Using Citizen Science to build baseline data on tropical tree phenology

Large-scale and long-term understanding of the phenology of widespread tree species is lacking in the tropics, and particularly in the Indian subcontinent. In the absence of baseline information, the impacts of climate on tree phenology, and thus on trophic interactions downstream of tree phenology, are also poorly understood. Citizen scientists can help bridge this gap by contributing simple, technology-based information over large spatial scales and over the long term. In this study, we describe an India-wide citizen science initiative called SeasonWatch, with preliminary insights into contributor behaviour and species phenology. Over a period of 8 years, between 2011 and 2019, cumulative contributor numbers have increased every year, although consistent contribution remains constant and low. The phenological patterns in the 4 most-observed species (Jackfruit Artocarpus heterophyllus Lam., Mango Mangifera indica L., Tamarind Tamarindus indica L., and Indian Laburnum Cassia fistula L.) are described, with discernible seasonal peaks in flowering and fruiting. Seasonal peaks are influenced by tree phenology reported in the south Indian state of Kerala, which has the maximum number of contributors and most number of observations per contributor, comprising 89% of all observations. We look in detail at the flowering phenology of one particular species, Cassia fistula, which appears to show aberrant phenology, reflecting a potential shift away from historical baselines. Latitudinal patterns in the phenology of widespread species such as Mangifera indica are also discernible from 4 seasonal bioblitz events organised during 2018-19, with trees in lower latitudes exhibiting flowering and fruiting phenology earlier than the higher latitudes. We conclude that there are signs of shifts in phenological patterns, as in the case of C. fistula, and extend a call for action to sustain long-term interest and participation by contributors to develop a baseline for common tropical tree species that can be used to understand long-term consequences of climate change on tropical tree phenology.

widespread species such as Mangifera indica are also discernible from 4 seasonal bioblitz events 23 organised during 2018-19, with trees in lower latitudes exhibiting flowering and fruiting phenology 24 earlier than the higher latitudes. We conclude that there are signs of shifts in phenological patterns, as in 25 the case of C. fistula, and extend a call for action to sustain long-term interest and participation by 26 contributors to develop a baseline for common tropical tree species that can be used to understand long-27 term consequences of climate change on tropical tree phenology. 28

Introduction 29
Cyclic patterns of growth and reproduction -or phenology -of living organisms are seasonal and highly 30 sensitive to the environment (Sparks and Carey, 1995). Discernible changes outside of the known 31 variability in the phenology of organisms are often indicative of underlying changes in large-scale climate 32 patterns. In temperate regions, higher temperatures are related to the onset of spring phenophases, such as 33 flowering and leaf unfolding (Menzel et al 2006). Increasingly, dates of bud-burst have been 34 demonstrated to advance when compared with long-term averages due to advancement in the warmer 35 season, increase in winter and spring temperatures, and even effects of urbanisation such as pavements 36 and light at night-time (e.g. Menzel et al 2006, Fu et al 2012, Chen et al 2016, ffrench-Constant et al 37 2016. In the tropics, recent long-term phenological studies have shown that the onset, persistence and 38 frequency of reproductive phenophases are affected by solar irradiance (Babweteera et al. 2018;Chapman 39 et al. 2018;Wright and Calderón 2018). In other tropical systems, reproductive phenology has been 40 reported to be affected by precipitation; fruiting intensity often increases with higher rainfall (Dunham et 41 al. 2018;Mendoza et al. 2018). However, the impacts of changes in baseline environmental conditions on 42 tropical tree phenology are poorly understood. 43 The bulk of our understanding of phenological patterns in plants, and their response to climatic change, 44 comes from Europe and North America (Bertin 2008). Apart from contemporary studies, this 45 understanding is supplemented by a number of historical data sets collected by hobbyists and naturalists, 46 6 observed tree. 'Leaf', 'flower' and 'fruit' are further categorised into finer phenological stages (Table 1). 117 The quantity of each phenophase is noted as one of the following -'None', 'Few' or 'Many'. 'None' 118 corresponds to the absence of the phenophase. If a phenophase is observed on 1/3rd or less of the tree 119 canopy, it is marked as 'Few' by a contributor, indicating a lower volume of the tree having the reported 120 phenophase. If a phenophase is observed on more than 1/3rd of the tree canopy, it is marked as 'Many' by 121 a contributor, indicating a larger volume of the tree having the reported phenophase. 'Many' is considered 122 as the phenophase peak at the level of each individual tree. On any given day, only one of these categories 123 can be reported for a phenophase. Along with phenology observations, contributors can optionally note 124 the presence of herbivores, pollinators and dispersers, and also make additional notes on natural history, 125 interactions of interest etc. All phenology data contributed to SeasonWatch are available to any 126 contributor or non-contributor, upon request. 127 Based on these contributed observations, we quantify the following: 128

1) Contributor behavior and data upload patterns 129
We summarise contributor behaviour in the form of overall participation (in making Regular and Casual 130 observations) from each Indian state, and for each year. We quantify contributor consistency as a measure 131 of continuity in Regular observations on registered trees. A contributor is defined as 'consistent' if she 132 makes phenology observations in 23 or more unique weeks in a year. 133

