind speed is high (personnel observation, Kerr Wildlife Management Area). It is certainly possible that O. virginianus density and other undetermined factors are related causing O. virginianus to migrate in and out these A. grandidentatum communities, but while they are present, they could consume most or all of the A. grandidentatum juveniles and juveniles of most other deciduous species, resulting in episodic intervals of no recruitment of the juveniles into the mature populations.
There were other herbivores in the area of the study including various rodents, but the enclosures were capped with the same wire as the sides, thus excluding species climbing over to get to the seedlings or juveniles. Feral hogs were present in the area indicated by digging and rooting, but caged or uncaged A. grandidentatum plants were not disturbed. We expected survival would be an inverse exponential function as was the case for other species [34], but found inverse linear functions described survival fairly well. We used these inverse linear equations to show expected time of population extinction and showed significant differences between protected and non-protected A. grandidentatum juveniles, thus suggesting herbivory as the cause of most of the mortalities. However, the time to population extinction was 5.9 -11.9 yrs for populations in the open and 12.6 -60.5 yrs in enclosures and we think this time interval would be long enough for some of these individuals to become established and avoid the possibility of mortality by O. virginianus herbivory. Continued examination of temporal changes in the populations would be judicious to confirm recruitment or extinction.