Long-term reproductive phenology in the Chihuahuan Desert linking legacy and contemporary datasets

Arid and semi-arid regions of the southwestern USA are especially sensitive to changes in temperature as well as drought frequency and intensity. Phenology is an integrated and salient indicator of plant responses to climate change. We examine an 18-year dataset of monthly observations of reproductive phenology for three species of perennial grasses (Bouteloua eriopoda, Sporobolus flexuosus, and Muhlenbergia porteri) and one deciduous shrub (Prosopis glandulosa) on the Jornada Basin long term ecological research (LTER) site in southern New Mexico, USA. We compare and contrast timing and duration of flower and fruit production for the three grass species with the co-existing shrub that has historically displaced grasses in this system (P. glandulosa). We examined field observations of reproductive phenology made monthly along two 75-m x 1-m belt transects to yield counts of individual plants in each of five phenophases:

1. Dormant
2. Leaf-out with no reproductive structures
   and producing
3. Flowers
4. Seeds
5. Fruits for each species/site combination

The appearance of first fruit for grasses occurred consistently in August or September, although the number of plants producing fruit was highly variable from year to year. P. glandulosa demonstrated remarkable synchrony in the production of first leaves (in April), flowers (in May), and fruit (in June) across four sites. The appearance of first fruit and peak fruit production was most often in the same month for P. glandulosa and S. flexuosus. The time from first to peak fruit production in B. eriopoda was one month at two sites and in the same month at a third, suggesting local variability in rainfall or soil moisture dynamics may play a role in peak fruit production. To bridge LTER monthly data on reproductive phenology with phenology observations at finer temporal resolutions, we are performing a cross-walk of monthly LTER observations with observations using NPN protocols conducted weekly and pheno-cam greenness indices captured daily. We present our study design as an expansion to the Jornada LTER phenology study protocol that could be extended to other sites within the LTER network to facilitate improved predictions regarding plant responses to changes in climate.