2) Tree behaviour and phenological patterns over time and space 134
We summarise average weekly phenological responses of tree species over a period of 6 years. We 135 assessed Regular phenology data for patterns of phenology using data contributed between 1 January 136 2014 and 31 December 2019. Since contributors are encouraged to record phenology once a week, we 137 also consider the week as the smallest unit of time to report average patterns across species. Tree 138 behaviour was quantified as the average proportion of trees displaying a particular phenophase in a week. reported as 'few' or 'many') with those in full bloom (i.e., 'open flowers' reported as 'many'). The 144 denominator in these proportions is the total number of individuals for which phenology was reported in 145 each week. We visually compare peaks in the proportion of trees showing any flowering with those that 146 were reported as being in full bloom to understand whether overall flowering patterns in C. fistula are 147 different from that which is anecdotally expected. 148

4) Patterns over space 149
Country-wide spatial phenological patterns in tree phenology were based only on Casual observations 150 during the four bioblitz events. We plotted the spatial locations of phenophases of interest in the most 151 observed species on a map to understand latitudinal variation in species phenology. 152 153

Results: 160
We summarise the overall and regional patterns of contributors and species phenology in SeasonWatch.

1) Contributor behaviour and data upload patterns 165
The total number of unique contributors with at least one valid phenology observation has been variable 166 since 2011, with an increase in both individual and schools since 2017 (Fig 1a). The number of consistent 167 contributors has remained constant and low, since 2011, even though there is an increase in total number 168 of contributors (Fig 1b). The south Indian state of Kerala has the largest number of contributors as well as 169 the largest number of observations per contributor (Fig 1c). 170

2) Tree behaviour and phenological patterns over time 171
The species with the most number of observations differed between the two types of observation. 62

Discussion: 206
Our first description of a long term, India-wide citizen science dataset contributes towards a basic 207 understanding of seasonal phenology of common, widespread trees. Across temperate latitudes, historical 208 information on plant phenology has been used as a reference to compare contemporary phenological 209 changes and also identify climatic correlates of these changes. In the tropics, and especially in India, a 210 similar baselines for species phenology across latitudes and seasons is lacking (Ramaswami et al 2019). 211 The citizen science study presented here provides information on four widespread, common tropical 212 species, which can potentially serve as references for future changes in phenology. For instance, in A. The initial description of tropical tree phenology merits further investigation in terms of underlying 244 environmental correlates, such as temperature, precipitation, and photoperiod. Given the large spatial 245 skew in SeasonWatch data, correlational inferences at the scale of the country are likely to be affected by 246 high variability. There is a better scope, however, for exploring temporal patterns of environmental 247 changes on tree phenology, at least from the state of Kerala, with the addition of a few more years of data.

Conclusion: 255
Citizen science has the potential to contribute immensely to mainstream scientific understanding of 256 tropical tree phenology via building baselines. However, a citizen science project such as SeasonWatch 257 faces several challenges, including -ensuring sufficient sample sizes per species per season from different 258 locations, as well as data quality and accuracy. Managing these challenges is especially difficult in the 259 Indian context given the high diversity of languages and the unequal access to the internet, as well as 260 mass and social media, and at times literacy. It is therefore not possible to sustain such programmes 261 without initiating and maintaining long-term partnerships with regional groups and organisations. Despite 262 increasing outreach in multiple languages to promote awareness about the project, providing 263 technological support free of cost for contributors, and simple visual design of the user interface created 264 by taking inputs from contributors during the testing stage, data quantity remains low and skewed to a 265 single state in the country. Furthermore, mechanisms to ensure data quality need to be implemented at 266 multiple stages of data contribution. With this paper, we extend a call for action to sustain long-term 267 interest and participation to develop a baseline for common tropical tree species that can be used to 268 understand long-term consequences of climate change on tropical tree phenology.    N  u  m  b  e  r  o  f  t  r  e  e  s  o  b  s  e  r  v  e  d  p  e  r  s  p  e  c  i  e  s  p  e  r  w  e  e  k  v  a  r  i  e  s  t  h  r  o  u  g  h  o  u  t  t  h  e  y  e  a  r  .  T  h  e  p  a  t  t  e  r  n  s  r  e  p  o  r  t  e  d  f  o  r  s  p  e  c  i  e  s  i  n  F  i  g  3  a  r  e   t  h  u  s  b  a  s  e  d  o  n  v  a  r  y  i  n  g  s  a  m  p  l  e  s  i  z  e  s  p  e  r  w  e  e  k  .  L  o  w  e  s  t  n  u  m  b  e  r  o  f  t  r  e  e  s  o  b  s  e  r  v  e  d  a  r  e  u  s  u  a  l  l  y  b  e  t  w  e  e  n  w  e  e  k  s  1  0  a  n  d  3  0  o  f  t  h  e   y  e  a  r  -c  o  i  n  c  i  d  i  n  g  w  i  t  h  s  u  m  m  e  r  h  o  l  i  d  a  y  s  i  n  s  c  h  o  o  l  s  i  n  s  o  u  t  h  e  r  n  I  n  d  i  a  .  S  a  m  p  l  e  s  i  z  e  s  a  r  e  a  s  f  o  l  l  o  w  s  f  o  r  t  h  e  s  e  s  p  e  c  i  e  s  ;  n  o  t  e  t  h  a  t   Y  a  x  i  s  v  a  l  u  e  s  v  a  r  y  -a  )  M  a  n  g  i  f  e  r  a  i  n  d  i  c  a  ,  b  )  A  r  t  o  c  a  r  p  u  s  h  e  t  e  r  o  p  h  y  l  l  u  s  ,  c  )  T  a  m  a  r  i  n  d  u  s  i  n  d  i  c  u  s  a  n  d